Mag. phil. Manuel Sedlaczek

“Conjectures on the Origin, Spread and Development of the Lingua Franca”

A historical linguistic study

DISSERTATION

zur Erlangung des akademischen Grades

Doktor der Philosophie

Studium: Anglistik und Amerikanistik

Alpen-Adria-Universität Klagenfurt

Fakultät für Kulturwissenschaften

Betreuer und Erstgutachter: Em. Univ.-Prof. Mag. Dr. Allan Richard James Institut für Anglistik und Amerikanistik Alpen-Adria-Universität Klagenfurt

Zweitgutachter: Univ.-Prof. Mag. Dr. Alexander Onysko Institut für Anglistik und Amerikanistik Alpen-Adria-Universität Klagenfurt

Oktober 2016

i Affidavit

I hereby declare in lieu of an oath that

- the submitted academic paper is entirely my own work and that no auxiliary materials have been used other than those indicated,

- I have fully disclosed all assistance received from third parties during the process of writing the paper, including any significant advice from supervisors,

- any contents taken from the works of third parties or my own works that have been included either literally or in spirit have been appropriately marked and the respective source of the information has been clearly identified with precise bibliographical references (e.g. in footnotes),

- to date, I have not submitted this paper to an examining authority either in Austria or abroad and that

- the digital version of the paper submitted for the purpose of plagiarism assessment is fully consistent with the printed version.

I am aware that a declaration contrary to the facts will have legal consequences.

(Signature) (Place, date)

ii TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION ...... 2

1.1 THE STUDY OF THE LINGUA FRANCA ...... 2 1.2 HISTORICAL LINGUISTICS AND ANTHROPOLOGY ...... 5 1.3 THE CREATING OF HISTORY ...... 6 1.4 THE RE -INTEGRATION OF HISTORY ...... 9 1.5 THE IMPORTANCE OF HISTORY ...... 11 CHAPTER 2: PROBLEMS CONCERNING THE LINGUA FRANCA ...... 13

2.1 THE CONCEPT OF ´HISTORICAL RECONSTRUCTION ´ ...... 14 2.2 THE PROBLEM OF HISTORICITY ...... 16 2.3 ATTEMPTS AT SOLVING THE PROBLEM OF HISTORICITY ...... 18 2.4 THE VASTNESS OF THE MEDITERRANEAN (COMMERCE NETWORK ) AND ITS CONNECTION WITH THE LINGUA FRANCA . 22 2.5 THE LOCATION OF ORIGIN ...... 24 2.6 HISTORICO -POLITICAL APPROACH ...... 32 2.7 CLARIFICATION OF THE CONCEPT ...... 37 2.8 HISTORICALLY CHANGING PERSPECTIVES ...... 41 CHAPTER 3: LINGUISTIC ANALYSIS ...... 44

3.1 THE SIMILARITY OF ROMANCE SOURCE LANGUAGE ...... 48 3.2 ORIGINAL ROMANCE SOURCE LANGUAGE ...... 51 3.3 CATALAN -O CCITAN SIMILARITIES ...... 53 3.4 PRONUNCIATION AS A SOURCE LANGUAGE INDICATOR ...... 56 3.5 TEXT COMPARISON ...... 58 3.6 INDISTINGUISHABLE BASIC GRAMMATICAL FORMS ...... 60 3.7 THE COMPLEXITY OF SIMPLICITY ...... 61 3.8 (LIMITATION OF ) EXPRESSIONABILITY ...... 63 3.9 CONCLUSION ...... 63 CHAPTER 4: HISTORICAL ANALYSIS ...... 65

4.1 THE BEGINNING OF COMMERCE ...... 67 4.2 COMMERCE AS A CATALYST OF ORIGIN AND DEVELOPMENT ...... 70 4.3 DATE OF ORIGIN ...... 71 4.4 HISTORICAL AND LINGUISTIC ANALYSIS ...... 75 4.4.1 Earliest Catalan-Arabic language contact ...... 75 4.4.2 Linguistic interpretation of Catalan-Arabic historic contact ...... 77 CHAPTER 5: SOCIO-LINGUISTIC EXCURSUS [AN AUSTRIAN-RUMANIAN PIDGIN] ...... 81

5.1 EARLY DEVELOPMENTS ...... 83 5.2 SUPERSTRATUM LANGUAGE DETERMINATION ...... 85 5.3 LIMITS OF LANGUAGE CONTACT ...... 88 5.4 THE VANISHING OF THE AUSTRIAN -R UMANIAN PIDGIN ...... 90 5.5 LIMITATIONS OF THE AUSTRIAN -R UMANIAN PIDGIN ...... 91 5.6 CONCLUSION ...... 91 CHAPTER 6: THE NATURE OF PIDGINS ...... 92

6.1 CATALAN -A RABIC LANGUAGE CONTACT ...... 94 6.2 IMPORTANT ASPECTS OF PIDGIN LANGUAGES ...... 95 6.3 VOCABULARY AND GRAMMATICAL SHIFTS IN PIDGINS ...... 97 CHAPTER 7: POLITICAL ANALYSIS ...... 106

7.1 PIRACY AND SLAVERY ...... 106

iii 7.2 VARYING LINGUA FRANCA COMPETENCE ...... 108 7.3 SLAVERY NOT RESPONSIBLE FOR EXTENDED PIDGIN STATE ...... 110 7.3.1 Reasons for Lingua Franca downgrade ...... 111 7.4 THE EXPANDED PIDGIN STATE ...... 112 7.5 THE CREOLE (-ESQUE ) STATE ...... 113 7.5.1 Reasons for the Creole and Creole(-esque) state development ...... 115 7.6 VOCABULARY EXPANSION THROUGH MULTI -LANGUAGE INPUT (THROUGH PIRACY AND SLAVERY ) ...... 116 7.7 VARIATIONS IN PRONUNCIATION ...... 117 7.8 POSITIVE ASPECTS OF SLAVERY ...... 118 7.9 MAJOR POLITICAL CHANGES AND FINAL DEVELOPMENT ...... 119 7.10 POLITICAL AND ECONOMIC CONNECTION ...... 122 7.11 POLITICO -ECONOMIC ASPECTS AND UNIQUE LINGUISTIC TRAITS OF RELEXIFICATION ...... 125 7.11.1 Economic facilitation of Lingua Franca use ...... 126 7.12 POLITICAL APPREHENSIONS ...... 127 7.13 THE LINGUA FRANCA DEVELOPMENT AND POLITICS ...... 130 7.13.1 Economic and political interdependencies ...... 132 7.14 POLITICAL -ECONOMIC HISTORIC ANALYSIS ...... 135 7.15 CONCLUSION ...... 139 CHAPTER 8: THE DEVELOPMENT OF THE LINGUA FRANCA ...... 141

8.1 PRACTICAL EFFECTS OF LANGUAGE CONTACT ...... 142 8.1.1 Catalan historical comparison ...... 143 8.2 THE IMPACT OF FREQUENT LANGUAGE CONTACT ...... 144 8.2.1 Attempt at an explanation ...... 146 8.3 PIDGIN DEVELOPMENT BASED ON FREQUENCY ...... 147 8.3.1 Austrian-Rumanian Pidgin comparison ...... 148 CHAPTER 9: A QUESTION OF DEFINITION ...... 150

9.1 DIFFERENT PROCESSES ...... 151 9.2 A MISLEADING THEORY ...... 152 9.2.1 Disproving this theory ...... 154 9.3 CONCLUSION ...... 160 CHAPTER 10: THE GEOGRAPHICAL SPREAD AND DEVELOPMENT OF THE LINGUA FRANCA ...... 162

10.1 14TH CENTURY DEVELOPMENT OF THE LINGUA FRANCA ...... 162 10.2 ANALYSIS OF LANGUAGE INPUT ...... 164 10.3 THE DYNAMIC CYCLE OF LANGUAGE INPUT ...... 166 10.4 LINGUISTIC COMPLEXITY OF LANGUAGE INPUT ...... 167 10.5 FACILITATION OF INTELLIGIBILITY ...... 169 10.6 LINGUA FRANCA VARIETIES ...... 171 10.6.1 The spread of one single Lingua Franca ...... 172 10.7 CONCLUSION ...... 173 CHAPTER 11: THE SPREAD OF THE LINGUA FRANCA ...... 174

11.1 SPREAD AMONG CATALAN SPEAKERS ...... 174 11.2 SPREAD AMONG THE LOCAL POPULATION ...... 179 11.2.1 Spread via merchants ...... 183 11.2.2 Spread via slavery ...... 186 11.2.3 19 th century spread and development ...... 190 11.3 GEO -POLITICAL ANALYSIS OF THE LINGUA FRANCA SPREAD ...... 191 11.3.1 Language border expansion ...... 193 11.3.2 Expansion into cities ...... 194 11.3.3 Expansion into ships ...... 195 11.3.4 Conclusion ...... 196 11.4 SPECIFICS OF LINGUA FRANCA SPREAD ...... 196 11.4.1 Specific place of origin ...... 197 11.4.2 The importance of Algiers (as a specific location of origin) ...... 198

iv 11.5 SPREAD INTO OTHER LOCATIONS ...... 199 11.5.1 Parallel spread...... 200 11.6 LINGUISTIC ANALYSIS AND INTERPRETATION OF SPREAD ...... 202 11.6.1 First language contact ...... 203 11.6.2 Further linguistic development...... 205 11.6.3 Linguistic re-definition of Lingua Franca origin ...... 206 CHAPTER 12: LINGUA FRANCA STABILITY ...... 208

12.1 ARABIC ENFORCING STABILITY ...... 209 12.2 CATALAN ENFORCING STABILITY ...... 209 12.3 RELEXIFICATION AS A SIGN OF STABILITY ...... 211 12.3.1 The process of French relexification and Lingua Franca replacement ...... 212 12.3.2 Colonization as reason for replacement ...... 213 12.3.3 Arab population ensuring stability ...... 215 12.4 ECONOMIC EXPLANATION FOR THE LINGUA FRANCA STABILITY ...... 216 CHAPTER 13: ANTHROPOLOGICAL APPROACH ...... 218

13.1 NOTION OF PURITY ...... 218 13.2 PRIMARY AND SECONDARY SPREAD ...... 219 13.3 GROUP DYNAMICS ...... 219 13.3.1 Dynamics of political-economic systems ...... 221 13.3.2 Group dynamics and origin ...... 222 13.4 THE IMPORTANCE OF THE INDIVIDUAL IN GROUP DYNAMICS ...... 223 13.5 PIDGIN EXPANSION THROUGH INDIVIDUAL DYNAMIC ...... 224 13.6 INTERACTION OF INDIVIDUAL AND GROUP DYNAMICS ...... 225 13.7 THE PROCESS OF UPDATE ...... 226 CHAPTER 14: THE UNDERESTIMATED INVOLVEMENT OF THE JEWISH POPULATION AND JUDEO-ARABIC ...... 228

14.1 PROBLEMS WITH THE THEORY OF LINGUA FRANCA INTEGRATION ...... 231 14.2 AN ALTERNATIVE THEORY ...... 233 14.4 JEWISH MERCHANTS AND LINGUA FRANCA ORIGIN ...... 236 CHAPTER 15: CONCLUSION ...... 239

BIBLIOGRAPHY ...... 240

v Chapter 1: Introduction

The study of Pidgins and Creoles offers a great variety of different topics and issues in the field of linguistics. One of such issues is the designated term “lingua franca” which was defined by the Pidgin and Creole specialist Hugo Schuchardt and basically describes “any widely-spread commercial argot” (Schuchardt, 1979: 32). A lingua franca, therefore, is a language that is known and spoken by a considerable number of people for communication purposes only and therefore is not necessarily spoken with a native-speaker competence. One may think of English as a current lingua franca, as English is widespread in the world and has also become a means of communication for people with different mother tongues. Many people whose mother tongue is different use English as a lingua franca to communicate with people of other nationalities and therefore other linguistic backgrounds. Fascinating as it is, the spoken form of English used as a lingua franca by these people may or may not be similar to the distinctive structure of Pidgins and Creoles itself. However, the term “lingua franca” originally derives from a very old Mediterranean language which was indeed called the Lingua Franca. The Lingua Franca was a commercial or trading language used especially at sea and in the ports of the Mediterranean basin. The Lingua Franca is also said in its turn to form the genesis of Pidgins and Creoles itself. Some theories suggest that the Lingua Franca is the Ur-Pidgin from which all Pidgins and Creoles derive. Ur-Pidgin or not, the Lingua Franca is an essential factor in the study of Pidgins and Creoles and many theories about the role of the Lingua Franca for Pidgins and Creoles have evolved. In any case, the Lingua Franca deserves closer attention and should also be studied seriously by anyone focusing on the analysis of Pidgins and Creoles as such, whether English- lexified or other language-lexified.

1.1 The study of the Lingua Franca

The study of the Lingua Franca is not new. Schuchardt (1909) was the first to seriously study the Lingua Franca, but there is still a lot to be discovered. One such aspect is the question about the origin of the Lingua Franca. Many scholars have made different approaches to this question and have come up with numerous theses, however, not much agreement has been reached concerning its origin. There unquestionably remains the problem of the genesis of the Lingua Franca. It is not

2 clear yet how or when the Lingua Franca came into existence and by what conditions and under what circumstances it was created. Many questions remain unanswered. Many opinions exist and scholars have wondered “how light could be thrown on the whole issue” (Le Page, 1961: 126). One suggestion was that detailed work needs to be done which should be “descriptive and historical” and “every effort should be made to” cover not only the language itself “but also the history and structure of the society” and additionally its development by means of historical and comparative linguistics (Le Page, 1961: 126ff). Because of this state of affairs, further research needs to be carried out and history should be looked at more carefully. Much has been done about uncovering history in the 19 th century, however, there are a few things one should be aware of. Historical writing, or to be more precise, writing about the discoveries of the 15 th century, which will be one important aspect of this paper, started only in the middle of the 19 th century and as a consequence also adapted the approaches of this century. Everything not European was excluded and history was seen exclusively from the European point of view. This “European-centredness” (Chaunu, 1979: 206) had the effect that other countries were only considered after Europe had discovered them. “Europe was the only place which counted” (Chaunu, 1979: 206), and everything outside of Europe was simply ignored. This “European-centredness” will be avoided here and history will be looked at as a whole. The approach will be to give a complete picture of the events and circumstances of the 14th century and even the time before that back to the 11 th century to avoid the impression of a “disjointed history” (Chaunu, 1979: 210). It was only from the middle of the 20 th century that reconstructing history in Pidgin and Creole studies has become a major aspect when “the need to be able to move about easily in time” (Chaunu, 1979: 204) emerged. Until then history had basically been rejected and it was in 1971 when Edwin Ardener, an anthropologist, complained about this ´lack of history´ (1971: 209-241) in linguistics, especially in historical linguistics itself. In my opinion, however, his claim is true until today. It was, of course, not only E. Ardener who complained about the lack of history in recent studies. According to J. Arends (2002), who made a similar claim twenty-five years later, “very little historical linguistic work has been done” (2002: 50) in Pidgin and Creole studies. Arends states that recent studies are rather “sketchy” and only in very few fields has historical work “become more detailed and precise” (2002: 54). Most scholars, with a few exceptions according to Arends, have simply neglected historical aspects even though in most cases the historical evidence is available.

3 Therefore Arends demands “an historically realistic theory of creole formation” (2002: 56). Such a theory, according to Arends, should “be in agreement with the historical facts” (2002: 56). Arends also mentions that historical evidence should play an essential role in the “evaluation of theories based on purely linguistic considerations” (2002: 56). The largely neglected history of current theories is due to the erroneous belief that the evolution and development of a language is (“paradoxically”) regarded “as an ´a-temporal´ process” (Arends, 2002: 58), where time seems to be irrelevant and therefore completely ignored. Arends states that it is not only “unwarranted to maintain this kind of ahistoricism” but also “hardly defensible at any time” (2002: 58). His final statement is that “historical correctness is not a frequently found feature” (2002: 56). Another scholar who asks for more historicity is J.E. Wansbrough (1996), who states that linguistic models are still far too theoretical as they are “methodologically useful but historically exiguous” (1996: 148). Wansbrough also states that the theoretical construct of historical linguistics, the comparative method, has been in use “since this has never proved especially difficult” but it only serves “to provide regularity with a diachronic dimension” (1996: 156). So it may only give (predetermined) answers to predetermined questions and the main theory still remains speculative, as this kind of “reconstruction can never be other than conjectural” (Wansbrough, 1996: 156). Wansbrough therefore calls for the “historical reconstruction of contact” (1996: 151), as this can eliminate the speculative factor to a certain extent and unlikely possibilities can be ruled out. Others who suggest an approach dealing much more with history are e.g. Robert Le Page (1961) who claims that linguistic work should be both “descriptive and historical” as mentioned above, and should cover the “history and structure of the society in which it is spoken” (1961: 127). A substantial history of the origin and development of the Lingua Franca is lacking, and if it is dealt with at all in literature, it is only considered briefly. Because of this “lack of history” in approaches to the problem, the starting point of this survey will be the classical anthropological approach.

4 1.2 Historical linguistics and anthropology

At this point, it should be clarified why anthropology will be used in this study. Anthropology has developed out of many sciences, but one of the major elements is natural history. Natural history is the very basis of anthropology and, at its core, is a science of observation which has greatly influenced anthropology. The method of anthropology is employed because of its strong observational character and because a language and its development in time may be observed. However, an anthropological approach consists of many other sciences such as history, linguistics, sociology and others as well. Furthermore, as anthropology developed out of many sciences it has also developed further into numerous sciences including sociolinguistics and historical linguistics. These sciences may be seen as a combination of anthropology and linguistics. The impact of linguistics on anthropology shows that both share common interests just “as in linguistics and in social sciences the role of language is assumed and asserted” (Dell Hymes, 1971: 49). Anthropological studies share a close connection as a science and can easily be combined with each other. Related subgroups of anthropology like sociolinguistics and historical linguistics have much in common and could easily be combined as well. However, there does not seem to be a cooperation of sociolinguistics and historical linguistics “although there is a great need for that” (Hymes, 1971: 50). A cooperation of sociolinguistics and historical linguistics would lead us to insightful conclusions, because sociolinguistics is the study of language in culture and society and there can be no separation between language and society, and historical linguistics is concerned with the origins of a language. The conclusion is that a connection between sociolinguistics and historical linguistics is desirable, because language is not separable from society. The historical linguistic approach is important to gain insights into a (past) society as it considers changes in society which, as a result, also explain changes in language. Ironically, neither natural history nor anthropology nor its subgroups take history itself into consideration even though anthropology developed in part from the science of history. This trend has continued even when anthropology developed into several other subgroups including historical linguistics. History, or to be more precise, the Classical study of History, is missing in anthropology, sociolinguistics and even in historical linguistics.

This paradox can be ascribed to the anthropology of the 19 th century. The concept of “history”, as well as “historicity” has been rejected by many anthropologists and consequently by historical 5 linguists and sociolinguists because of a misunderstanding. This misunderstanding which was meant to reject models that offered fake historicity actually caused the rejection of history itself. The models that were created out of this misunderstanding rather “created” history than actually reconstructed history. The most prominent model, the Neogrammarian model, failed to predict history and instead created ´a “possible” past but not a “real” past´ (Ardener, 1971: 211). Linguists, then, turned their back on history as well because it could not offer them anything, so the argument went, as a consequence of the lack of competent models. There was indeed a problem with history(icity). Many models in linguistics, especially the Neogrammarian model, were not models of history, as they do not offer history but generate it from forms that were not previously attested. They were a kind of guess, historically speaking, a theory of history without attested facts about history. Therefore, “the Neogrammarian model generates more ´history´ than it puts in” (Ardener, 1971: 215). It produced history and therefore failed history. Historical linguists were, of course, not satisfied with only a guess about history i.e. that there was no real history that could be deduced from this model. Historical linguistic models were also based on the method of reconstructing history, but unfortunately this too was rather a method of construction and did not in itself reveal history. A historical explanation, the most obvious solution, however, was not sought. This, by contrast, will be the main intention of this paper, i.e. to give a historical explanation of the origin, development and spread of the Lingua Franca.

1.3 The creating of history

Many linguistic models offer a historical interpretation and are essentially ´neutral in regard to history´ (Ardener, 1971: 225). E. Ardener (1971) claims that history was rather created than reconstructed. His most essential point is that history “cannot surely be (re)constructed without the historical records” (1971: 225). If we have the historical records, or at least know what has happened in the course of history, we discover that without any historical records we are “replicating” history and the conclusions that we draw about history, in the absence of historical documents, could be the wrong conclusions or, in other words, could not have happened at all. This is the problem of reconstructing history.

Now, we should not totally reject historical linguistics studies as they still offer valuable insights about the past. We must only be aware that many models of historical linguistics, especially the

6 Neogrammarian model and as a consequence the comparative method, are generative. They do not, however, generate real history and there can be no “winning outright in the attempt to generate ´the past´” (Ardener, 1971: 229). History cannot be generated, the past can only be uncovered by “chronologically marked documentation” (Ardener, 1971: 231) and also by historical evidence. We need to bear in mind that “the innovations in a language need not bear any resemblance to what happened historically” (Ardener, 1971: 232). It is a wrong belief that history can be reconstructed from a model alone, so we should not “confuse a model with reality” (Ardener, 1971: 236). All of the above does not mean that historical linguistics and the comparative method it incorporates is invalid. Not everything about this model should be rejected and, more importantly, “what is true of one model of history” (Ardener, 1971: 231) is not necessarily true for another model, it just means that we have to consider this problem of history when we are using such a model, i.e. the comparative method. Hall (1966) also states that not everything about the Neogrammarian model has to be rejected. He claims that its parameters have been “poorly stated” (1966: 114) and that scholars who rejected or criticised this method have done so for the wrong reasons, just as history has been rejected for the wrong reasons. However, the Neogrammarian model has been modified ever since.

The most basic proposition of the Neogrammarian model is that sound change is regular as long as it is not disturbed by other factors. The comparative method, which is the reconstruction of the ancestral forms of related families of a language, is based upon the principle of sound change and by this principle “it was possible to bring order out of chaos in the history of language and to trace the development of linguistic structures and their change through time from a common source” (Hall, 1966: 114). This is now a standard view and also a valid practice in historical linguistics. Languages are considered genetically related if they ´show similarities in morphological and syntactical structures and basic vocabulary´ (Hall, 1966: 114-5). If this is the case we can assume that these languages “must have come from a common source” (Hall, 1966: 114-5). Vocabulary alone is said to be the least reliable source as words can be borrowed very easily. Languages are then related if all of these factors can be found in two or more languages. However, without historical sources we can not be absolutely sure. Linguistics has done much to help to understand how languages work and how they are related, but as we have seen, historical data have been left

7 out even though they are indispensable. Of course this is also valid the other way around. History alone cannot explain all the details about language as only linguistics is able to uncover the details that show patterns and other structural changes. Even though the present study attempts to uncover the history of the Lingua Franca, linguistics will still be an essential part and will explain those things that history itself cannot. Hall (1966) phrases this controversy very well as “it is generally wise to be cautious in assuming the effect of a substratum on the history of language, unless we have specific and detailed evidence to render such an assumption likely” (1966: 113). Hall tells us to be careful of assuming anything without any historical proof and states that history can be an indicator of language relations. However, it is still linguistics that can show what effects a substratum language has had.

As Alleyne (1988) has mentioned, unrelated languages might share the same structures, and without any historical evidence, we might draw wrong conclusions. Hall (1966) further expands on this issue and mentions another example. Similarly to English, many American Indian languages have a type of word formation like the English “compound” words. In many American Indian languages (verb-)compounds are possible. In each language several hundreds of such “compounds” exist. Hall argues that “on the basis of this situation” (1966: 113) a linguist could, several decades later, be misled to believe that there was a substratum influence of Amerindian in English looking at the data several decades later. Such an assumption would be far from the truth. However, if this linguist does not consider history, he might come to such a (wrong) conclusion. Hall’s point is that the only way to know for sure is by resorting to history and to examine the historical possibility and, if possible, also confirm the historical plausibility of an event or development. After all, one could easily be led to the wrong conclusions about the origin and development of a language especially “in the absence of history” (Hall, 1966: 115). If we have historical data we can be quite sure, but in other instances where historical data is lacking, we are not able to make “a clear-cut decision” (Hall, 1966: 123). This is the so-called “problem of historicity”. Our argument so far is that to be certain about the origin and development of the Lingua Franca we need to know history itself. The problem about the Lingua Franca is that we do not know much of its previous history. Before we can jump to any conclusions about possible connections between historical events and the Lingua Franca we need to reconstruct history, i.e. with the help of historical documents. Many scholars have

8 proposed the idea that there might be a connection between the Lingua Franca, a Portuguese Pidgin and Pidgins and Creoles such as Todd (1990), Whinnom (1965, 1971) and Thompson (1961). None of them, unfortunately, have made the attempt to show a clear historical connection or have tried to unfold history. As a consequence, this theory still remains a possibility, a likely idea, a speculation. My argument here is that, if we uncover and reconstruct history and its details, we might just be able to know more and can uncover the origin of the Lingua Franca itself.

1.4 The re-integration of history

Since historical evidence is immanent in the reconstruction of the Lingua Franca, this thesis will start with a historical survey of the Lingua Franca. Furthermore, a sociohistorical description of the Lingua Franca, how it developed, what it was (a trade language of commerce) and how much importance and impact it had on Mediterranean trade will be included. The aim of this thesis is to give a detailed historical description of the events that led to the creation of the Lingua Franca, which will be accomplished by a historical survey of the different developments of the Lingua Franca in the course of history including a historical description of the social circumstances.

As we are using historical linguistics we need to have a closer look at the actual study of historical linguistics. Jan Voorhoeve (1961) treats the problem of historical linguistics extensively. For him historical linguistic studies can be divided into “internal and external linguistic history” (1961: 99). External linguistic history, for him, is the circumstances in which a language came into being. Internal linguistic history is how such a language developed including the process of creolisation which he claims is not taken into consideration very often for the lack of reliable records (1961: 99). Here again we see the problem of historicity. However, we will take into account both, so called internal linguistic history, which in our case will be the starting point of our short survey and also external linguistic history, as we will observe the events that led to the creation of the Lingua Franca through time. However, there is once more the problem of the few reliable sources, the problem of historicity. As we can guess, we encounter the problem of historicity the further we go back in time. Recent history is quite well documented. However, it is difficult to find revealing sources for linguistic aspects a few hundred years back in time. Nevertheless some sources do exist, a fact that has often been neglected in research on the Lingua

9 Franca.

First, the question why history should be studied should be answered too. Many have questioned the significance of history. However, Alleyne (1988) perfectly describes the need for a historical study. He states that there is a huge argument against the study of history and historical origins. The argument, which is used, is that history is “not only difficult to ascertain” (1988: 2), but is also uninteresting as it says nothing about the language and its society in its given cultural framework as there are forms that are similar but not necessarily related with each other. History in itself does not explain anything so, the argument concludes, it is useful rather “to look at how a term functions” within its environment and explain its existence in terms of that environment (1988: 2). As we have seen before, this is the main argument against the study of history and it could not be further from the truth. History is an essential part of a culture, and of course, of language itself. History and language are intertwined with each other and cannot be separated that easily. Completely ignoring history and trying to recreate it by some abstract linguistic formula (alone) does not lead us anywhere. History is not at all irrelevant and no theory or model should ignore this fact. What has been claimed to be “history” is often little more than speculation. We must keep in mind that ´history is a huge part of the present and to know and understand the present, history is necessary´ (Alleyne, 1988: 3). To be able to understand what has happened and how things were back then can give us valuable insights into linguistic developments as well. We need to uncover the past to be quite certain what “evolutionary processes” (Alleyne, 1988: 3) were involved and what factors determined the development of the Lingua Franca. Also to be sure about the present we need a detailed account of history. It is essential to ask where a culture or cultural traits come from, and, in order to be able to answer such questions, we need history. In order to understand, in our case the Lingua Franca, we must study how it originated, and we have to consider its linguistic traits and features. We must study how it works, its social settings, its own mechanisms and its development from the past, as far as we know it, to the present. Only then will we be able to draw insightful conclusions. This has hardly happened in the past. Very few people have attempted such an approach, so very little is known today. This present study is, of course, only an attempt. However, I will try to prove that we can learn a lot from history in linguistic studies. We must study history, specifically the forms that survived and the conclusions that can be drawn from them. The historical

10 perspective, which is the main aim here, is a perspective of continuity and chronological development, so the main focus will be on documentation and mainly written sources. It is important to note at this point that documents are indispensable for this kind of approach. History tells us where to look and, most importantly, language is always worthy of historical study.

1.5 The importance of history

We should now look at an example of why history is important. Keith Whinnom (1956, 1968, 1977, 1984) has suggested that Portuguese Pidgin might derive from Sabir, a supposed version of the Lingua Franca of the Mediterranean. This idea has then been proposed by Thompson (1961) who said “what could be more exciting that we should prove that this Grammar was a development of that of a Mediterranean lingua franca ” (1961: 113). At this point we must state that Whinnom, for his part, has rather merely suggested this, as he has not fully proven his theory. He has written articles about the Lingua Franca which, unfortunately, are rather short and many questions remain, presumably because of the absence of reliable documents. Many other scholars have then taken over this idea, but they as well have not proven it. Sabir and the connection to Portuguese Pidgin is mentioned often, but only mentioned. If this idea is mentioned it is not further developed or reference is made to statements of other researchers like Whinnom (1965, 1971, 1984) or Thompson (1961). The only statement that is repeated over and over again is that Portuguese Pidgin might derive from Sabir. At this point history comes in. If we take only a very short look into history we can easily prove that Portuguese Pidgin did not derive from Sabir. Portuguese Pidgin developed during the age of discovery through the import of slaves into Portugal and then later developed further in the colonies which we can date as the fifteenth or early sixteenth century. Sabir, however, is a different story altogether. This Pidgin developed in the mid nineteenth century when the French established colonies in Algiers. We further know from history that France did not establish colonies before Portugal. In fact, the French followed the Portuguese a full hundred years later. Hall (1966) explains very well that the Pidgin called Sabir could be found along the northern African coast. This Pidgin, however, was a French Pidgin as its form betrays its Southern French origin. The word /sab ir/ as is “shown by the vowel i is corresponding to the close e of other Romance languages” (1966: 6). The Portuguese Pidgin word would be /sábi/ or /sábe/, the Lingua Franca word is /sab ir/ but came into existence several hundred years earlier. This word derived

11 ultimately from “the Romance word /sapé^re/ “to know” “(Hall, 1966: 100). Sabir was a North African French Pidgin that “was formed through contact of speakers of French” and “speakers of several different African languages” (1966: 25). African slaves “had deliberately been mixed together on slave ships, slave markets” and later on the plantations (1966: 25). History does not allow this kind of conclusion as Sabir, a French Pidgin, developed almost four centuries later than Portuguese Pidgin. Also, according to Hall (1966), there were no traces of Sabir in Portuguese Pidgin. These two Pidgins have rather similar forms which could be an indicator that both derived from the Lingua Franca. However, history tells us that Portuguese Pidgin can not have derived from Sabir, so the similarities must have another cause. In fact, Sabir took over the complete Lingua Franca vocabulary, but then, in the mid 19 th century, slowly but entirely replaced the Lingua Franca vocabulary with French vocabulary.

Finally, it has been hopefully shown in detail why history plays such an important role not only of the present thesis but also for historical and sociolinguistics in general. With the help of history, this paper attempts to answer the major question of the origin and development of the Lingua Franca. The sources for the sociohistorical aspect will be based on the work of J. E. Wansbrough, C. Foltys and G. Lang who have dealt in detail with the sociohistorical settings of the Lingua Franca. The basis for the linguistic study can only be based on Hugo Schuchardt and Guido Cifoletti, who have not only analysed several documents, especially Guido Cifoletti, but have also provided indispensible linguistic insights without which this present study would not have been possible. However, there will be reference to the work of others too such as Barbara Collier, Rachael Selbach and Elizabeth Tonkin.

The reason for the high proportion of historical content is that too many assumptions about a possible past have been made, so the attempt will be made to look at what has really happened through historical records and other documents including the external and internal linguistic history. Stating the plausibility of a theory, as many scholars have done, is not enough, as a theory cannot be proven or disproven by simply rejecting something or looking the other way, claiming that it is unlikely or absurd. A detailed look is necessary, and therefore instead of simply producing a possible past, history will be looked at, as it is necessary to look at what really happened. As Wiener has stated when he pointed out the importance of historical study “there is much we must leave, if we like it or not, to the un-´scientific´, narrative method of the

12 professional historian” (Wiener 1948, 1961 edn.: 163-164).

Chapter 2: Problems concerning the Lingua Franca

The Lingua Franca is quite an old language that has existed for half a millennium. However, very little is known about the origin, development and spread of the Lingua Franca and therefore many opinions concerning the origin of the Lingua Franca exist. One opinion is that “the Lingua Franca originated at the time of the Crusades (A.D. 1095ff.) on the Jerusalem battlefields” and continuously spread westward “along the shores of the Mediterranean” (Hancock, 1977: 283) through military and merchant activity as well as piracy and slavery later on in the 16 th century. According to Wansbrough (1996), the Lingua Franca established itself as a language of commerce ensuring communication between the cultures of the Mediterranean. However, despite continuous contact and establishment of the language, very few documents and linguistic evidence have survived about the Lingua Franca. The history of the Lingua Franca itself is quite obscure as there are references to it every now and then, and the probability that such a language was spoken throughout the Mediterranean is quite high, but there is hardly anything concrete to show. The Lingua Franca shares the fate of very many Pidgins and Creoles that are known today that never were “adequately recorded” (Hancock, 1977: 279) or described. It seems the Lingua Franca was such a matter of course on the one hand that nobody felt the need to write about it or to just mention it and was at the same time efficient enough on the other hand that people used it without hesitation. Presumably, the Lingua Franca was not a language of prestige, i.e. not used in science or education and thus neither scholars nor educated people saw the necessity to further indulge in the study of such a language. The Lingua Franca most likely developed out of commerce, as we will see. It was a neutral language that was simple, quick and effective. It was used primarily for trade, but then later it was used for communication between slavers and slaves, and even in the field of diplomacy. The Lingua Franca was practical and effective and this just might be the reason for its “insignificance”. It never was adequately recorded or studied except for the 19 th century when the need was suddenly there. The Lingua Franca has survived centuries without even being noticed or regarded as important which doubtlessly shows the low status of this language, a situation which is known from so many Pidgins and Creoles.

13 2.1 The concept of ´historical reconstruction´

In the course of the historical reconstruction of the origin of the Lingua Franca many uncertainties have arisen due to a lack of documentation, due to the need to resort to other unspecific documents and also due to a lack of interest at various stages in the history of the language. In short, in tracing the origin of the Lingua Franca one seeks a “resolution of an historical problem, that is altogether unsatisfactory,” (Wansbrough, 1996: 25), as many questions remain unanswered. Reconstructing the origin and development of the Lingua Franca through history is a complicated task as many details, unfortunately, remain unclear due to historical gaps created through limited sources or limited documentation itself. Therefore, historical reconstruction of the Lingua Franca is frustrating as “evidence is often meagre or indeed unavailable” (Wansbrough, 1996: 77). The gaps in history are “so frequent and so exasperating” that “the very fact of recurrence might be thought to entail some regularity” (Wansbrough, 1996: 77). The safest conclusion that can be drawn from this recurrence of the Lingua Franca in the course of history is that what is unattested can be regarded as “gaps in a continuum” (Wansbrough, 1996: 77).

What is known about the Lingua Franca can rather be regarded as limited compared to what is unknown due to the historical gaps that permeate the centuries of the Lingua Franca historicity, and according to Wansbrough (1996) the reconstruction of the Lingua Franca is therefore quite often carried out “within these admittedly stringent but ineluctable limits” (1996: 7). Furthermore, even though some sources are available there is still the possibility that these documents do not deal with the matter in question in a satisfactory manner or that what is relevant is not expressed clearly. Therefore, within all these limitations, it is important to “extrapolate beyond what is in the documents” (Chaunu, 1979: 283) and dare to make “bold interpretations” (Chaunu, 1979: 248). In most cases the most necessary documents that are relevant to the origin of the Lingua Franca are missing, either because a document may have been lost in history or is simply not available. In these cases interpretations are one possibility of solving such a problem, especially the problem of the origin of the Lingua Franca. This is where searching for revealing sources needs to be extended to history itself as history allows further and, probably, more accurate interpretations. By looking not only at the available (relevant) sources, but at the same time also at the historical events that accompanied these developments, it is possible to make logical 14 interpretations and gain further insights. To make such logical interpretations, “circumstantial evidence” (de Granda, 1976: 13) needs to be consulted. Only by demonstrating the historical plausibility of certain developments is it possible to prove “that the postulated causal series actually could take place” (de Granda, 1976: 17). It is through “historical analysis” (de Granda, 1976: 17) that the obstacles of limited availability and limitation of these documents itself can be overcome. After all, most developments of important historical events, in this case the origin and development of the Lingua Franca, are not a coincidence. The formation of the Lingua Franca was not unintended or merely an accident. The Lingua Franca was the logical consequence of a sequence of historical events and, then, became only one further event in a chain of many other historical developments. What is most essential, and most difficult, is to not confuse the “real conditioning factors with the monetary situation” (Chaunu, 1979: 291). Often clarity about an event is lost in the fog of history due to insufficient documentation, and therefore it is necessary to combine all the available existing sources and affirm their historical plausibility. Only if an event, or a set of events, could really have happened in the course of history can interpretation begin. This is a quite often forgotten problem concerning history.

Wansbrough clearly states that “as nearly always, the real problem is in depicting not the product but the process” (1996: 77). In the case of the Lingua Franca it is not difficult to state its existence, but the long and continuous duration and continuity of its existence and, most of all, an exact statement of the origin of the Lingua Franca or the events or processes that led to its (supposed) creation is hardly possible. Even a statement about the events in history that influenced the Lingua Franca or its course in history is very difficult. What events changed the Lingua Franca and why it was construed as secondary is a very complex question that cannot simply be answered without resorting to history, historical analysis and historical plausibility itself. From the available documents alone it is only possible to state that the Lingua Franca reappears time and time again for centuries, as there are references to the existence of the Lingua Franca in all kinds of documents. However, a clear statement from such documents alone can not be deduced because most of these references are merely side notes or not long enough to give detailed information about origin, development or diffusion of the Lingua Franca. Through documents alone the view of the Lingua Franca is limited and very little can be stated about its existence and even less about its origin. This is why history is such an important aspect in the

15 reconstruction of the origin of the Lingua Franca, because if these small bits and pieces of information that are available about the Lingua Franca are viewed in a historical context and also match with (supposed) historical events and even confirm historical plausibility, then our knowledge about the Lingua Franca is extended and further conclusions can be drawn. To explain, analyse and verify the relevance of historical events to the Lingua Franca is what is so difficult about unravelling the origins and development of the Lingua Franca.

The attempt will be to solve “unavoidable historical questions” (de Granda, 1976: 13). Such historical questions simply arise during the study of the Lingua Franca and it is most important to attempt to answer such historical questions as they are inseparable from history and, of course, the Lingua Franca itself. Even though many questions about the Lingua Franca may remain unanswered, this does not mean that they will or have to remain unanswered forever. Questions usually evoke answers and, in the end, it is answers that we seek, especially the answer about the origin and development of the Lingua Franca.

2.2 The problem of historicity

Even though “for some languages concrete historical evidence is lacking” (Hancock, 1977: 282), through linguistic clues we are able to reconstruct their history such as the time or circumstances of their origin. For other languages it is the other way around. We possess historical evidence of their existence but lack any concrete linguistic data. The Lingua Franca is such a case where documentation clearly proves its existence early in time (about the 16 th century), but where very few linguistic traits have survived. Linguistic documents about the Lingua Franca are sparse and everything before 1600 hardly qualifies as what could be called a “linguistic description” (Wansbrough, 1996: 149). The majority of the written linguistic evidence that is available is not “much older than a century or two” and “early documentation is especially scant” (Hancock, 1977: 278). For the even earlier period there is no substantial linguistic evidence as the Lingua Franca has not been “adequately recorded throughout its development” (Hancock, 1977: 279). For the most part the Lingua Franca is only mentioned or hinted at. These so-called hints are often “reports of pilgrims blessed with the gift of tongues” (Wansbrough, 1996: 148). It may be necessary to mention that at least these hints “are plentiful” (Wansbrough, 1996: 148). These reports are about the lives of saints, crusaders, pilgrims, about expeditions, merchants and

16 garrisons overseas and the like. There are various passages where saints were speaking a lingua franca to communicate with other people along their way. Many of these documents state that often there was no need for an interpreter, as some possessed “the gift of tongues” (Wansbrough, 1996: 148) or had received “the grace of languages” (Hancock, 1977: 285). However, as no examples of these languages are given, it remains unclear if it really was the Lingua Franca or if it was some lingua franca of the time such as “Vulgar Latin” or “Targumic” or some other trade or contact language. As likely as it could really have been the Lingua Franca, there still is no final evidence.

While reconstructing the later development of the Lingua Franca is already a difficult task, it is easy to imagine that it is nearly impossible to reconstruct the earlier period or even the beginnings of the Lingua Franca from linguistic evidence alone. For the most part “the crucial evidence is lacking” (Wansbrough, 1996: 16), which can be explained by the “limited sources” (Wansbrough, 1996: 17) on the one hand and that evidence was lost in the course of history on the other hand. Furthermore, both quality and quantity “of the documentation varies considerably across time and space” (Wansbrough, 1996: 45). Depending on what century is looked at, the sources are either dense or hardly available. However, as a general rule it can be stated that the further we go back in time the less documentation is available. There are also temporal ´lacunae´ (Wansbrough, 1996) and other gaps that might create a false impression, i.e. the impression that there was no continuity in trade and communication in the Mediterranean basin. This image can be ascribed to the “random density pattern” (Wansbrough, 1996: 52) in which the Lingua Franca has been documented. However, this so-called “lack of continuity” created due to the lack of documentation is “nothing more than a product of the historical residue” (Wansbrough, 1996: 52). The Lingua Franca has not been recorded adequately enough throughout the centuries and for some time in history evidence, in form of documentation, is lacking completely. However, “it would of course be irresponsible to infer from lack of typical documentation the absence of such activity” (Wansbrough, 1996: 52). From the lack of documents alone we cannot conclude the absence of continuity in the use of the Lingua Franca. Especially not as the Lingua Franca reappears time and time again in the course of history after centuries of missing documented proof. Stating that the Lingua Franca was not in existence in the time where there is no documentation or other evidence available, is unacceptable because “after all we cannot consider the absence of documents in Lingua Franca as proof of its nonexistence”

17 (Whinnom, 1984: 302).

2.3 Attempts at solving the problem of historicity

There are many lacunae and gaps concerning the history of the Lingua Franca and therefore a lot of speculation has been undertaken about the origins and development of it. Linguistic documentation is far too “inconclusive” to allow more than a hypothesis, as “much still remains unproven” (Hancock, 1977: 278). This is also the crucial point, because to reach a conclusion we need more than a hypothesis. After all, “it is nonetheless important to recall that [a] conclusion cannot rest upon a hypothetical construct” (Wansbrough, 1996: 6). A hypothesis based upon linguistic evidence alone, unfortunately, is not enough to reach a conclusion, as the available linguistic evidence does not sustain a final conclusion. Therefore, there has to be an “escape from this linguistic prison” (Wansbrough, 1996: 8). Instead of basing reconstruction of the Lingua Franca on linguistic evidence alone we should seek an alternative. The origin and development of the Lingua Franca “might be constructed by resort to accessible but different data” (Wansbrough, 1996: 17) so that lacunae, or historical gaps, can be “filled from other kinds of source material” (Wansbrough, 1996: 52). Resorting to other documents, which are quite often unrevealing or even ambivalent, is not without problems, but due to the historical gaps the reconstruction of the Lingua Franca also needs to rely upon such ambivalent materials (Wansbrough, 1996: 40). Ironically, this rather leads to an abundance of source texts and although Wansbrough (1996) states that “It is, ironically the abundance of documentation that generates a range of defensible interpretation rather than a single “correct” version” (1996: 6) and thus would consider the various historical documentation about the Lingua Franca rather as a disadvantage. Further source texts do offer valuable insights about a fair range of topics concerning the Lingua Franca and its (social and political) environment. Chaunu (1979) states that “one of the greatest problems is the question of the sources available to us” (1979: 308). One possibility is to resort to poetry and ordinary prose. However, one must be careful when dealing with such ambivalent sources, as it is essential to decide what is important and what is irrelevant. What might be useful and what might not be helpful is a question not easily answered. Literary evidence may be very helpful and easily understandable but it “can also be deceptive” (Wansbrough, 1996: 40). Undoubtedly there is a dilemma one has to face when “dealing with

18 distinct genres” (Wansbrough, 1996: 40) as in this case linguistics and prose or poetry. The problem is of course that these genres allow a combination only to a certain degree, which affects the reliability of such documents. One must be aware, for poetry (including stage plays) at least, that an author may have distorted the linguistic evidence to achieve a comic or dramatic effect or he may have misinterpreted the linguistic evidence due to a lack of knowledge of the Lingua Franca or he may have adapted the Lingua Franca to make his work easily understandable for an audience that had no knowledge of the Lingua Franca. The samples, written in verse form, of the Lingua Franca that can be found in poetry were “hardly ever produced by writers thoroughly acquainted with the language” (Whinnom, 1984: 296). These verses can hardly be interpreted or be regarded as representative of the Lingua Franca as they were on the one hand “written with [the] intent to amuse their Romance-speaking readers” (Whinnom, 1984: 296) and on the other hand only show “the writer’s ignorance and native linguistic habits” (Whinnom, 1984: 296). Poetry does not necessarily reflect the reality of the Lingua Franca in the earlier period. Also, in the 16 th and 17 th centuries it was customary to portray the Lingua Franca in a slightly distorted way. As Schuchardt (1979) says about the 16 th century “this was the century when one loved to hear all manner of broken Italian, naturally with the appropriate caricature, from the stage” (1979: 33). It seems clear from this statement that the Lingua Franca was not (always) portrayed accurately. This was definitely on purpose, just like in the famous play by Moliere “ Le Bourgeois Gentilhomme ”. According to Schuchardt, the inaccuracy of the portrayal of the Lingua Franca in Moliere’s play, and also in other plays as well, serves “to add [a] dramatic effect and which is therefore not very reliable in its details” (1979: 34). Concerning prose, the situation is quite comparable. We are confronted with a similar, but not quite the same problem in a different medium of documents, i.e. in eyewitness reports and travelogues where references to the Lingua Franca and sometimes even linguistic evidence can be found. The travelogue itself is quite an old art form that was ´popular even in the Arabian Empire around the 7 th /8 th century´ (Wansbrough, 1996: 149) where it had its literary peak. The few references and linguistic evidence of the Lingua Franca that one can find in travelogues date from the 7 th to the 19 th century. Travelogues were very popular not only in the Arabian Empire but their popularity continued far outside Muslim territory, albeit in later centuries. Unfortunately, even if there is something written that could be classified as “linguistic evidence”, it is mostly not written

19 by a linguist and presumably distorted, misinterpreted and most likely adapted to the author’s mother tongue. Hancock (1977) mentions that references to such languages were always made by “non-native speakers” who always were “speakers of lexically unconnected languages”, and therefore the languages were “represented in the orthography of the recorders` tongue” (1977: 278). As a consequence “this standardization leads to error”, as pronunciation can easily be mixed up and the orthography of two different languages is never, or hardly ever, the same, which again leads to a distortion of the language described (1977: 278). This is especially true for the Lingua Franca. Another source of prose would be the eyewitness reports that were popular from the 16 th to the 19th century. These reports were mostly writings about the experiences and observations that captured Europeans made during their life in captivity. These reports were often written “with a Eurocentric and one-sided view of Lingua Franca” (Selbach, 2008: 36). What is striking about these reports is that the perception of the Lingua Franca depended on the mother tongue of the author. Those people that were least familiar with the Lingua Franca always wrote more positively and impartially about it. The language available to the author did not only influence the attitude towards the Lingua Franca but also ´the status, perception and even the assessment of the source of the lexicon of the Lingua Franca´ (Selbach, 2008: 44). According to Selbach, French authors more readily suggested French as the main source of the Lingua Franca, Italian authors most often stated that the Lingua Franca was a bastard or broken (Italian) tongue and only authors of other languages were more tolerant in this respect (2008: 44). What is also noteworthy at this point is that only people that had no immediate knowledge of the Lingua Franca wrote about it, especially in the earlier period, as that was the time of travelling abroad and writing about the experience gathered. In the later period the view of the Lingua Franca was always influenced by other factors, which led to a distorted view of the language. Those persons who had knowledge of the Lingua Franca or knew it well enough, hardly ever wrote about it because they did not see the sense in doing so. Other people that also had knowledge about the Lingua Franca never wrote about it because most of them were illiterate, as they belonged to the lowest social classes. The Lingua Franca was at most a peculiarity and never truly studied before 1800. This is quite amazing given the time it had been in existence.

Another possibility is to resort to history itself. “Historical rather than linguistic evidence must provide the principle lead” (Hancock, 1977: 279). Only by reconstructing the early history of the

20 Lingua Franca and concentrating on relevant historical aspects and situations that stimulated its use, spread and development and then comparing it with the linguistic situation can a conclusion be reached. It is most essential to increase the available “historical knowledge” (Hancock, 1977: 279) to establish the necessary “historical links” (Hancock, 1977: 283) in order to be able to connect aspects of history and historical events that were responsible for shaping the linguistic traits of the Lingua Franca. Many linguistic puzzles could be solved if we knew enough history and, of course, what was relevant for the history of Lingua Franca. This could be the key to successfully explain the linguistic peculiarities of the Lingua Franca. The problem that must be identified at this point is definitely the very huge amount of (historical) data that emerges. It is this vast amount of historical data that makes quantifying the data a real problem. The question of what is important and what is not relevant cannot easily be answered. Ironically it is this vastness, this “abundance of documentation that generates a range of defensible interpretations rather than a single ´correct´ version” (Wansbrough, 1996: 6). Furthermore, the interpretation of earlier materials poses an additional problem as these materials “suffer from uncertain chronology” (Wansbrough, 1996: 19). So we have to choose between either the lack of linguistic evidence or the abundance of historical evidence or the adapted version of poetry and prose. Making a choice is definitely difficult. All of these sources are important to reconstruct aspects, peculiarities and features of the Lingua Franca. However, all cause immense difficulties in finding an appropriate, even probable, version of the development of the Lingua Franca. At all times we must remember that historical “reconstruction of contact is of course frustrated by spatial lacunae and a notoriously conjectural chronology” (Wansbrough, 1996: 63). What can be done about all these difficulties is to fill the gaps as well and complete as possible by resorting to historical evidence. After all, historical reconstruction is certainly the best alternative for reconstructing the development of the Lingua Franca, as the historical point of view offers many more perspectives and can more easily fill gaps and lacunae that neither linguistic evidence nor poetry or prose can, due to their restricted documentation. This especially applies to the earlier period, as linguistic evidence is available only from the 15 th century onwards. Therefore history can help a great deal but only if one knows where to look. As was mentioned before, linguistic evidence alone is too sparse to reach a final conclusion. However, if we can reconstruct the history of the Lingua Franca we can use the linguistic evidence that is available to check if our theory can hold.

21 2.4 The vastness of the Mediterranean (commerce network) and its connection with the Lingua Franca

The Lingua Franca has been described by the sources as a Mediterranean language used by merchants (Pananti, 1817 quoted in Cifoletti, 2004: 245) and corsairs alike (Dan, 1637 quoted in Cifoletti, 2004: 204 and Aranda, 1656 quoted in Cifoletti, 2004: 206). Therefore, to discover the origin of this language, the Mediterranean area involved in commerce has to be analysed to unravel the most plausible time and place the Lingua Franca originated in, which presumably lies within one of the Mediterranean commercial networks. However, the problem is that Mediterranean trade from the earliest beginnings, 11 000 BC (Gertwagen, 2014: 154), to the commercial low between the 7 th to 10 th century (Valérian, 2014: 80) and its revival during the 11 th century (Gertwagen, 2014: 159), to the high point of European dominance in the medieval period (Valérian, 2014: 87) and the flourishing of piracy which existed in several commercial networks that existed in parallel but interconnected to each other (Ruiz, 2014 and Kea, 2014 and Doumanis, 2014 and Pollard, 2014) which stretched from Africa and the Sahara (Kea, 2014), to England, Holland and Norway (Gertwagen, 2014), to the Indian Ocean Pollard, 2014), India itself and then as far as China i.e. Asia (Doumanis, 2014), and the Black Sea, Turkey, the Atlantic and even to the New World (Ruiz, 2014). Thus its interconnectedness can only be described as vast and the Lingua Franca being a Mediterranean language becomes difficult to retrace. Even though each of these trade systems was interconnected, the Mediterranean was a “closed system” (Chaunu, 1979: 81) i.e. a (economic) system in which commerce was concentrated and functioned within its own bounds. As all of these networks were closely connected to one another and never truly isolated, trade definitely flowed and merchants were travelling through and thereby connecting them, with some merchants travelling even from one end of the network to another e.g. Jewish merchants (Catlos, 2014: 399). Given the technological advance at this time, it is quite impressive linking one system to the next thus creating the vast trade network that propelled commerce.

This circumstance highlights two main problems (with Lingua Franca studies). First, it is very difficult to pinpoint the origin of the Lingua Franca because the Mediterranean system as a whole, as well as every single (closed) commercial network, had the potential to create a Pidgin like the Lingua Franca (which could be found) in any area and in any century beginning even

22 before the Greek period (within any place of the whole trading network). Second, assuming that the Lingua Franca was used by merchants that travelled across the whole trading network, which does explain why the Lingua Franca is reported in places such as Cornwall (Collier, 1977: 284), the western shores of Africa (Huber, 1999: 15), North Africa, Turkey, and the slave fields in Portugal (Alleyne, 1998: 28) etc., makes pinpointing the (linguistic) origin of such a contact language most difficult. However, according to the sources collected by Schuchardt (1909), Foltys (1984) and Cifoletti (2004) the main agglomeration points were Algiers, Tripoli and Tunis, all of them on the North African shore. Throughout all the sources the (Arabic) North African (shore)cities are always brought into contact with the Lingua Franca, not the European ports or cities, not even the European ships (only the ships of North African pirates and corsairs). However, for finding the origin of the Lingua Franca it is essential to know the time and place where contact (of what kind) plausible for the origin of the Lingua Franca occurred as well as the circumstances that led to or facilitated its creation. The place is important but also the corresponding scenario, i.e. that what was happening culturally, commercially and contact-wise. This might allow a plausible explanation of how two different language speakers came into regular contact with each other, who these different speakers were, i.e. profession, mother tongue, and who brought their vocabulary and who used their grammar that finally created the Pidgin known as the Lingua Franca. This seems almost impossible, especially if the interconnectedness of the Mediterranean with other trading networks such as the Sahara trading route, the Indian Ocean, the Northern Sea and the Atlantic Ocean is considered. Not only was contact intense but also these trading networks fluently flowed into each other so that a merchant from Holland could travel to the Indian Ocean and back. Although quite many of merchants travelled across the whole interconnected area there were, of course, areas in which trade concentrated. Immediately the unavoidable question of the exact location in this vast sea of possible origins, in which the Lingua Franca definitely originated in, arises. To identify the point in time and the location at which the Lingua Franca was created (in), any material that is closely connected to the Lingua Franca is valuable. A main reference text is, of course, the very first linguistic study conducted by Hugo Schuchardt (1979: 26) in the late 19 th century (1880). However, it would not have been possible to study the Lingua Franca so precisely without the source texts about this language. Many people across the centuries have mentioned the Lingua Franca, few people have written a little about it and even fewer people have given examples. However, these source texts allowed the recreation of the main linguistic aspects of the language and thus can be considered as the

23 most essential tool in Lingua Franca studies and it can not be stressed enough how important these texts are for exploring the origin (and development) of the Lingua Franca. Furthermore, the source texts allow the use of another method that can be employed to identify the origin of the Lingua Franca. As there are so many possible dates of origin (that can be deduced from the source texts) it is indispensible to circle the exact date via a linguistic analysis, and analysing the main components (vocabulary and grammar) of the Lingua Franca should make it easier to pinpoint the exact date and place of origin of the Lingua Franca.

2.5 The location of origin

Linguists that have studied the source texts of the Lingua Franca all state that it is, in fact, a Romance-Arabic Pidgin i.e. Arabic grammar with a Romance vocabulary which does, at least, restrict the area of possible origin to the Arabian Empire. Unfortunately, the Arabian Empire stretched over an enormous area of this vast trading network and Arabic was spoken almost everywhere in the Southern Mediterranean (Valérian, 2014: 83 and Doumanis, 2014: 452 and Chaunu, 1979). Therefore, at least the North Sea, i.e. England, the Netherlands, by extension Iceland, and the North-west European shore including France and Portugal can be excluded as well as India and China. However, this still leaves a vast area in which both Arabic and Romance speakers, predominantly merchants, were to be found, i.e. the whole Southern Mediterranean area. The unavoidable question is who of the Europeans came into contact in which part of the Arabic (speaking) world. One approach would be to look at the sources that state where the Lingua Franca was spoken. Fortunately, there are many sources that explicitly state where the Lingua Franca was used, and whenever one source mentions the Lingua Franca it also states what people spoke it or, at least, that it could be used to communicate with a whole variety of different language speakers. The conclusion could easily be reached that the Lingua Franca was spoken widely where it is reported and hardly at all where there are no accounts. Unfortunately, this is not entirely the case, as some sources mention the Lingua Franca in one place, but other sources state the nonexistence of the Lingua Franca in the same place. This circumstance could be interpreted as contradictory or, what is more likely, as the uneven dispersion of the Lingua Franca even in areas where the Lingua Franca was said to be spoken regularly by many (which would explain the different levels of awareness of some speakers over others). One can conclude, therefore, that in the same place the

24 Lingua Franca may be noticed by one person but not by another and this “discontinous valence” (Mallette, 2014: 340) could be due to the Lingua Franca being rather exclusively used by specific groups. The sources mentioning the Lingua Franca always state that people (or groups) who spoke the Lingua Franca were merchants, Christian slaves, Arab/Turkish slavers, corsairs or pirates, and renegades of all nationalities. Most typically, they were to be found especially in Tunis, Tripoli and Algiers, including the local population due to contact to Christian slaves, i.e. captives. The main general areas are ships and consequently cities and ports, and markets which, in most cases, were the ports themselves and the so-called bagnos or prisons used for captives and slaves (Selbach, 2008: 33). There are a few exceptions, i.e. sources that only once mention a place the Lingua Franca was spoken, e.g. in Cornwall (Collier, 1977: 284), although only one person here, a merchant that had been to the Mediterranean basin, could speak the Lingua Franca. Another source indirectly states that during the Crusades one received the “grace of languages” and could communicate with a variety of people, but even if the Lingua Franca is meant, the particular text as a whole is too insubstantial to conclude that it could have been the Lingua Franca, and Hancock further mentions that this is rather Vulgar Latin (Hancock, 1977, 285). Another article mentions that a language called lingua franca was spoken on the west coast of Africa (Huber, 1999: 15) and also ship-logs indicate a possibility for the Lingua Franca to be spoken in West Africa, although by merchants not by a locals (Blake, 1942), which can be explained via the reference to the Sahara route which was described as “the second face of the Mediterranean” (Kea, 2014: 425). Traders would go to the Mediterranean and back to (West) Africa to buy and sell wares, but also slaves (Kea, 2014 and Chaunu, 1979). Looking at the reports and documents about the Lingua Franca, it can be curiously noticed that some locations are described as an integral location for the Lingua Franca, other locations mention the Lingua Franca in some reports but not in others, and some locations never mention the Lingua Franca. There are locations (mentioned) in the reports that always, sometimes or never mention the Lingua Franca, i.e. Algiers and Tunis are always mentioned as being Lingua Franca speaking locations since they were the slave headquarters, while Tripoli and other locations further east are sometimes mentioned to have speakers of the Lingua Franca, as opposed to other reports which do not mention the Lingua Franca at all. The Mediterranean northern basin is hardly mentioned, only Venice in the 18 th century (Cifoletti, 2000); the African West Coast is only mentioned once (Huber, 1999) and the inner countryside of North Africa is not mentioned at

25 all. Most curiously, although some reports never state that the Lingua Franca is spoken anywhere, there are reports that state that the Lingua Franca is spoken in every port (La Condamine, 1731 quoted in Cifoletti, 2004: 216) and, most of all, as a trade language for commerce and its merchants (Cifoletti, 2004 and Foltys, 1984 and Schuchardt, 1909). This peculiarity why the Lingua Franca was distributed unevenly or known irregularly can be explained through the correct background knowledge. The Lingua Franca is unlikely to have been spoken in locations in which it is never ever mentioned (in five centuries). Even if merchants or sailors were travelling to or living in these locations, it can be assumed that the Lingua Franca was not known in such a location, i.e. Cornwall, West Africa, the Sahara, India and China. Locations in which the Lingua Franca is mentioned by some but not by other reports could indicate a Lingua Franca presence which was probably limited to a small circle, i.e. people from or in contact with the Lingua Franca milieu. People that spoke the Lingua Franca were from the lower classes, i.e. people from the cities or “cittadini” (Folty, 1984: 12), and “merchants, sailors, notaries, ship owners, renegades and captives” (Rossetti, 2005: 7). The groups thus associated with the Lingua Franca are primarily merchants, sailors, pirates, renegades and presumably local merchants who had little (language) contact with other groups, even to the rest of the population. It is unlikely they would have heard about the Lingua Franca as it (most probably) existed only in specific places which can easily be overlooked. Those (reports) that all mention the Lingua Franca in a specific location indicate that such a location was definably a centre for the Lingua Franca in which it was more strongly present and spoken not only by merchants but by parts of the local population (of the city, not the countryside). Therefore, two different patterns are recognizable from the sources. For one, the Lingua Franca has been spotted within the commercial limits of the Mediterranean including Cornwall, the Sahara and the Mediterranean shores. The second reoccurring pattern is the enormous stability noticeable in the development of the Lingua Franca. The question of how reports about the Lingua Franca in Cornwall, the Sahara or the West African coast add up arises, and the explanation is that while it was primarily local Arabic speaking traders in the ports and cities that used the Lingua Franca, there were non-European travelling merchants (Jewish, Arabic and Turkish speaking) in the Mediterranean as well. Furthermore, both the Sahara and the West African coast, an extension of the Sahara commercial network, were closely linked to the Mediterranean network with travelling merchants from the Sahara coming to the trading centres (Kea, 2014: 425) where they most definitely acquired the Lingua Franca and brought it back

26 through the Sahara network to the West African coast. However, there the Lingua Franca was really only spoken by merchants that had connections to the Mediterranean (as Christian merchants only found a sea route to West Africa in 1450 (Blake, 1942), which is when the Lingua Franca is mentioned there). As for the Cornwall report (Coates, 1971 quoted in Foltys, 1984: 23 and Cifoletti, 2004: 220-221), it is said that Arabic speaking merchants or travellers became lost and ended up there, which can be explained via the fact that the Mediterranean commercial network stretched as far as the Netherlands and England, i.e. Cornwall, so that it was not unusual that either Arabic speakers were spotted in Cornwall nor that the report states the presence of a travelling merchant from and living in Cornwall who spoke the Lingua Franca. As European travelling merchants were rather the norm, this clearly indicates that merchants were familiar with the Lingua Franca and brought their knowledge back home, regardless if they used it or had opportunities to use it or not, and that people in Cornwall, i.e. non-merchants (locals), did not speak the Lingua Franca. Rather the Lingua Franca was not spoken in England itself, probably not in Europe at all, indeed only by these merchants that were involved in the Mediterranean trade. However, at the same time did the local population at least know about the Lingua Franca or even what the Lingua Franca sounded like. All of this allows the assumption about both the enormous spread of the Lingua Franca, but also that its spread correlates with the trade routes of travelling merchants, as these were the group that primarily spoke the Lingua Franca, had access to the whole Mediterranean (where the Lingua Franca is reported) and were in linguistic contact with other people, i.e. Arabic speakers. The fact that the Lingua Franca might have been spoken on the whole Mediterranean coast is known from the reports and documents collected and quoted in Cifoletti (2004), Foltys (1984), Schuchardt (1909), Selbach (2008), Collier (1977) and Lang, 1992, and according to these documents the Lingua Franca must have been known and used within the whole Mediterranean commercial network (which was mainly sea-based). As almost every merchant and sailor spoke the Lingua Franca, the language was also linked to different groups. These groups could reach any city that had a port, and consequently the Lingua Franca must have been present in the whole Mediterranean and had the potential to be spoken in every port in every city of the Mediterranean. It is known that merchants made stops during their travels, which includes other ports than the major ones (as well), with merchants (and sailors) speaking the Lingua Franca which must have created language situations in which the Lingua Franca was used. However, that this would have left an impact, i.e. the Lingua Franca settling there, is another matter.

27 If the local people in these locations spoke the Lingua Franca as well, is a question for debate, as the majority of the documents do not report or only vaguely mention the Lingua Franca competence of the local population. After all, especially in commerce, language situations can arise in which people communicate with each other through a Pidgin, but as soon as the trade is completed and these people leave this communication situation, the Pidgin is not used again outside of the trade situation. Thus, using a Pidgin language as a means of communication does not necessarily mean that the local population will adopt it as well (as in ´The Austrian-Rumanian Pidgin´, see below). Simply because merchants and sailors spoke the Lingua Franca consistently also does not necessarily mean that people anywhere would adapt the Lingua Franca and use it themselves, i.e. that the Lingua Franca was truly spoken by the whole local population in any port or city in the Mediterranean. That the Lingua Franca was used, however, seems most likely and it also seems very probable that at least in the ports themselves the Lingua Franca was known by some people even if it was not spoken by the whole population. Presumably, the Lingua Franca remained a language for specific purposes among the multilingual repertoires of these speakers that lived more stably in the port cities and engaged in trade. Furthermore, the Lingua Franca would be used among merchants speaking different mother tongues to trade but nothing else, i.e. (only) to communicate with another merchant. The same could be true for sailors who were able to speak it among themselves but had no interaction with other people outside this group that was intensive enough for others to adopt the language. Concluding, the reports could refer to the merchants and sailors themselves stating that basically any of them in the whole Mediterranean could speak the Lingua Franca, which does not include the local population of the Mediterranean cities but at least the local merchants who traded with the travelling merchants, and also sailors (who were to be found everywhere). Thus the reports and documents about the Lingua Franca (being spoken in the whole Mediterranean) mainly refer to specific groups but not the population of the Mediterranean itself. Unfortunately, the documents do not specifically state where in the city, in which milieu or in which places exactly the Lingua Franca was spoken but rather which group used the Lingua Franca. What researchers tend to overlook is that some documents may only claim what seems obvious, that the Lingua Franca was spoken by merchants and sailors but what this further implies is that (consequently) the Lingua Franca must have been spoken in the port areas and markets which is the natural habitat of this group and presumably the place in which the Lingua Franca was mainly spoken. Therefore, this might have caused a situation where those who were

28 writing about the spread of Lingua Franca in the whole Mediterranean were in contact with the ports whereas those that had hardly any contact to this milieu never came into contact with the Lingua Franca. Thus, the Lingua Franca had been overlooked by many authors writing about a city not as accurately as Haedo (1612), Renaudot (1830), Rehbinder (1798) etc. (quoted in Cifoletti, 2004 and Foltys, 1984 and Schuchardt, 1909) (and also) because such people would not have gone where merchants or sailors went. In any case, merchants resided in so-called merchants` quarters (Chaunu, 1979: 262). The concept of the fonduk , a merchant inn, creates the impression that it was a group-exclusive place (Abulafia, 2014: 149), and probably also located in a sealed off area. According to the reports all citizens of these commercial cities knew the Lingua Franca, i.e. merchants, slavers and pirates or sailors, but not the Christian slaves who had yet to acquire the Lingua Franca. This would be an indication that those not involved in trade did not speak or even know of the Lingua Franca as well as an indication that Arabic speakers are the ones that know and use the Lingua Franca quite widely. Therefore, only reports about cities that were frequently visited by Europeans (Rehbinder, 1798 quoted in Cifoletti, 2004: 228) were centres of Lingua Franca use where the Lingua Franca itself was to be found everywhere. Although mentioned as being in use in every port of the Mediterranean and especially on the North African coast, the Lingua Franca only directly appears in some documents, and other reports about these locations never mention the Lingua Franca at all. The Lingua Franca probably existed on a much larger scale in the Mediterranean but remained unrecognized in the documents. However, even if all ports in the whole Mediterranean were places in which the Lingua Franca was spoken (and there is evidence to believe that), it was probably an exception for the local population to adopt the Lingua Franca. This only occurred in the major commercial cities of the Southern Mediterranean (i.e. adopting it into their mother language as the case of the Iberian Jewish population could suggest, based on the correlation between (parts of) the Jewish language and the Lingua Franca (vocabulary) or in the case of slave strongholds such Algiers, in which the local population adopted the Lingua Franca to communicate with European slaves). It could have been possible that the Lingua Franca was used by the local population to communicate with the (travelling) merchants, and thus the Lingua Franca was known but not used to a great extent other than for trade purposes. Maybe this is the reality for many other European and smaller southern Mediterranean cities where the Lingua Franca was spoken but remained limited to the ports and specific groups and may thus have remained unnoticed, as has been indicated above.

29 As a comparison, the situation in Venice of the 18 th century was that the local population definitely knew about the Lingua Franca but not many locals spoke it (Cifoletti, 2000: 14). They were able to recognize it, as merchants from elsewhere in the Mediterranean presumably used the Lingua Franca, but this would not have led the local population to adopt it, but rather to the (travelling) merchants adopting a Venetian vocabulary and expressions thus facilitating a Venetian variety of the Lingua Franca which could have become adapted within the central Mediterranean due to the huge Italian influence. However, as hardly any documents exist, as always, it can not be proven and only remains a speculation. Another possible, but not very likely, explanation would be that people used the Lingua Franca widely, but this had never been properly recorded, because nobody (ever) ventured in those parts and wrote about it, as only the major cities were of interest to travellers and adventurers, not small port cities. Furthermore, local inhabitants that knew about it could have never considered this circumstance important enough to mention or may even have been illiterate in any case. The sources, and there are only a few, that explicitly refer to the local population speaking the Lingua Franca always confirm (one of) the three major Southern Mediterranean cities (Algiers, Tunis and Tripoli). Nonetheless, there is only one (indirect) report about 18 th century Venice, which seems to be an exception, considering that there are no other documents reporting the Lingua Franca to be spoken in any European city. This is not that unusual considering that popular travel reports were aimed at a European audience about the life of non-European cities and not about European cities themselves. These sources state that there existed a close connection between the local population and the Lingua Franca (only) in these specific locations, with almost everyone speaking or at least understanding the Lingua Franca because of its strong presence in the location. According to the documents, the Lingua Franca settled (or became adopted) due to the massive scale contact that existed between merchants, slaves (and sailors) and the local population and thus these locations were becoming centres, or fixities, of the Lingua Franca themselves. This serves as a perfect explanation for the second pattern noticeable about the Lingua Franca in the source texts. The second pattern which is recognizable is the considerable stability of the Lingua Franca, which seems to be tied to the commercial cities of Algiers, Tripoli and Tunis, with Algiers as the seemingly most important commercial city with the highest Lingua Franca presence. Algiers is also the most plausible place of origin of the Lingua Franca as it was the most western trading city and the first major city to be opened up by the Catalan trade network. Reports of Haedo

30 (1612) and Rehbinder (1798), both quoted in Cifoletti (2004), mention that all of the locals speak the Lingua Franca in Algiers, although in other (Arab) cities, i.e. the trading and slave centres (or “hotspots”) the Lingua Franca definitely became a fixity as well. Furthermore, through the reports and source texts it becomes quite clear that the Lingua Franca did not spread into the countryside. In fact, it never left the borders of a city anywhere, a deduction being that the speakers of the Lingua Franca themselves did not leave or use the Lingua Franca outside of the city, which, as a description, fits local merchants and slavers or pirates. Most local merchants lived in the city, and slavers, who were also pirates, must have been in the city as they needed the Mediterranean Sea to conduct their business. Also the travelling (European) merchants only operated in the Mediterranean Sea. Thus it would seem unlikely for them to reach into the mainland. Even the designation “local population” is restricted and can be applied only to the population within a city but not to the countryside or the population of a whole nation. The reason can be found in slavery itself, as it was the reason that promoted language contact between the local population and Christian slaves, and these slaves were kept only in cities, not in the countryside. Hence the influence of the Lingua Franca really never reached outside the city limits. As a whole, the Lingua Franca seems to be closely tied to the slave trade, especially if we look at the source material on slaves, stating that they all encountered the Lingua Franca during captivity. However, these sources were either travel accounts or even diaries about captivity. Furthermore, these reports are all from the 16 th -19 th century when people took interest in the happenings of the Mediterranean, piracy itself had its peak and highest impact on the Mediterranean, and commerce mainly consisted of the slave trade (Selbach, 2008: 33-34), which was paying for the wars countries waged (Abulafia, 2014: 145, 147). Most curious is the fact that ´by the 14 th century slavery was omnipresent and had even become a part of bureaucracy that when a Christian merchant was robbed, the state paid a high percentage back to the merchant´ (Backman, 2014: 174), which, in turn, also shows the importance of trade itself, as refunds were offered to keep commerce flowing. There are, furthermore, the other sources that state that the Lingua Franca was used by merchants and sailors, indicating its strong connection to commerce (Cifoletti, 2004, Schuchardt, 1909 and Foltys, 1984).

31 2.6 Historico-political approach

A rough outline of the political development could give further clues as to which language (Old Catalan or Old Occitan) was used in the region and became the source language of the Lingua Franca. Old Catalan would have had the political influence and (social) backup to be the source language of the Lingua Franca as ´the Crown of Aragon was a powerful thalassocracy which dominated trade via Catalan merchants´ (Abulafia, 2014: 146, 148). Looking at the conditions of the Lingua Franca, i.e. used for commerce, spoken by merchants (and sailors) on ships and in ports, spread across the Mediterranean etc., and comparing this to the Catalan language shows that Old Catalan fulfils the same requirements. Old Occitan, at least concerning the superstratum language influence, could have had some importance as well. However, this was the language of the troubadours who indeed travelled around, but only from royal court to royal court in the European inland, and maybe even England (www.languedoc-france.info/1904_troubadours.htm). Furthermore, the Occitan kingdom was also land-based although there was one major port in Marseilles which had importance in the Mediterranean trade (Valérian, 2014: 84) and “Marseilles was linked to North African trade and corsairing” (Greene, 2014: 97). On the other hand there is no mention of Occitan as a language dominating or spreading out into the Mediterranean. Arguably, as Catalonia and Occitania became one kingdom at some point, one could assume that at this point Marseilles would be used for commerce, maybe even as an extension of Barcelona, and Occitan speakers would venture out into the Mediterranean. However, Catalan became the dominant language (even before this happened). Furthermore, Occitan as a language was so similar to Catalan that there would hardly be any difference noticeable for Lingua Franca speakers, as they could easily communicate with both Catalan and Occitan speakers. Subsequently there would have been no adoption of the Lingua Franca by Occitan speakers. Occitan was also not the language of commerce in the Mediterranean: this was, in fact, Catalan. Evidence can be gathered from the Catalan Consulate of the Sea (1494), a collection of maritime laws written between the 12 th and 14 th century in the Catalan kingdom (Valérian, 2014: 87) which was the predominant law of the Sea, the “ de facto standard for the conduct of maritime matters“ (Backman, 2014: 181) and was composed (and executed) in the Catalan language.

The Catalan kingdom in the 11 th to 12 th century started to gradually expand its territory in the Mediterranean to Malta, later Majorca, Valencia and Sicily, thus establishing commercial and territorial dominance in the Mediterranean. It was also very much engaged in trade although it 32 never established a complete hegemony over commerce, as Italy for instance was also very active in trade (Valérian, 2014: 80). However, Catalan commercial dominance was such a success that the Aragonese monarchy by the 13/14 th century had established a supremacy (Abulafia, 1997 quoted in Valérian, 2014: 82) and with it Catalan had become the official language of commerce itself in almost the whole Mediterranean, at least in the west and central areas, as seen with the Consulate of the Sea (1494). Catalan was also the language used on ships, i.e. for ´merchants and sailors, which were more or less the same person´ (Wansbrough, 1996: 25) as many merchants, in the beginning periods, travelled on their own (Chaunu, 1979: 261). Furthermore, sailors would have had to speak Catalan too to be hired or they had to speak something close, i.e. the Lingua Franca. Geo-politically, the Catalan kingdom expanded seawards and ultimately became a thalassocracy (Abulafia, 2014) and, more importantly, held strong commercial relations with North Africa. Assuming that close and constant language contact between speakers of different mother tongues is the condition for the origin of a Pidgin, Catalan is the only Romance source language of the Lingua Franca which comes into question. The possibility that Occitan merchants ventured to North Africa, created the Lingua Franca and spread it into the Mediterranean is definitely possible due to the control of commercial networks and large-scale commerce implemented by the Occitan kingdom in western and central Mediterranean Arabic speaking regions (Valérian, 2014: 84). However, the Occitan kingdom was then not known to be a thalassocracy and although commercial dominance was a reality, “we can certainly conceive of commercial networks that did not have the backing of naval power and were not thalassocracies within any meaningful sense of the term” (Abulafia, 2014: 141-1421). Occitan was never used officially at sea, on ships or in commerce as Occitan was not common outside the European inland. Furthermore, the possibility that Occitan was the source language for the Lingua Franca seems to be very unlikely. There is too little evidence that Occitan traders ventured out into the Mediterranean in great numbers, consequently bringing with them the Occitan language. It is definitely not the case that it could have rivalled Catalan and most definitely not to such a degree that could have created a fixed Pidgin, yet alone facilitate its constant development. The Occitan kingdom also became part of the Catalan kingdom and vanished at some later point in history to become French due to the increasing French influence in the 19 th century (Abulafia, 2014: 147). Thus it is very unlikely that Occitan is the source language of the Lingua Franca (as it did not have contact to such a necessary degree for creating

33 a Pidgin with Arabic). The other possibility just seems too complex and simply too unlikely to have happened in this kind of way, i.e. that North African merchants ventured to Marseilles or Barcelona, created the Lingua Franca, and from there the North African merchants adopted the Lingua Franca, even brought and used the Lingua Franca back home, where it spread locally and at the same time became adopted by other travelling European merchants too and spread across the Mediterranean. Despite the common belief of Arab merchants preferring not to venture to European ports due to their religious belief (Skinner, 2009 quoted in Abulafia, 2014: 143), they did of course travel within the Mediterranean to trade, although this might have been a later development as Cifoletti’s article about Venice in the 18 th century suggests (Cifoletti, 2000). However, even in great numbers it would be unlikely that the Lingua Franca was created in this kind of scenario, due to two fascinating aspects about Pidgin formation: the superstratum language does not adapt its own Pidgin and, in a trade situation, the superstratum language is the language of the visitors, not the local language of that location (cf. the discussion of Austrian- Rumanian Pidgin below). Visitors that do not speak the language of the area use their own language to communicate and astonishingly, in a communication (language) situation, the local people (of that area) - merchants in this specific case - quickly adapt to it as they learn the meaning of the new words by trial and error. They then reassemble these new words according to their own grammar, and thus, after a very short time when communication is established, it has become clear which is the super- and which is the substratum language. The most fascinating aspect about this is that this seems to be a very common approach of people with the intention to trade, probably even of people trying to set up communication as well as the common pattern of such an approach. Therefore, the Lingua Franca would have been a completely different language with Arabic as the superstratum language and Catalan, Occitan or Italian as a substratum language and thus would not have looked like the Lingua Franca that is documented. Even in the case of relexification, the grammatical traits, which are visible even today, would have been completely different. Furthermore, the area in which the Lingua Franca would have been predominantly spoken would have been rather on the Northern Mediterranean shore: i.e. the Southern European shore and due to many different European languages many more varieties of the Lingua Franca would have been the result if the Lingua Franca had spread across the Mediterranean which is rather unlikely. Arabic speakers spreading the (Romance-superstratum) Lingua Franca in the Northern

34 Mediterranean definitely occurred (to some degree), but spreading a Lingua Franca with Arabic as a superstratum language seems rather unlikely, because Arab merchants traded much with European traders in their own region, North Africa. Thus an Arabic-speaking superstratum Pidgin would be less useful than a Romance superstratum Pidgin, (although it could be Europe if the Lingua Franca were an Arabic superstratum Pidgin). The theory that Arab merchants created and further developed the Lingua Franca but it was spread mainly by European speakers, correlates much more with the known historical processes and events and hence the possibility of Southern Europe as the place of origin for the Lingua Franca very likely just did not occur. The possibility that Italian was the source language for the Lingua Franca would definitely be given as, geo-politically, Italy and consequently Italian did have contact to North Africa and, consequently Arabic. However, this happened to a limited degree at rather small ports in the Mediterranean basin because Italy’s main destination was the Aegean Sea and the Black Sea (Kahane & Tietze, 1958: 9, 11-12). Thus Italian as a source language can be excluded (both linguistically, as there are linguistic differences, and) geo-politically, as contact in North Africa was not established to a sufficient degree, although early on in Mediterranean history Italy, i.e. Pisa and Venice did have colonies or rather settlements in North Africa (Chaunu, 1979: 262) and therefore contact in some form in North Africa must have existed. However, this presumably was the kind of, or the intensity of contact that does not lead to the creation of a Pidgin but the learning of a set of vocabulary necessary for a successful (trans)action, arguably close to an ad- hoc Pidgin, but nothing similar to close or intensive contact between speakers of different mother tongues which could have led to the creation of either a fixed Pidgin or the Lingua Franca. Furthermore, Italian, just like most Romance languages before the 12 th - 14 th century, only started developing around the 13 th - 14 th century and was heavily bound to and influenced by Latin before this time (www.orbilat.com/Languages/Italian/italian.html) and could therefore not have shaped the Lingua Franca in the form that is recorded, as there are too many differences in morphology. Furthermore, a relexification, from Old Italian to Old Catalan or Spanish, which completely altered the form of the Lingua Franca without leaving visible traces is quite unlikely as the grammatical traits of the Lingua Franca are also not close enough to Latin but also not to Old Italian. Curiously, Italian only developed its independent character in the 13 th to 14 th century, and before this time it looked similar to both Old Catalan/Occitan and the Lingua Franca, although only concerning verb forms which, like most other Romance languages, had conjugation and infinitive endings with /-ar/ or /-r/ in general. Old Italian, as it was called, shares,

35 in fact, more features with Old Occitan and Old Catalan than would be expected. Although Italy too would fulfil the contextual requirements (of Pidgin formation) there are also more differences to prevent the conclusion that Old Italian was the source language of the Lingua Franca (orbilat.com/Languages/index.html). Its main base of operation was the Aegean Sea and the Black Sea, due to Catalan being present in the western Mediterranean. Consequently, Catalan is indeed very likely the Romance source language of the Lingua Franca

Possibilities, perhaps opportunities, for a Pidgin to arise existed aplenty and any of these situations could have been the origin for the Lingua Franca, and so a different method of approach is needed. Looking at the politico-economic changes in the Mediterranean over the last millennium, the following picture of Lingua Franca origin and development could be given: The possibilities that are present at this point for the source of the development of the Lingua Franca would be either Catalan-Aragonese in the western Mediterranean, with influence reaching even into the central Mediterranean, or Italian in the eastern parts of the central Mediterranean. Both of them seeming quite plausible as both involve a Romance language and Arabic, and both of them were far spread in the Mediterranean, i.e. had the same opportunities for expansion, which is most interesting given the enormous ´interconnectedness of the Mediterranean trade network´ (Kea, 2014: 435). The Lingua Franca could have originated in this time in any of these places, including the Turkish North African provinces. Although the Ottoman Empire was dominant, Turkish speakers were only a minority consisting of administrative representatives and thus North Africa was a predominantly Arabic-speaking environment (Kahane & Tietze, 1958; 20). Furthermore, since Arabic was quite present in this part of the Mediterranean, it would rather have been an Arabic Pidgin developing than a Turkish one. Furthermore, a Turkish Pidgin would linguistically be too far away from the Lingua Franca. However, at this point in time the Lingua Franca could have been influenced by Turkish in the sense of taking over some concepts and loanwords i.e. “yoldach” turkish soldier “nelle Reggenze magrebine era il termine tecnico per indicare I soldati del contingente turco stanziato là” (Cifoletti, 2004: 59).

36 2.7 Clarification of the concept

As the Lingua Franca is known to be a Romance-Arabic Pidgin, consequently it was created by Romance and Arabic speakers that came into contact with each other in, presumably, an Arabic speaking area due to the tendency of Pidgins to adapt the language of the visitor as the superstratum language and not the local language of the area it originated in (which, in most cases, is the language of a minority). Evidence that Arabic speakers and not Christians used the Lingua Franca can be found in the source texts (Schuchardt, 1909 and Cifoletti, 2004 and Foltys, 1984 and Collier, 1977 and Selbach, 2008 and Lang, 1992), where captives only heard of the Lingua Franca once their captivity started and they had to learn it and talk to the local population in Lingua Franca, cf. e.g. Haedo (Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Foltys, 1984: 13-14 and Schuchardt, 1979: 35-37), Rehbinder (Rehbinder, 1798 quoted in Schuchardt, 1979: 39-40 and Cifoletti, 2004: 224-229 and Foltys, 1984: 25-26) and Renaudot (Renaudot, 1830 quoted in Cifoletti, 2004: 248-255 and Foltys, 1984: 28-29). Many Europeans only came into contact with the Lingua Franca through their imprisonment and enslavement and it is mentioned that the local population already spoke the Lingua Franca. It is an important fact that people from e.g. Algiers already knew the Lingua Franca but not the Europeans that were captured, which can be seen as an indicator for the origin of the Lingua Franca. The enormous difference is that the local Arabic population used the Lingua Franca but not the European population, which clearly shows that not only was the Lingua Franca an Arabic Pidgin (with a Romance vocabulary) but it also originated, developed and spread from Arabic-speaking North Africa. It seems most likely that the Lingua Franca established itself in North Africa and judging from the general tendency of Pidgin languages to remain in both the original social as well as the geographic area, only extending through the people involved, it seems only logical to assume that the Lingua Franca origin is closely tied to North Africa. The Lingua Franca, however, in regard to Pidgins and Creoles, is somewhat of an exception as it is one of very few Pidgins that have gained such a vast spread in an area of this size, i.e. the Mediterranean, and have even extended their social field. The Lingua Franca definitely extended also due to the considerable spread of piracy in the 16 th century (Selbach, 2008: 33), making it widely known among the North African population. The Lingua Franca also became relexified by Spanish and Italian in locations that had contact with other European nations i.e. Tunis, Algiers and Tripoli; e.g. Tunis, and presumably also Tripoli, with Italian and Algiers with Spanish (Cifoletti, 2004: 32 and Schuchardt, 1909), 37 causing the origin of the Spanish and Italian variety of the Lingua Franca, which (I claim) was rather a relexified continuation of the previous (Catalan) Lingua Franca but with increased popularity.

This circumstance only adds further confusion, as looking at the sources stating who actually spoke the Lingua Franca and who supposedly used it as a mother tongue differs immensely. People that supposedly spoke this language as a mother tongue were said to be all Europeans, and hence the Lingua Franca from the perspective of Arabic speakers meant the language of the Europeans and/or Christians. In other words the term refers to Christians, including people that were captured in the Mediterranean Sea and brought into the so-called bagnos (prisons) in Tripoli, Algiers, Tunis, etc. during a period in which piracy truly exploded in North Africa and hundreds of thousands were made slaves, forced to work and held for ransom (Selbach, 2008: 34). Curiously, the European sources never report the Christians using the Lingua Franca to communicate among each other, only to communicate with the slavers, although primarily the slavers themselves as well as the local population used this language in order to command or insult the Christian slaves, i.e. to communicate with Christians (Haedo, 1612 quoted in Cifoletti, 2004: 197-198; Renaudot, 1830 quoted in Cifoletti, 2004: 248-255). Christians were called francos in North Africa, as were, curiously, Muslim serfs in Iberia, but also “extra-peninsular Latin settlers (“Francos”)” (Catlos, 2014: 369): the language of the Christians was also called “Franco” (Haedo, 1612 quoted in Foltys, 1984: 13; Aranda, 1656 quoted in Cifoletti, 2004: 207), which would imply the meaning of Lingua Franca being the “language of Christians” or rather the language which the Christians were supposed to speak. The sources, however, allow, or actually suggest, a different conclusion. i.e. that it was Arabic- and Turkish-speaking North African slavers that truly used the Lingua Franca, claiming to speak the language of the Christians (i.e. all Europeans) and that Christians then forcedly learned this Pidgin, leading to the further misconception of it being a Christian language. The confusing aspect about the whole situation is that the designation “Lingua Franca” implies that it supposedly was the mother tongue of all the Christians, although it actually was the language that the slavers used to communicate with the Christians (slaves, merchants, sailors etc.), falsely believing they themselves were speaking the language of the Europeans, or more precisely the language of the Franks “lisan al-afrandz”, literally meaning the foreigners

38 (Schuchardt, 1979: 32). Interestingly, the term “Ifranj” was applied to “tribal military elites originating on the periphery of the Islamic world” (Catlos, 2014: 367). It seems unlikely that many Turkish or Arabic speakers, with the exception of diplomats and wealthy merchants, i.e. people that had a more sophisticated form of contact with Europeans, would have known that there were several different European languages and that what they spoke themselves was a Pidgin language. The Lingua Franca, therefore, was the language Arabic speakers used to communicate with Christians and merchants with the wrong assumption that they would speak the Christians´ mother tongue itself. If this complex situation is looked at carefully, a certain pattern can be detected, i.e. the Lingua Franca, despite its designation as the ´language of the Europeans´ among Arabic speakers, actually is the language used by Arabic speakers who believed that they were speaking the mother tongue of the Europeans but communicated in this language to trade with all Europeans (Haedo, 1612 quoted in Cifoletti, 2004: 197-198 and Foltys, 1984: 13-14 and Schuchardt, 1979: 36-37; Renaudot, 1830 quoted in Cifoletti, 2004: 248-255 and Foltys, 1984: 28-29). Furthermore, analysing the locations the Lingua Franca was mentioned to be spoken in, it is the North African shore, but there are a few exceptions of people knowing the Lingua Franca outside the Southern Mediterranean, e.g. Cornwall (Coates, 1971 quoted in Foltys, 1984: 23 and Cifoletti, 2004: 220- 221), but no communication networks based on the Lingua Franca are mentioned, and the Lingua Franca also appears not to be spoken on the European shore. However, the exception is Italy, but attested only since the 18 th century and there the Lingua Franca was known but not spoken by the whole population (Cifoletti, 2000: 14). The reason the Lingua Franca was not spoken in Barcelona, Marseilles or in Italian cities is very likely because the Lingua Franca was created in North Africa(n ports). If the Lingua Franca really stemmed from the term franco as in “Frankish”, it would coincide with the true franc or “Franconian”, i.e. the language itself and the name “Lingua Franca” would mean ´the language of the francs´, which would imply that Franks themselves must have spoken this language, although, contrarily, it was Arabic and Turkish speakers that spoke the Lingua Franca. The term itself, according to Schuchardt (1909) (as has already been pointed out above), stems from lisan al farandz which means ´die Sprache der Fremden´ (the language of the foreigners) and assumingly was first used in the 8 th century by Arabic speakers. The cultural group so represented really were the Franks (Schuchardt, 1909: 446-447), although this was only one cultural group that came to represent a whole continent, a ´pars pro toto´, similar to the

39 designation Arabs or Turks which can represent any culture within the vast cultural variety of Turkish or Arabian culture. Thus the term Lingua Franca was originally an Arabic term, translated into, or more precisely Europeanised, into the phonetically similar Romance equivalent franc or franco , possibly a Catalan term, to which the European term lingu a or lengua was added. Subsequently, the Italians took over franc(o) which became franca and hence the name Lingua Franca, ´la lingua franca´ in Italian, became known. It could also have been a Spanish term, however, lingua seems rather Italian, lengua would probably be more Spanish. Originally, the term afarandz (the foreigners) must have designated the inhabitants of the Iberian Peninsula, those that the Arabic speaking people first came into contact with. This could be an indication of the language implied in the designation “Lingua Franca” i.e. the language of the inhabitants of the Iberian Peninsula, which would include Catalan. Catalan, although said to be fully developed before the 11 th century, was not documented as a fully developed language until the 12 th century (orbilat.com/Catalan/Catalan.html) and until then presumably looked a lot like Latin, as did Italian until the 14 th century (www.orbilat.com/Languages/Italian/italian.html). Thus the origin of the Lingua Franca can be excluded at this point in time (and only the term remained). Later in time, the term was translated as franc(o) and applied to all Christians, i.e. Europeans, although originally only Iberians were designated by it probably due to the fact that more and more Europeans became present in the Mediterranean (commerce). Ever since the origin of the Lingua Franca, especially around the 14 th -16th century when the Lingua Franca became closely tied to the slave trade, the term franco was applied to the supposedly Christian language, the Lingua Franca. Theoretically, it seems plausible that the same term which originally applied to all Christians later acquired the meaning of the language (supposedly) spoken by Christians, i.e. a Romance Pidgin. However, the term franca , and in further consequence the name “Lingua Franca”, first appears in the documents and sources only in the 16 th century, 1516, by Reftelius (1516) “ lingva franca ” (Reftelius, 1737 quoted in Cifoletti, 2004: 216) and, later, in Rehbinder (1798) “ Lingua Franca ” (Rehbinder, 1798 quoted in Cifoletti: 2004: 225) and especially in documents by Italian authors, i.e. Lorenzo Bernardo (Lorenzo Bernardo, 1591 quoted in Foltys, 1984: 13. Other documents name the Lingua Franca le franc (Dan, 1637 quoted in Cifoletti, 2004: 204) and Haedo calls it “franca, o hablar franco” (Haedo, 1612 quoted in Cifoletti, 2004: 197 and Foltys, 1984: 13), but he also refers to the Lingua Franca as “lengua franca” (Haedo, 1612 quoted in Cifoletti, 2004: 201). However, the majority of the sources call the Lingua Franca “Langue Franque” (d´Arvieux,

40 1735 quoted in Cifoletti, 2004: 209; Dapper, 1686 quoted in Cifoletti, 2004: 210) or “Langue franco” (Aranda, 1656 quoted in Cifoletti, 2004: 207). The conclusion that can be drawn from the sources stating the designation of the Lingua Franca is that two separate varieties, a Spanish and an Italian variety should have existed by the 16 th century as (hablar) franco is mentioned in Haedo (1612), franca and lingua franca appear in the Italian, but also English and German, texts, most likely due to becoming Italianized in the 16 th century, with the Lingua Franca definitely known by the 18 th century, e.g. Venice (Cifoletti, 2000: 14). There is also the French designation langue franque which seems to have become popular, presumably due to the rising importance of French since the 17 th century. It appears that the whole term “lisan al-farandz” (Schuchardt, 1979: 32) was translated simply as franc , became franco , then expended to langue or lengua franca/o as well as lingva franca until it reached the more universally accepted designation Lingua Franca . The Pidgin language becoming known as “Lingua Franca” led to the widely spread belief that the Lingua Franca was actually an Italian Pidgin.

2.8 Historically changing perspectives

As we can see through this diffusion of terms, what is effectively known about the Lingua Franca is truly the most important aspect about studying this language. To fully understand the Lingua Franca, its origin, its spread, its development and ultimately its demise depends on what knowledge we have, or to be more precise in this case, how much is known about the Lingua Franca. To be able to achieve such a task, everything that is known, everything that has been passed down about the Lingua Franca in texts and reports including everything loosely connected to the Lingua Franca, even those seemingly unimportant bits and pieces found in history that had no direct connection to the Lingua Franca but still affected its development, have to be considered, otherwise no complete picture about the origin, spread and development of the Lingua Franca can be given. Before this task can be completed a significant problem that has to be faced is not the question of how many bits and pieces have to be used but the problem of looking at the past using ideas and beliefs of the present instead of looking at historical facts and considering the ideas and beliefs of that time. Michael Herzfeld (2014) has described this as the “diachronic paradox” (2014: 122) and states that it is a “focus on the past, or rather on ideas of what the past might have been as

41 well as representations of what it almost certainly was not” (Herzfeld, 2014: 122). Although he talks about tourism giving a somewhat distorted picture about the past, the concept of imagining history which, in the imagined form, did not happen and is thus giving a misrepresentation, can still be applied here. Especially in the case of the Lingua Franca misconceptions about its purpose and use could prove a fatal mistake, as looking at history from a today’s perspective including the beliefs and opinions of the present time, which may not have existed at an earlier time. Time, in different forms, may give a distorted view of the perception of the Lingua Franca in earlier centuries and about the Lingua Franca itself. Furthermore, this is true both ways, as what might have been intended in the past might simply be overlooked from today’s perspective, as Pidgin languages may have a much better image in present times which, however, does not mean that every writing about them in earlier centuries was full of contempt. Unfortunately, the Lingua Franca was quite often falsely viewed with these doubtful depictions and from a more recent perspective these quasi-racist beliefs and consequently wrong descriptions could be interpreted as correct observations. The consequences of falsely interpreting history would be that the Lingua Franca were looked at as a lingua franca would be today, as with the case of English, and consequently fail to show the required traits of a modern lingua franca, creating an obscure view of the Lingua Franca of the past and making an already hidden past forgotten or maybe even non-existent. There are many beliefs of today claiming the Lingua Franca to be based on different languages (primarily Italian and Spanish due to the 16 th century relexifications), which does influence where in history answers are sought after, resulting not only in a search in possibly the wrong place and a different time but which might also influence the result and create an origin different from the actual origin of the Lingua Franca. The problem about the Lingua Franca origin is that any clue or assumption that is followed up will lead to an abundance of places that would serve ideally as the Lingua Franca place of origin and even more times in the last millennia that would be suitable. Concerning the time and place of origin, even the assumption of the Mediterranean proves to be difficult as the Mediterranean was connected via one considerable trading network and serves as a suitable place itself. What is even more questionable is that the time can be specified even less, as the possibility of origin has existed since the beginning of trade itself. What can actually be done about this problem is the question that consequently arises, as mere speculation will lead nowhere, only further away from the actual origin. The only strategy that can lead to the origin of the Lingua Franca being illuminated is a compilation of everything that

42 is known about the Lingua Franca through texts and reports, an analysis of the language itself, i.e. its linguistic traits, as well as the gathering of information about the background of the times and places the Lingua Franca is reported to have existed; i.e. to interpret all the material and research. Hence one can (hopefully) come to a conclusion about the origin, spread and development of the Lingua Franca and all the details about it that lead to a plausible explanation of its creation, spread, development and demise. Considering how little is known about a language that appears to have been spoken in the whole Mediterranean, it seems to be a good idea to gather all available source material including the sources less relative or only indirectly concerning the Lingua Franca. This might create what Wansbrough has called the “abundance of documentation” (Wansbrough, 1996: 6), making it more difficult to uncover the origin of the Lingua Franca. However, it is a process that is most necessary for the (true) origin of the Lingua Franca to be revealed and for a plausible explanation of the origin, spread and development of the Lingua Franca to be presented. It is crucial to discover the (true) origin of the Lingua Franca via research, comparing the Lingua Franca to historical circumstances and linguistic features of certain languages and thus eliminating all languages and times that do not fit the linguistic and historical descriptions of the Lingua Franca. Before speculation about any details about the Lingua Franca can even begin, both the sub- as well as the superstratum language that created the Lingua Franca have to be known, and only then can the Lingua Franca be further analysed and a comparison to the known facts can lead the way to reaching a plausible explanation about its date of origin, its spread as well as its development. Basically, the strategy is to take all the sources and texts about the Lingua Franca, analyse which language the Lingua Franca (most likely) stems from and/or correlates with, analyse the history of this language and when this language came into contact with the other Lingua Franca source language. This should lead to a suitable time and place in history for this event, the creation of the Lingua Franca, to have happened. Furthermore, the (process of) development of the Lingua Franca, a process that lasted for centuries until the Lingua Franca’s final demise, can be recreated via historical, geo-political, and even anthropological means as well as, of course, through linguistics and Pidgin field studies. Anything that provides a further clue needs to be included and the interplay of all these different perspectives or approaches should at least give an indication to the true origin, spread and development of the Lingua Franca.

43 Chapter 3: Linguistic analysis

Notable linguists such as Schuchardt (1909), Cifoletti (2004) and Foltys (1984) make an attempt, via (indispensible) linguistic analysis, to interpret these facts in order to reach a conclusion about the Lingua Franca source language. Although little is truly provable and much remains unclear, it was the research of these individual linguists which achieved great progress and insight within Lingua Franca research. The realization that much still has to be discovered should raise an awareness of the high probability of misinterpreting facts and ´imagining history´. After all, there is little proof as to how and why the Lingua Franca was created; thus, the hypothesis developed in this study may simply remain, if not mere speculation, then one other attempt to give a plausible explanation of the Lingua Franca origin and development. With this established, let us look at the known facts and what can further be discovered with this knowledge. The Lingua Franca appears to be a Romance-like language, as its vocabulary is foremost of Romance origin, and indeed, the Lingua Franca has, over many centuries, been viewed as a primarily Italian language. However, it has been condemned as inferior due to its more simplistic-appearing grammar, although in fact the Lingua Franca is not a Romance language. It is not an unrightfully so-called “broken” or “bastardized” variety of a Romance language, be it Spanish, French, Portuguese or any other Romance language; it is not simplified Italian, and it is most definitely not an abomination as many people, both travellers and linguists of the earlier centuries (but occasionally even in the 20 th century), have claimed it to be. This fact might not be relevant for the present time when much is known about Pidgin languages and their uniqueness. However, lingual racism was a great issue during the times the Lingua Franca existed when it was consequently depicted as inferior and either ignored or criticised in a despising way. Therefore, it is relevant to address this issue, to analyse the effects on the Lingua Franca and to consider this when reading texts written about the Lingua Franca. Consequently not every report might be accurate, either shifted in favour of what the writer wanted the Lingua Franca to be and look like or giving a misguided picture about the Lingua Franca itself either due to the writer’s inexperience in linguistics and a (somewhat) limited understanding of this Pidgin or notoriously racist beliefs about this language. Of course, this seems unnecessary from today’s perspective but ´imagining history´ or misinterpreting historical motives indeed works both ways, in this case the admiration for Pidgins and Creoles which, unfortunately, had not existed in the times of the Lingua Franca. This is a fact

44 that has to be considered when analysing what was written about the Lingua Franca centuries ago and it consequently must be judged as to how reliable a source really is. In fact, most sources are so-called travel reports, written by adventurers, merchants, priests and sometimes passengers, i.e. people that were not linguists in the modern sense and could (or would) not give a precise description of the language. Schuchardt (1909), especially Cifoletti (2004) and Foltys (1984) have written in detail about this problem and have analysed most of the known reports about the Lingua Franca, and all of them state that a few sources do not give a very accurate description, and in some cases the descriptions given by people can even be disregarded as nonsense (e.g. some theatre plays) or not depictions of the Lingua Franca itself. The explanation for the inaccurate description of the Lingua Franca might be because many people most likely saw the Lingua Franca as an impure form of their own language and produced a corrected form of the Lingua Franca with their own language system. An example would be ´bono´ and its considerable spelling variation (Selbach, 2008: 46), which would allow the assumption that the Lingua Franca was any of these languages i.e. Italian, French, Spanish, Portuguese, Turkish and even Greek. There are many examples of the Lingua Franca, but especially linguists such as Schuchardt (1909), Cifoletti (2004) and Foltys (1984) have done groundbreaking work in this field, analysing quite a number of sources and thus validating or falsifying the authenticity of both the professionalism of the author´s depiction itself and the Lingua Franca presented in these texts, i.e. questioning if they concern the real Lingua Franca or are rather an interpretation of the author, a learned type of stage Italian or utter nonsense as in the case of some theatre plays. Fortunately, the other texts can be considered as valid Lingua Franca source material, as they can be legitimated as authentic depictions of the Lingua Franca, i.e. Encina (Harvey, Jones and Whinnom, 1967 quoted in Foltys, 1984: 10), Paolo Giovo (Ferrero, 1959 quoted in Foltys, 1984: 10), Haedo (Haedo, 1612 quoted in Foltys, 1984: 13), Aranda (Aranda, 1656 quoted in Foltys, 1984: 16) and several others. Fortunately, eminent linguists such as Schuchardt (1909), Foltys (1984) and Cifoletti (2004) are experts in the field of the Lingua Franca studies: Hugo Schuchardt, or as Cifoletti has called him “il grande Schuchardt” (Cifoletti, 2004: 10), was the first linguist to analyse the Lingua Franca and consequently the first to discover that the Lingua Franca was, in fact, Arabic grammar with a Romance vocabulary. Today, it is a known fact that Pidgins are created through language contact of two different languages (no mutual language connection is required). For the time of Hugo Schuchardt, however, this was a groundbreaking and also taboo-breaking discovery as many

45 scholars, among them linguists, rejected the idea of ´mixed´ languages itself. Considering the diverse opinions that were current about the Lingua Franca (and languages in general) at that time, Hugo Schuchardt has of course achieved a great deal in linguistic analysis and language studies and has inspired further studies in the field of Pidgins and Creoles. Schuchardt too gives the most plausible explanation as to how and why the Lingua Franca was created, where its name stems from and provides a most convincing linguistic analysis including an explanation for the choice of the infinitive in the Lingua Franca: “It is the European who imposes the stamp of general currency on the infinitive” (Schuchardt, 1979: 28). He also analyses the source languages of the Lingua Franca and reveals the Arabic grammatical element and the Romance-based vocabulary (Schuchardt, 1979: 27). Schuchardt also states that the language that has supposedly given the Lingua Franca its vocabulary is assumed to be Italian (Schuchardt, 1979: 33), although he does not seem to be truly convinced about this being the source language and thus he describes the Lingua Franca as “the communicative language formed of a Romance lexicon” (Schuchardt, 1979: 26). Cifoletti (2004), almost a century later, has continued the study initiated by Schuchardt and analysed the Lingua Franca in depth, especially the Arabic element. Among many valuable insights, he discovered the Berber element noticeable in the vocabulary of the Lingua Franca, although the influence of the Berber language, as well as Turkish, is in itself limited (Cifoletti, 2004: 33) and the Lingua Franca appears to have had a predominantly Romance vocabulary. Cifoletti (2004) does mention that both Spanish and Italian influenced the Lingua Franca but never explicitly names a Lingua Franca source language and thus leaves the impression that the source language could have been another Romance language as well. Cifoletti (2004) further mentions that the Lingua Franca even created (two) new words on its own i.e. cazeria (Cifoletti, 2004: 73), forar (Cifoletti, 2004: 74, 199) and further considers the possibility of griboilla being a ´Neuformation´ or own creation as well, although he himself is doubtful about this term (Cifoletti, 2004: 74). Hugo Schuchardt (1979) also talks about gribouilla but calls it a “blend” and states that it is “French but has the meaning of Ital.[Italian] garbuglio ” but could have derived from Old French grabouil or (19 th century) French gribouilette (Schuchardt, 1979: 41). Schuchardt (1979) does not list any new own creations and Guido Cifoletti (2004) only lists two. Although new and hybrid creations emerging from language use is very typical for a Pidgin, so- called ´Neuformationen´ or words independently evolving out of the Pidgin itself without the influence of the sub- or superstratum language appears to be a rather rare case.

46 Concerning the Romance source language, if various source texts are considered, the suggestion is offered that the Lingua Franca differs in its core elements. This can be explained from a geographical perspective, depending on which part of the Mediterranean is looked at, either Spanish or Italian (and a mixture of one of them with Turkish or Greek influence) seems to be the source language. This, in fact, has been stated by Schuchardt (1979), noticing the differentiating Italian and Spanish influence, he states that the “Lingua Franca with Spanish and Italian coloration blends them together with one another in different gradations so that perhaps only the western and eastern extremities are monochromatic”, calling Algiers “the geographical mid- point” of the Spanish and Italian influence (Schuchardt, 1979: 35). This results in the Lingua Franca being “Italianized” or “Hispanicized” (Schuchardt, 1979: 41), and later in the “Frenchification” of the Lingua Franca (Schuchardt, 1979: 43). Cifoletti (2004) has in detail confirmed that the Lingua Franca varied geographically and has also analysed the Italianization and Hispanicization of the Lingua Franca (Cifoletti, 2004). Already a (minor) feature of Schuchardt (1979), and with it of Lingua Franca studies, is the inclusion of a historic-political analysis which is offered on several points in his Lingua Franca analysis, giving both a historical background as well as a political explanation for some of the linguistic features (or choices) of the Lingua Franca (Schuchardt, 1979). This is indeed a most fascinating approach, which is further explored in this study. Cifoletti (2004) has collected a good number of source texts, among them very rare works and references to the Lingua Franca which show the existing differences between the Lingua Franca varieties that are referenced in many sources, but also the geographical aspect which becomes clear when considering date and location the Lingua Franca is being referenced in. Furthermore, one aspect noticed by Foltys (1984) is that some sources are actually referencing some sort of Italian but not the Lingua Franca itself, which is due to the observers either quoting Italian spoken by a non-Italian speaker or the observers lacking an Italian language competence themselves. Foltys (1984) has also had quite an impact on further linguists and Carles Castellanos (2010) has made a study based on the research of Foltys (1984) in which similarities between Catalan and the Lingua Franca are suggested. Following the linguistic studies of today, new considerations about the Lingua Franca Romance source language have appeared, as there are a few linguists such as Castellanos (2010) that state that the Lingua Franca is, in fact, a Catalan language. This might just be the missing piece of the puzzle (Castellanos, 2010). Nevertheless, the opinions vary, which shows just how similar

47 Romance languages are as all of them present a plausible source language for the Lingua Franca.

3.1 The similarity of Romance source language

At first sight, the Lingua Franca seems to be a Romance language, close to Italian, with Spanish influences, similar to Latin with a number of Arabic, Berber and Turkish loanwords. Although the Lingua Franca is similar to all Romance languages, including Latin, which becomes clear if the forms of several Romance languages, including Latin, are compared to the Lingua Franca itself 1, it seemingly is none of them in total. It could almost be assumed that this Pidgin, which appears to be not fully similar to any Romance language, does not originate from one Romance language alone but is rather a cluster language, put together from every single Romance language, most likely at a point when Romance speakers were forced to create a common means of communication among each other. This, historically speaking, could have happened both during the time of the crusades and at the peak of slavery on the North African coast in the 16 th century when speakers of all different Romance languages were coming together as slaves in the prisons or bagnos of Algiers, Tunis and Tripoli (Selbach, 2008: 33-34). This theory, although possible, must be rejected because, although the Lingua Franca is reported to be a mixture of languages containing Italian, Spanish, French or Occitan, Latin, but also several Arabic dialects presumably further including Berber, Turkish and even Greek, this mentioning of several Romance languages is due to a feature of language similarity which is the grammatically, lexically and phonetically close relationship of Romance languages among each other, as they all stem from Latin or the Latin vernacular. A very important point here is that the Latin vernaculars were much more close to each other during the Middle Ages than they are today, which would further explain why several Romance languages are mentioned. The coincidence that the European languages all stem from Latin or the Latin vernacular and therefore, in basic grammatical structures and especially in pidginized form (in grammar), which basically consists of core grammatical traits, indeed look quite similar, serves as an explanation as to why the Lingua Franca is described as a mixture of languages. Furthermore, it is also one factor that explains why Italian has often been named as the language of origin for the Lingua Franca, including the fact that ´Italian itself was most popular and widely spread in the 16 th century´ (Chaunu, 1979: 15, 305).

1 As for instance shown in https://en.wikipedia.org/wiki/Mediterranean_Lingua_Franca 48 Another important aspect that serves as an explanation is that the Lingua Franca was influenced by different languages depending on its location (Tripoli, Algiers, Tunis) including Italian in Tunis and, presumably, Tripoli, and Spanish in Algiers, leading to an Italianization and Hispanicization (Cifoletti, 2004: 7-14). The circumstance that Italian has been named as the source of the Lingua Franca is due to the fact that the Lingua Franca was a Romance Pidgin and was Italianized, which made it very close to Italian. Furthermore, Italian became really present and popular in the Mediterranean during the 14 th to 16 th century, with the Italian Renaissance at the height of its influence. Catalan more or less declined by the 16 th century with its recession starting even a century earlier (orbilat.com/Languages/Catalan/Catalan.html). The Lingua Franca being a Romance based Pidgin and of Romance (i.e. Latin) descent and thus also close to Latin, prevents the unambiguous identification of its origin language, as there are far too many similarities between the Romance languages that one could categorically rule out all the other Romance languages and leave only the correct one. The belief that the Lingua Franca is most likely Italian is a very plausible theory, although (judging from the linguistic analysis) so is Old Catalan. At the time when the Lingua Franca was noticed, which was rather late in the 16 th century, no precise studies were undertaken of it, and only much later, when the Lingua Franca was actually considered a subject worthy of study and research, could many facts about the history of the Lingua Franca be revealed which had been either considered irrelevant, remained unnoticed or simply forgotten before as no observer then assumed that the Lingua Franca was more than a recent and a temporary development and dated back for such a long period of time (being dismissed as a curiosity). Basically, observers at that time, starting presumably in the 16 th century but attested since the 17 th century (Dan, 1637 quoted in Cifoletti, 2004: 204) and especially in the studies (i.e. sources) of the 17 th to 19 th century, always assumed the Lingua Franca to be Italian, or to be more precise, corrupted Italian i.e. “Italien corrompu” (D´Arvieux, 1735 quoted in Cifoletti, 2004: 209) or “Italiano corrumpo” (Pananti, 1817 quoted in Selbach, 2008: 41 and Foltys, 1984: 26). There are, of course, always further characterizations concerning which other (Romance) languages are involved as well, and the descriptions of the Lingua Franca in these sources vary enormously, designating its origin from Spanish, Portuguese to French to Latin and even Greek. Curiously, the Lingua Franca has only once been explicitly stated as Catalan by the various sources consulted in this study. In detail Clough called it “a mixture of Catalan, Limousin,

49 Sicilian and Arabic, with other roots, especially Turkish” (1876 quoted in Foltys, 1984: 36). However, Foltys (1984) adds that this source is not reliable (1984: 36). Only much later in the 21st century have some academics (seriously) considered the possibility of Catalan influence. The earliest student of the Lingua Franca that was very close to this assumption was La Condamine (1731 quoted in Foltys, 1984: 22), although La Condamine stated that the Lingua Franca was (partially) Provencal. Considering the considerable similarities between Occitan and Catalan (although they do differ from each other in certain aspects) allows the assumption or at least leaves open the possibility that he could have referred to Catalan, as it would seem that he may have chosen Occitan simply due to the remarkable similarity to Catalan itself. In fact, Catalan and Occitan were so similar that a differentiation is difficult, especially in a Pidgin, as many basic linguistic traits as well as vocabulary are shared and can easily have looked the same when pidginized (with Catalan and Occitan alike). However, Foltys (1984) points out that the choice of Provencal by La Condamine (1731) is rather due to his nationality (La Condamine, 1731 quoted in Foltys, 1984: 22) and thus most likely would have made no distinction between Provencal, i.e. Occitan, as both languages are said to be French, and Catalan. Castellanos (2010) states that due to the similarity between these languages, it is possible to consider the Lingua Franca as Occitan- Catalan (Castellanos, 2010). It is quite difficult, from a linguistic point of view, to unambiguously identify both Catalan and Occitan (in the Lingua Franca) as well as to determine the language of origin due to the similarities of the Latin vernaculars which make it most difficult, if not impossible, to verify which language is the correct one. However, most sources and linguists state Italian as the main Romance source language for the Lingua Franca which, with a few exceptions, seems to be the unified opinion about the Lingua Franca language of origin. Thus a whole chain of varying descriptions, depictions and opinions have made it very complex to see through this Gordian knot of information to pinpoint the origin of the Lingua Franca, having been described as Italian, Spanish, Occitan, Greek as well as the language of the Europeans, slaves and corsairs alike. It is none of them in its basic form, but a Romance-Arabic Pidgin stemming not from Latin itself but from one of its vernaculars. Although claiming that the Lingua Franca does not match Romance languages in full does not necessarily mean that the Lingua Franca does not have a Romance language as its superstratum. However, due to having much in common with other Romance languages, it is indeed difficult to determine the original Romance source language. What is unusual as well is that it is rare for a Pidgin to distance itself so far from its source language that it becomes difficult to distinguish.

50 Perhaps the second source language responsible for the creation of the Lingua Franca can assist in finding a solution. Experts such as Schuchardt (1909) and Cifoletti (2004) but also others such as Rossetti (2005) and Castellanos (2010) agree that the second source language must be Arabic, and therefore, it must have been Arabic that came into close contact with a Romance language, which would even go back to Latin. It would also be possible that the Lingua Franca was a Latin- Arabic Pidgin. However, the Lingua Franca is closer to its vernaculars than it is to Latin itself and, consequently, Latin is also not the source language. Even though a Romance language must have been the source language for the Lingua Franca, none of the Romance languages, such as Spanish or French, not even Italian, which is closest to the Lingua Franca, seem to match the Lingua Franca completely, and it almost seems that it is another Romance language that simply has been overlooked either because it had died out or become irrelevant on a political and/or economic scale.

3.2 Original Romance source language

To fully identify the source language of the Lingua Franca all Romance languages, including Latin, need to be compared with it, which should indicate which language shows the most striking similarities, i.e. which is closest to the Lingua Franca. Curiously enough, Latin also seems to be quite similar to the Lingua Franca, which, despite being the origin for all Romance languages, is somehow unexpected, as the origin of the Lingua Franca presumably dates long after the demise of Latin. What is truly fascinating about this relative closeness is that the historical possibility and plausibility is given, as contact between Latin and Arabic definitely existed at some earlier point in history. However, when taking a closer look it becomes clear that Latin cannot be the source language of the Lingua Franca as there are too many differences, linguistically speaking. With the form close to Latin but Latin not being the source language the only possibility would be the source language of the Lingua Franca being a direct descendent of Latin. Latin just shares a lot of linguistic features with the source language of the Lingua Franca, but the source language of the Lingua Franca is a Latin vernacular, and the question, of course, is which one. Expectedly, there are similarities between these Romance languages, e.g. French, Portuguese and Spanish words do match the Lingua Franca words but only sporadically. This situation can be explained because all known Romance languages are still similar to each other in some regards

51 and consequently share some forms that do not seem to have changed over the course of history. Probably these are old structures that have been preserved in dialectal forms and thus have stayed the same, which could also explain the similarities to the Lingua Franca. Naturally, there are similarities to other Romance languages, as they are closely related but also because many Romance languages such as Spanish, Italian and French have indelibly influenced the Lingua Franca and caused a shift in its vocabulary and even grammar. Arabic traits can be identified too, such as the /ar/ and /ir/ in verb endings, the final letters never ending in /–o/ or /–e/ but rather in an unstressed /–a/ (Cifoletti, 2000: 17). The Arabic influence is present in the Lingua Franca phonetics and morphology (Cifoletti, 2004: 21-41) and the repeatedly occurring reduplication e.g. ´bono bono´. Reduplication as such is a most common feature in Pidgins and Creoles as it allows putting stress on meaning and it is also a common feature in every language, in some languages more than in others. Interestingly, Arends (2003) states that “reduplication is not a feature of Arabic” (2003: 229) but instead quotes Underhill (1976) stating that reduplication is “´a major grammatical and stylistic device of Turkish´” (Underhill, 1976: 436 quoted in Arends, 2003: 229). Dickins, Harvey and Higgins (2002), on the other hand, state that “lexical repetition is the most commonly used form of rhetorical anaphora in Arabic” (2002: 131) and there is also “the repetition of the same word” (2002: 108) which would indicate that both Arabic and Turkish support reduplication forms which would explain why reduplication was so common in the Lingua Franca in both the eastern and western variety. Further similarities between Arabic and the Lingua Franca include grammatical traits ´found in verb systems as the system of tenses in Arabic can also be context depended. One verb can be the present tense but also the future and the past tense because Arabic combines tense and aspect in its verb system. English, for instance, has verb tenses that consistently relate to natural time´ (Dickins, Harvy and Higgins, 2002: 99). This is a matching link to the Lingua Franca tense system which basically consists of the infinitive (later with the additional qualifier bisognio to indicate future tense) which can, dependent on context, take on any tense necessary. Furthermore, “Arabic can shift word order freely” (Dickins, Harvey and Higgins, 2002: 115) and in the Arabic word order ´the verb can be right at the beginning´ (Dickins, Harvey and Higgins, 2002: 119) which is because “punctuation in Arabic is less systematic…[as] traditionally, Arabic had no punctuation whatsoever” (Dickins, Harvey and Higgins, 2002: 115). These are remarkable similarities to the Lingua Franca which seems to have adapted most of these features, presumably because Arabic speakers would have used their own grammar with Romance vocabulary and

52 once regular grammatical patterns were established many of the Arabic grammatical features had become an inherent feature of the Pidgin. Pidgin languages, such as the Lingua Franca, tend to remain close to their original source languages and do not differ significantly from them (although easily shifting their vocabulary in general). It is only in the development process to Creole languages that differences with regard to the source languages occur (as new grammatical traits tend to be created). Hence, with the Lingua Franca being an extended Pidgin it should (despite the shift of vocabulary, i.e. Hispanicization and Italianization) remain close to its source languages and resemble the original source language to a quite high degree. The reason why the Romance language which created the Lingua Franca has been overlooked as a source language is due to the fact that its original form had become replaced by Spanish, although before it dominated the Mediterranean for centuries. The language in question is Catalan which unfortunately entered a phase of regression in the 16 th to 19 th century, but has since been revived in the 19 th century becoming more close to Spanish in the process (orbilat.com/Languages/Catalan/Catalan.html) Because of that it can easily be overlooked when considering the source language of the Lingua Franca.

3.3 Catalan-Occitan similarities

Linguistic analysis indicates that the Lingua Franca was a Catalan Pidgin, to be more precise it is Old Catalan, a language that existed from the 8 th to 16 th century. However, before Catalan can be classified as the Lingua Franca source language, it must be mentioned that great similarities can be identified between (Old) Catalan and (Old) Occitan (and the Lingua Franca). Lindsay (2009) considers both Occitan and Catalan as originally one language originating around the 5 th century which only began to differentiate themselves in the 13 th century (Lindsay, 2009), and also ´Medieval Catalan can be considered as the literary dialect of Occitan i.e. Lemosi´ (orbilat.com/Languages/Catalan/Catalan.html). The most promising vernacular of Latin, therefore, is both Catalan and Occitan as these two languages show not only a very close resemblance to the Lingua Franca but are also very similar to one another. What crystallises more and more when analysing these languages is that although all Romance languages have a certain linguistic closeness to Latin, only Old Catalan and Old Occitan considerably match the traits of the Lingua Franca, while the other Romance languages show occasional correspondences. Catalan and Occitan show remarkable similarities with the Lingua Franca not only in vocabulary

53 but also in sentence structure and word order. In fact, it goes as far as the situation that a sentence of Old Catalan or Old Occitan can with ease be translated into Lingua Franca, and indeed, there is not much that needs to be changed in the sentence. Old Occitan is a language that existed at the same time as Old Catalan and also shared its fate: it declined too and was revived in the 19 th century (orbilat.com/Languages/Occitan/index.html). Both of these languages are spoken today, although not by the whole population (in their regions) and, what is more important, differ to the original languages, having been adapted more closely to Spanish and French respectively (orbilat.com/Languages/index.html). Linguistically, it is difficult to differentiate the languages because of several reasons, since both languages declined and because of this there can always be a minor detail that might have been missed or has been forgotten in time that could make a major difference in the analysis. Although research on these languages has been conducted and therefore there might only be a slim chance of overlooking significant details, one can not exclude this completely. Both languages come from an (almost) identical linguistic background, were geographically situated close to each other and at some point in time were spoken in parts of the same kingdom, under the Crown of Aragon (Chaytor, 1933), with “Catalan territories on the French sides of the Pyrenees (Roussillon and Cerdagne) and Montpellier” (Abulafia, 2014: 146). This brought the two languages close together linguistically. One aspect to mention further is that individual similarities to or even matchings with the Lingua Franca are found in some dialects of Occitan today such as Languedocien and Provencal, which is due to these dialects being more conservative in the sense of being closer to medieval Occitan (Lindsay, 2009). However, no dialect possesses the full spectrum of grammatical traits and similarities which would lead to the assumption that some older forms of Old Occitan might be preserved in the dialects of today’s Occitan. This is a point where, linguistically, two options are presented as the Lingua Franca origin – either Old Catalan or Old Occitan, as both of them are quite plausible. The problem, at least linguistically, is that the Lingua Franca is very similar to both Old Catalan and Old Occitan, and a language that is very similar to another is almost impossible to fully exclude as a possible source language in a Pidgin as only basic grammatical and vocabulary traits, which usually already share similarities in the (developed) languages themselves, are found in a Pidgin language. The Lingua Franca itself is a Pidgin language, meaning that its grammatical core consists of so-called simplified, i.e. basic, grammar which becomes visible through the limited grammatical traits (and also vocabulary) and thus makes a further distinction between two

54 possible source languages difficult. The distinction of potential source languages can further be obfuscated by the fact that the Romance languages functioned as a superstratum, i.e. lexifier language, while grammatical substratum influence might have favoured certain Arabic patterns. Both languages (technically) contribute to the aspect of grammar due to (Arabic) grammar being influenced by (Romance) grammatical traits, i.e. the substratum grammar also adapting superstratum grammatical traits which are part of the vocabulary. Transitive/intransitive verbs or verbs requiring a specific case being taken over and implemented into the other language thus lose their original grammatical traits and are replaced with the grammar of the target language, or basic grammar takes over the role of more complex grammatical traits e.g. first person singular for first, second and third person singular and plural. The Lingua Franca uses the singular form for both singular and plural in nouns but also verbs can refer to the second person plural while still using the first person singular. Prepositions serve as a context marker, but differentiating only between the first and second person singular ( mi, ti) and this is not ideal for a linguistic differentiation of very similar languages. Linguistically, it is quite difficult to state what grammar defines the Lingua Franca, i.e. (the same form including) infinitive, phonology and basic word order can be found in both Catalan as well as Occitan. However, even though Occitan shows many similarities and parallels to the Lingua Franca it seems to have less similarities than Old Catalan. Looking closer, while some differences can be detected, there is nothing absolutely unambigiuous to determine which language verily created the Lingua Franca. At best, it can only be mentioned that Old Catalan shows more similarities to the Lingua Franca than Old Occitan does, which is not a complete proof. Interestingly, Old Italian, at least with its verbs, also shows similarities to the Lingua Franca, but by far not as many as Old Catalan or Old Occitan do (orbilat.com/Languages/Italian/Italian.html). By further comparing other words and analysing the results, the astonishing fact can be discovered that, indeed, (Old) Catalan and (Old) Occitan words match the Lingua Franca. (Old) Occitan is a language that is so similar to (Old) Catalan that linguistically the source language could be classified as a Catalan-Occitan origin, as done by Castellanos (Castellanos, 2010), because these two languages are almost indistinguishable on a basilectal level. However, historically, only (Old) Catalan can be classified as the source language for the Lingua Franca as it was the Crown of Aragon whose sea-based empire, i.e. thalassocracy, and trade network reached across the Mediterranean basin (Abulafia, 2014) and had Catalan as its official language.

55 3.4 Pronunciation as a source language indicator

What is most fascinating, but also a noteworthy problem, is that the pronunciation of the Lingua Franca does not match either Catalan or Occitan, i.e. verb endings are pronounced with the consonant /r/ (which is the most typical feature of the Lingua Franca verbs). Arguably, the correct pronunciation of Lingua Franca verbs, if there was any in a (pronunciationally non-fixed) Pidgin, is not known and speakers seemed to pronounce the Lingua Franca words according to their own nationality. However, I would argue here for a specific aspect of Pidgin languages, i.e. that they do not vary much from their source languages, meaning that features such as pronunciation, word stress, sentence stress (and structure) etc. are simply adapted, as these features are a fixed part of the already established source languages. These features, if not all features, hardly vary or differ from the source languages in the first Pidgin stages, i.e. ad-hoc, fixed and extended, although it can be a completely different case with Creoles, and while words from the other source language are pronounced differently this is still according to the (other) source language pronunciation. When learning a new language, ideally the new grammar is learned as well, however, when there is no fixed grammar to learn one must resort to the grammar one already knows. In the case of Pidgins, usually the grammar of one (substratum) language and the vocabulary of the other (superstratum) language are combined into a new (Pidgin) language. The superstratum language is adapted with grammatical traits of the other source language, usually grammar that the other language speakers insist upon using, e.g. infinitives or ´duplication´ or grammatical traits that are bound to vocabulary such as tense auxiliaries, prepositions (word order) etc. Furthermore, in the creation of Pidgins there is no re-creation such as vowel shift, only a reordering or restructuring of already existing features (one source language (vocabulary) is taken over with grammatical features of the second source language coming in occasionally). In Pidgin languages, “new” words seemingly appear through endomorphic restructuring (suffixes, endings) which seems to create new forms of words, although it is the process of strictly keeping already existing grammatical forms and words from both source languages and putting them together in a new way, probably unknown to both source languages, which gives the impression of new words being created. Consequently, there is no shift in the pronunciation of words per se, as the pronunciation of both source languages remains unchanged. However, through the process of combining two languages into one (mixed) language, the pronunciation of the superstratum language is combined with the pronunciation of the substratum language, and thus an apparently new pronunciation is created, although, technically, it is an adaptation or 56 restructuring. Therefore, pronunciation in Pidgins stays close to its source languages, but alterations are noticeable due to the restructuring process in the combining of both source languages (or, in other words, due to transfer features from the substratum languages). A circumstance that further supports this claim is that although the Lingua Franca has always been a spoken, not a written, language many people have written down their encounters and experience with the Lingua Franca, i.e. consequently they have written down how they heard the Lingua Franca being spoken, i.e. the pronunciation of the Lingua Franca. Looking at the source texts remarkably supports this claim and a considerable amount of written evidence can be found: e.g. all of the travellers have written andar , bibir , cunciar not andare , bevere etc (Cifoletti, 2004 and Schuchardt, 1909 and Foltys, 1984). The only reports that do write this Italian pronunciation are Italians themselves, or it can be found in theatre plays, but not even there is a consistency visible as sometimes they write andare but other times they write andar . This is most likely less due to the traveller’s nationality but more aligned with the type of Lingua Franca variety they encountered (e.g. andar being the Spanish and andare the Italian variety). The explanation, I believe, lies in the fact that the Lingua Franca became Italianized at some point as Cifoletti (2004) has pointed out (and what led Schuchardt to consider Italian as a source language), and with it the Lingua Franca pronunciation shifted towards Italian. This is why there is variation between the sources and variation even within various source texts. A curious aspect is that only the most common Italian words are affected, i.e. basic verbs, which could be a clear indication for the Italianization of the Lingua Franca as Italians speaking the Lingua Franca would pronounce the most basic words according to their own pronunciation. This would lead to a barely noticeable shift in vocabulary so that possibly even two forms of pronunciation could have existed simultaneously in the Italian variety. Although a few reports stating the existence of a more Italianized Lingua Franca variety can be identified as depictions of Italian dialects and not the Lingua Franca itself, as occasionally stated by Foltys (1984), there are valid documents (quoted in Foltys, 1984 and Cifoletti, 2004) showing depictions of different infinitives of the Lingua Franca. Italian infinitives are noted instead of Catalan or Spanish infinitives andare – andar, which can be considered rather as accurate depictions of the pronunciation of an Italian Lingua Franca than a simple Italianized spelling or a mother-tongue based typing mistake in the sources of the later centuries. Even Guido Cifoletti (2004) has mentioned that as unbelievable as some forms may sound, at least in the 19 th century they seemed to be a reality (Cifoletti, 2004: 202), allowing the assumption of these Italianized

57 forms to be correct. In all source texts, even in some of the Italian sources, the Lingua Franca is reported to end verbs with /ar/ or /ir/, i.e. to be more precise the Lingua Franca verbs are pronounced with a consonant /r/ at the end, with the rare exceptions of the Italianized variety ending with a vowel i.e. /e/. Consequently, it can be concluded that this is the correct pronunciation of these verbs. Unfortunately, this (ending with a consonant) is not how Catalan and Occitan (but Spanish and French in some cases) end their verbs. The infinitive of Catalan and Occitan verbs ends with a vowel (Romance languages orbilat.com/Languages). However, looking at the history of Catalan and Occitan, one notices that both languages were revived in the 19 th century and have changed in the progress. Through a historical-linguistic analysis one fact becomes clear: Old Catalan and Old Occitan, i.e. the original Catalan and Occitan languages, all show similarities to the Lingua Franca, even the same pronunciation, as both Old Catalan and Old Occitan end verbs in /ar/, /er/ or /re/ and /ir/ (Monk, 2013: 215), i.e. with the consonant /r/. Today, verb endings close with a vowel, i.e. /a/ in Occitan or with a schwa in Catalan, although some dialects of Catalan and Occitan (of today) have kept this original form and end with /r/ such as the Occitan dialect Aupenc (/ar/, /ir/ and /re/), but also Languedoc, Basque and Provencal (Lindsay, 2013) 2 and the Catalan dialect of Valencian (/ar/ and /ir/). Thus the pronunciation of Old Catalan (Monk, 2013: 215) 3, Old Occitan (https://en.wikipedia.org/wiki/Occitan_conjugation) and the Lingua Franca, except the Italian variety, can be considered as the same. Further similarities are sentence structure, word formation, verb endings and prepositions which are adapted one-to-one, as are content words, predominantly (main) nouns and verbs.

3.5 Text comparison

Comparing both Old Catalan and Occitan sentences given in Dante (1304) and Joam Rosenbach (1502) and examples provided by Mateu (2009), Alfajarin (2000) and Fischer (2003) with the Lingua Franca, astonishing matchings can be discovered. This fact shows that through these remarkable similarities the Lingua Franca was very close to Old Catalan/Occitan and could have been either created or already established in this time period. To further specify the date of the Lingua Franca origin, older texts of Old Catalan and Old Occitan need to be analysed, i.e. 7 th ,

2 This is also mentioned in: https://en.wikipedia.org/wiki/Occitan_conjugation 3 This is also mentioned in: https://en.wikipedia.org/wiki/Catalan_verbs 58 10th , 12 th and 14 th century texts, and, curiously, 7 th to 10 th century texts show a closeness to Latin with the occasional Catalan or Occitan word. However, the beginnings of Catalan and Occitan bear no resemblance to the Lingua Franca (at all), consequently the time period before the 10 th century can fully be excluded as the date of Lingua Franca origin, leaving only the 11 th to 14 th century as the earliest plausible date. The 11 th century would seem to be a possibility, but Catalan only became established as a language (different from Latin) at the end of the 10 th century and only appears in written form in the 12 th century and Occitan had its peak in the (late) 12 th century (orbilat.com/Languages/Occitan/Occitan.html). Language contact itself must have occurred beforehand, most likely since the re-establishment of trade in the 11 th century (Gertwagen, 2014: 159) around 1050, which is the first contact between Catalan and Arabic speakers in North Africa since the banishment of Christians in the Southern Mediterranean and the consequent diminishing of trade since the 7 th century (Chaunu, 1979: 68), but this can not have created the Lingua Franca. Therefore, arguably, the earliest plausible date for the Lingua Franca to have reached the recognisable state in which it became known would be after the 11 th century, assumingly the late 13 th to 14 th century, once Catalan had become fully established. However, Catalan becoming a fully developed language in itself (distinguishable from Latin) does not necessarily mean that the origin of the Lingua Franca is in the same century. Furthermore, older texts show that both Catalan and Occitan still had (some) Latin features before the 11 th century. The earliest resemblance to the Lingua Franca can be found in the 14 th century texts which would, most curiously, suggest that the Lingua Franca of the 16 th and 19 th century resembles (in part) the Catalan language of the 14 th century, allowing a direct comparison between Old Catalan texts, the 16 th century Lingua Franca and the 19 th century Lingua Franca and showing remarkable similarities, especially concerning Old Catalan language features present in all time periods. This, however, also does not truly prove that the Lingua Franca did exist at this time, only that some early (i.e. Old) Catalan features have survived in the Lingua Franca. The assumption that the Lingua Franca could have existed around the 14 th century, can also not be supported through the source texts as the earliest example is from the 16 th century, given by Haedo (1612), which confirms that the Lingua Franca existed as a developed language, i.e. as an expanded Pidgin in the 16 th century (Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Foltys, 1984: 13-14 and Schuchardt, 1979: 35-37), but not much earlier than this date. Furthermore, as noted by Cifoletti (2004), the events portrayed in Haedo allow the assumption that Haedo’s

59 descriptions of Algiers predate the date of publication and thus the observations can be dated to the century before (Cifoletti, 2004: 197). The poem of the Zerbitana (1353), also written in the 14th century, further shows similarities to Catalan i.e. ´the expression mia is very typical for Old Catalan as it has been changed to meua and meva in later centuries´ (Monk: 2013, 215), however, there is no unambiguous proof that this is the Lingua Franca itself (Zerbitana, 1353 quoted in Collier, 1977: 295). Schuchardt (1979) has analysed this poem and states that the only Arabic expression is barra followed by an Italian word fuor “ barra fuor ” (Schuchardt, 1979: 33) which would allow this assumption of a Lingua Franca connection but is not enough to be considered as proof.

3.6 Indistinguishable basic grammatical forms

Starting from the 5 th to 12 th century there was presumably only one Catalan-Occitan language but during the course of development two varieties developed, Catalan and Occitan, which still showed remarkable similarities to each other even in the 15 th and 16 th century (Lindsay, 2009), but analysing this circumstance of change further differences become noticeable in pronunciation, conjugation and declination. These differences are visible, e.g. in the conjunctive form in some (French and Spanish) dialects of today, and show more similarities to the Lingua Franca than others, such as Basque, while the dialect of Marseilles (Occitan) or Maritime Provencal (Lindsay, 2009) do not. This is a curious fact, as the Lingua Franca was a Mediterranean language and most likely took over the form of a language that had contact to the Mediterranean. The North-western French shore, but also Basque, however, was in close contact to England through trade (Chaunu, 1979: 305), while Marseilles had commercial connections to the rest of the Mediterranean, and thus it seems it should be the other way around with Marseilles being the dialect closer to the Lingua Franca, thereby making Marseilles seemingly more appropriate as a source language. Presumably, this occurrence is due to some linguistic features surviving or being preserved in the dialectal form. The Marseilles dialect could have been much closer to the Lingua Franca originally: however, after the revival of Occitan, much French influence became noticeable, which seems to have affected some dialects more than others. Presumably the preservation of old forms is related to closeness to the standard language, which is also one reason why the dialects of today vary so much among themselves, as some have much more contact to French and other dialects than others do (Lindsay, 2009).

60 Catalan, for instance, shows more features that are close to the Lingua Franca. Unfortunately, one essential problem arises: while there are, of course, differences in Old Catalan and Old Occitan, these differences can be found in the more complex grammatical structures such as conjugation, i.e. conjunctive, future and passive, tense auxiliaries etc., but not within simple or basic grammatical structures such as basic word order (SVO), word stress, pronunciation and intonation, i.e. those features the Lingua Franca is prominently composed of (due to the lack of highly complex grammatical structures such as passive conditional subjunctive). These (basic) grammatical traits of Old Catalan and Old Occitan remain so similar that linguistically they can hardly be differentiated, which on its own would not be a problem as ultimately these are minor complications. However, the true problem about this circumstance is that the Lingua Franca, in all of its complexity, is a language built upon basic grammatical structures.

3.7 The complexity of simplicity

What makes Pidgins (and Creoles to a certain extent) complex is the way basic grammatical structures are restructured, grammatical endings (endomorphs) are applied in a fashion unknown to the (superstratum) source language and vocabulary itself becomes more generalized with a single vocabulary item covering to many more meanings and contexts than in the source language, i.e. semantic extension, but also making single words only apply to and be used in one single context with one meaning, i.e. semantic restriction. Schuchardt (1979) gives many examples, one of which is cunciar which has the general meaning of ´to make´ as in ´ cunciar pace ´, but can also acquire the meaning of ´to build´ as in cunciar una casa (literally ´to make a house´). It can also mean ´to do´ cosa cunciar (literally ´what are you making´, i.e. doing) thus changing its meaning in different contexts (Schuchardt, 1979: 29). Regardless of the complex result, Pidgins only use the most basic forms of language, e.g. the Lingua Franca does not have morphologically marked tenses, and only uses the infinitive to indicate present, past and future - a simple basic structure that, ironically through its simplicity, becomes most complex to decipher. Later, in the 19 th century, the Italian variety developed a past and future which was indicated through bisogno as commented on abover (Schuchardt, 1979: 29 and Collier, 1977: 286).

Only possessive pronouns of the pronominal system are used and only in the singular, mi and ti, which serve as a substitute not only for all other forms but also for all other pronouns (including

61 reflexive, demonstrative etc.), and which makes expressing an idea easy but deciphering meaning a most complex task. Here one has to deduce grammatical rules themselves as they are not in the sentence itself, which could be one reason why communication among different language speakers works by itself as Pidgins orient themselves to existing forms making complex forms identifiable even through basic grammar. Structured idiosyncrasies or differences occur through contact which is a natural phenomenon that occurs when fully fledged languages establish and maintain communication. The intensity of such an occurreance is, of course, much higher in Pidgin languages as they are (almost) dependent on such new creations.

At this point, it is necessary to point out that without being based on an already established grammar, this would become a highly complex task and hardly be possible even for speakers of the superstratum and substratum language, but especially for speakers of languages other than the source languages as there is no declination, no plural and conjugation only existing in one single form, the infinitive. Furthermore, there are a number of prepositions which take on more complex grammatical attributes as indicators for meaning, tense and especially location. For speakers of other languages than the source languages, it becomes an almost impossible task to decipher the ambiguity of grammar as well as the intended meaning. One further aspect to note at this point is that the assumption that Pidgins consist of the vocabulary of one source language and the grammar of the other is not entirely true as such, because there are basic grammatical traits carried over by the superstratum language through its vocabulary, i.e. tense forms, infinitive, imperative, prepositions, tense auxiliaries etc., although this influences the grammatical structure of a Pidgin to a lesser degree. What makes the Lingua Franca, and Pidgins in general, structurally challenging is the lack of complex grammatical structures and hardly any morphological or syntactic differentiation or indication as to what grammatical form a word has, or is intended to have (or means). However, there still is a fixed lexico-grammatical meaning, which uses rather indirect means as a substitution for grammar to convey its meaning such as context, vocabulary, prepositions and even word order. What is so fascinating about this characteristic is that it makes the assumption that a Pidgin is a simple language quite ironic given the complex grammatical structures that are being substituted.

62 3.8 (Limitation of) expressionability

With that said, there is a certain limitation compared to other (fully fledged) languages. This so- called structured simplicity (i.e. basic grammatical structures) limits the spectrum of meaning that can be expressed, especially what is linked to tenses (e.g. as in German ich wäre gegangen ), fixed expressions and abstract phenomena. However, there is far more that can be expressed with rather simple means either through fixed phrases or, blandly put, more sentences, plus there is another useful aspect that boosts the lexical semantic flexibility of a Pidgin and that is the openness of its vocabulary. Not only can, in theory, the vocabulary of two languages be drawn upon but also words from other languages, when needed, can enter a Pidgin without delay. This can be confirmed by the document of Haedo (1612) which reports Portuguese vocabulary entering the Lingua Franca within a short amount of time after a high number of Portuguese soldiers became captives (Haedo 1612 quoted in Foltys, 1984: 13 and Cifoletti, 2004: 198). Consequently, none of the long processes common in language development occur in Pidgin development, for instance word borrowing, a process that can take years as to enter a language a foreign word needs to establish itself first before it becomes widely accepted. Admittedly, borrowing can be an instantaneous process in non-Pidgin languages as well, although borrowings may take a longer time until they become institutionally accepted (i.e. are adopted into grammar or dictionaries). This, however, can almost immediately be achieved in a Pidgin language and, most importantly, concerning Pidgin languages, the meaning of an existing word can be extended, created in context even, and defined through a preposition or a phrase with a fixed meaning, i.e. a fixed expression, which shows not only how much can be expressed with an often called “simplified” grammar but also how creative Pidgin languages are as they solve complex problems via simple but effective means. Ultimately, there are limits as to what or how much can be expressed; however, it is considerably more than would be expected from such a grammatically basic language.

3.9 Conclusion

Linguistically, there is the problem that, while simple and basic grammar can be analysed and identified, a source language for a Pidgin can not be fully differentiated if two languages are too close or too similar to each other and consequently share the same basic grammatical structures and forms. As Catalan and Occitan have almost identical linguistic features (on this linguistic

63 level, i.e. the basilect) there is little to no differentiation via basic grammar and structure (including prepositions, word order (SVO), pronunciation). In any case, there is too little differentiation as to fully prove which of the two is the source language. Even if there are any differences, it seems unlikely that these are major differences or, ideally, unique traits of one language, i.e. anything that causes an identifiable difference in basic grammar or at the basilect level. These differences, if there are any, most likely are not enough, linguistically speaking, to prove or disprove one language as the only possible source language in question. Although there is a tendency in favour of Catalan, there is no considerable substantial evidence to support this claim. Linguistically both languages are equally plausible as the possible source language due to the basic structure of the Lingua Franca. (However, analysing the Catalan and Occitan language further, especially in later decades, considerable differences become noticeable) Despite this stalemate, much can be deduced from the similarity of the Lingua Franca to the Catalan and Occitan language and many other important facts about the Lingua Franca can be concluded from this linguistic analysis, e.g. the estimated time period when the Lingua Franca originated, at what time this (Pidgin) language became established and an estimated location of the first plausible contact can be determined. As one source language, Arabic, was spoken on the southern Mediterranean shore, and both Catalan and Occitan were geographically relatively close to each other, spoken on the north western shore of the Mediterranean basin, the western part of the Mediterranean covers a plausible point of contact of both source languages (which can further be specified via different means). The origin of the Lingua Franca must have been in a location in which both languages came together and interacted with each other. Technically, concerning Arabic, this would include the whole southern shore of the Mediterranean basin as ´the Arabic Empire stretched from Byzantine as far as Spain in 711 and Arabic was such a widespread language that it supplanted Greek as a lingua franca´ (Doumanis, 2014: 452) causing ´the spread of Arabic within the whole Arabic Empire which also became the language of administration and trade´ (Valérian, 2014: 83). However, distance in these times was an issue; therefore, contact can not have occurred far off the Catalan or Occitan shore, which restricts the point of contact to the western part of the North African shore that could be reached by both primarily Catalan and Occitan traders. Both the Catalan language and Catalan commerce were slowly established and spread into the whole Mediterranean, but only reached a high point in the 13 th century. The Catalan and Occitan language did not, at least in the 12 th century, spread on the whole southern shore but only in the

64 western part, which is North Africa. The beforehand addressed problem of being unable to identify the source language of the Lingua Franca linguistically, due to linguistic limitations of this analysis, can be solved by using different means (other than linguistics) e.g. historical, geo-political and anthropological approaches. The evidence that cannot be determined linguistically might be revealed through different means such as a historical study, an anthropological and a geo-political analysis. The history of both Catalan and Occitan should hold more clues about which language could have created the Lingua Franca. Furthermore, a study of Mediterranean commerce, which facilitated language contact, is essential too.

Chapter 4: Historical analysis

Concerning the historical background of Catalan and Occitan, both languages originated as one in the 5 th century and separated in the 12 th century (Lindsay, 2009) and were, by then, still very similar to Latin and nowhere close to the Lingua Franca. Through political developments both Catalan and Occitan developed and gained their own shape until they were fully developed independent languages by the 12 th century. Catalonia also developed into a kingdom that, after commerce became rediscovered or rather re-established, soon realized the potential of trade, even though its kings only saw trade as a funding opportunity for their wars (Abulafia, 2014: 145, 146- 147). What can be said about the kings of that time and the royal court is that most of the time they were on the battlefield or directed their political agenda, arranging marriages, formulating pacts, peace treaties and declarations of war (Chaytor, 1933). Even in official meetings with Arabic speakers those kings never interacted with Arabic speakers directly as they had translators and thus never came close to creating a Pidgin language, not that there were not many negotiations (peace treaties or declarations of war etc.) and with them also contact with Arabic speakers. However, linguistically speaking, there never existed an opportunity, i.e. language interaction for the origin of any Pidgin, not to mention the Lingua Franca. Neither Catalan, nor Occitan kings created the Lingua Franca as both were too far away from interaction with Arabic speakers or speaking another language, as they had translators. Consequently, the Lingua Franca did not originate through kings. Catalan-Aragonese kings, however, made great efforts to expand

65 their territories and sent many expeditions into the Mediterranean (Abulafia, 2014: 145). However, multilingual situations must have been present, although not including Arabic, through the planned marriages which allowed princes and princesses from all over the country and Europe to be wed to the Catalan and Occitan offspring. Politically, one major difference between Occitan and Catalan occurred as of the 11 th century, i.e. the seaward expansion which changed the Catalan land based kingdom into a thalassocracy (Abulafia, 2014). To begin with, both kingdoms had one major port, i.e. Barcelona and Marseilles; Marseilles is mentioned to be an important port, presumably because in those times travellers and traders usually sailed from port to port stopping for supplies until they finally reached their intended destination as there was no direct travel within the Mediterranean yet. Consequently, (European) merchants and travellers had to take every opportunity to stop at a port, (sometimes staying there for days as travel was very much bound to weather conditions, loading and unloading, crew-hiring, etc.). Barcelona, on the other hand, became well known for merchants travelling to the Muslim parts of the Mediterranean, and the Barcelonan port had a size that almost expanded along the whole coast of the Catalan kingdom, which could be an indication for its importance as well as the extent of its commercial activity. This tendency would indicate that it was rather the Catalan language that expanded into the Mediterranean and its importance only grew within the following 200-300 years, starting with the re-establishment of trade to the 14th century. The Catalan kingdom, or Crown of Aragon, had become a sea based empire and stretched from the Iberian Peninsula to Greece, while Occitan remained a land based kingdom and was also governed by the Catalan kingdom for some time (Chaytor, 1933). The political circumstances allowed both kingdoms and languages to remain close to each other, Occitan troubadours were always present at the royal courts and the Occitan language always had a huge literary influence on Catalan. However, the Catalan language did have an influence on Occitan as well due to the Catalan reign for some time; consequently, there still was not that much of a (basic linguistic) difference between these two languages noticeable even after centuries of development. However, the political agenda/facts suggest Catalan to be the language of more importance in the Mediterranean as Catalan became the official language of commerce from the 13 th to 15 th century 4 and in its colonies (Malta in 1127, Majorca in 1229, Valencia in

4 (http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710b0c0e1a0/en_GB/i ndex619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextchannel=77595c43da8962 10VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aR 66 1238, and Sicily in 1282, later also Sardinia, became Catalan colonies). The settlement was so successful that Catalan not only became the main language but can be heard in the streets of Alghero, Sardinia, even today; furthermore, there were the so-called dominions or possessions in North Africa, Tunisia and Tunis, and Crete (Abulafia, 2014: 145-151) and thus Catalan was carried out into the Mediterranean as the language used on ships and even the language used to settle merchant disputes, i.e. the language of maritime law (Backman, 2014: 181 and The Consulate of the Sea, 1494). The Consulate of the Sea (1494) was a compilation of older sea laws with further alterations from the Catalan kings translated into Catalan. Catalan therefore became the (official) language of the Mediterranean Sea and of commerce itself, contrary to Occitan 5. Consequently, traders and travellers used Catalan to conduct their business, thus making Catalan quite a plausible choice for the Romance source language of the Lingua Franca. Though it might not be possible to unambiguously identify the Romance source language of the Lingua Franca linguistically, through an analysis of the history of the Catalan and Occitan language, including the political assets and impact regarding language, a fairly plausible theory can be attested. Not only did Catalan have much more presence and prestige in the Mediterranean, it also had the necessary resources, i.e. numbers of people for large-scale commerce, (repeated) interaction and long term relations with Arabic speakers, and therefore, continuity (and repetition) in language contact could be established.

4.1 The beginning of commerce

Beginning with the re-establishment of trade in the 11 th century (Gertwagen, 2014: 159), trade was revived with Christians finally being granted access to Muslim lands (Abulafia, 2014: 143), i.e. North Africa, and trade slowly recovered from its previous almost complete absence (Valérian, 2014: 80). The number of Christian traders (from Barcelona, Marseilles and several Italian cities) that travelled to North Africa, the Maghreb and the Levant to trade for goods such as raw materials and luxury goods increased constantly (Pollard, 2014: 465). In this time many different people traded, and it could be assumed that the nation that was most effective and

CRD&newLang=en_GB) 5 (http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710b0c0e1a0/en_GB/i ndex619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextchannel=77595c43da8962 10VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aR CRD&newLang=en_GB) 67 successful in trading was also the one that gained more presence and power, which would lead to the conclusion that the Crown of Aragon, i.e. Catalan, was most present in the Mediterranean trade. Via gaining power, Catalan as a language also received more and more prestige while the Crown of Aragon gained economic and commercial importance which facilitated contact among the Catalan merchants and the Arabic speaking merchants. In this light, it seems highly probable for Catalan to be the moving force of the Lingua Franca to originate and develop into an extended Pidgin (even though ad-hoc Pidgins must have arisen every day) through the consistent Catalan presence and language use in the area in which the Crown of Aragon conducted business and commerce on a huge scale. Concluding, the area of commerce of the Crown of Aragon, its main trading routes and ports, should coincide with the place where the Lingua Franca was used and that with the highest probability the Lingua Franca originated in North Africa through Christian Catalan traders. Christian merchants were getting access and travelling in large numbers to the ports of Arabic (speaking) North Africa, which as a consequence led to the origin of the Lingua Franca and further contributed to its spread and development in this established commercial area. The area to which the merchants travelled to should thus coincide with the area the Lingua Franca was created and spoken in, as it can be assumed that the Lingua Franca developed a close connection to and remained in this area for a long time. Thus Catalan-Arabic language contact must have been the basis for the origin of the Lingua Franca and the question is at what point in time this started the creation of the Lingua Franca. The Lingua Franca must have been created at a time when commerce and contact between Arabic and Catalan speakers was frequent, and developed presumably when Catalan itself was the official trade language, which must have been at the time when Catalan merchants ventured to North Africa regularly. However, this does not necessarily mean that no other merchants ventured to the same parts of North Africa but that the Catalan language was dominant and, more importantly, North African merchants used Catalan to communicate with (other) Romance merchants. Assuming that this happened when the frequency of commerce was high, presumably by the end of the 13 th century as ´the period from 1050 to 1350 was marked by both political and economic expansion´ (Catlos, 2014: 368) allows the further assumption that the Lingua Franca spread rather fast across the Mediterranean. The spread of the Lingua Franca should also coincide with/be directly proportional to the spread of Catalan commerce, as through the dominion of commerce by the Crown of Aragon the Catalan language spread as well, so there should be a

68 direct connection between these factors. Commerce was re-established in the middle of the 11 th century and was constantly but slowly increasing throughout the 12 th century, but only established itself on a broader scale in the late 13 th century and reached its peak in the 14 th century, at least concerning the Crown of Aragon. Consequently, the origin of the Lingua Franca would rather be within this later period. Furthermore, the origin and the spread of the Lingua Franca should lie close together as well, as they are closely tied to the spread of Catalan commerce and the Catalan language. The Lingua Franca was be tied to commerce and its development; consequently, its spread can only have happened when commerce was established and Catalan was (omni)present in the Mediterranean, which only was the case later on in the 14 th century. The establishing of language contact must have occurred early on, in the 11 th century, once Catalan established contact to North Africa; however, the most plausible date of origin for an expanded Pidgin such as the Lingua Franca is around the time of the Crown of Aragon, and with it Catalan, became commercially dominant in the Mediterranean. This would indicate that the Lingua Franca originated and/or evolved into a fixed or expanded Pidgin in the 14 th century. One aspect in history which is important to note is that with the re-establishment of trade in 1050 there were not masses of people trading, as ´the eleventh century only marked the beginning of territorial conquests of the Latin Mediterranean´ (Abulafia, 2014: 82). Only some merchants were venturing out into the Mediterranean, and thus a rather slow process started through the opening up of the Muslim shores in which a low number of merchants began to trade with North African merchants. However, the African merchants themselves were not low in number; in fact, there were large (port) cities that were known for their commerce. It was rather the Catalan traders that were few in number, as commerce between Christians and Muslims as well as Christians on Muslim soil had been forbidden for centuries before the re-establishment of commerce. Thus, the North African cities were unknown to most Catalan traders (an exception being the Jewish traders (Astren, 2014 and Catlos, 2014)), and, therefore, large-scale commerce was hindered until this very same commerce proved so lucrative and profitable that more and more merchants were setting sails to North Africa, as Jewish merchants had done long before. Establishing commerce of this size so that huge numbers of Catalan merchants were conducting large-scale commerce took a long time, if not centuries. The 11 th century was only the beginning of both, the re-establishment of commerce and the development and spread of Catalan as a(n) (independent) language (that differentiated itself from

69 Latin). Therefore, it seems unlikely for the Lingua Franca to have originated in the 11 th century, when in the middle of this century commerce was still on too small a scale to provide the necessary conditions and the Catalan language was not fully established in the Mediterranean yet. Commerce itself also only slowly developed and became slightly more common almost a century later in the 12 th century. It then developed further and steadily increased through the 13 th century to reach its peak in the 14 th century. Although contact of the same kind (Catalan-Arabic) existed since the re-establishment of trade, the Lingua Franca most likely developed its final shape not earlier than the 14 th century, as texts from 14 th century Catalan (and Occitan) bear some resemblance to the Lingua Franca, which would suggest that the Lingua Franca, definitely the form known from later centuries, could be from that time period. The possibility of earlier origin and development exists in form of indications found only in historical but not linguistic facts. In any case, this cannot be proven as no sources from earlier periods exist, and there are some points speaking against this theory such as the fact that the Catalan and Occitan language had not been fully developed before the 12 th century.

4.2 Commerce as a catalyst of origin and development

The potential for a Pidgin to arise has existed ever since contact between Romance and Arabic speakers had been established; however, for a Pidgin to truly arise, a lot more than just an occasional encounter between two people is needed, i.e. repeated continuous language contact which can be found in large-scale commerce. It is large-scale commerce that brought with it the necessity for a means of communication and the combination of circumstances that large-scale commerce entailed led to the creation of the Lingua Franca. Of course, it was not commerce as such but the contact that established itself between different people through conducting business on large scale and long term relations that created the Lingua Franca. This, however, probably did not happen in the 11 th century or even within the first century since the re-establishment of trade (for several reasons). In the beginnings, though lucrative, not many merchants could afford or dared such large scale investments; thus, many merchants were financed by the Catalan kingdom i.e. the Crown of Aragon which was still in a state of war and expansion. Consequently, funds for large scale investments were rather rare and not many merchants were supported considering the risks such a venture entailed. Gertwagen (2014) mentiones a voyage from 1184 in which the conditions of, and happenings during sea travel can only be described as most dangerous

70 (Gertwagen, 2014: 163). Furthermore, “merchants who took to the sea tacitly accepted piracy as one of the risks of doing business” (Backman, 2014: 170). Therefore, the (southern) trading routes, although being established, had yet to become safe and accessible to a broader public and a large number of merchants, which took some time. Commerce, therefore, remained a “highly risky, yet lucrative, business” (Pollard, 2014: 467). This whole process created quite a lot of changes, e.g. the rise of a merchant caste that was present in the whole Mediterranean with lodgings in various ports, the so-called merchant quarters, with warehouses reserved only for the merchant’s wares (Chaunu, 1979: 262) and the Catalan language becoming known and spoken in the whole Mediterranean due to the Crown of Aragon having expanded onto the most important islands necessary to establish a thalassocracy and also commercial ties and routes within the whole Mediterranean. In the 14 th century the Crown of Aragon had reached its biggest expansion, facilities were established, trade routes and a trade routine had become fixed and also the knowledge to accomplish such ventures i.e. ´nautical technologies and techniques (as well as the banking sector) were available to a large public´ (Gertwagen, 2014: 163). Once commerce and Catalan thrived (in the 14 th century), the ideal conditions for the creation of a Pidgin were met and this could be the (most plausible) point of origin of the Lingua Franca. This would also be the most plausible and solid theory (from this perspective), as an origin earlier than the 14 th century cannot be deduced from a linguistic point of view. Both the historic and linguistic analysis strongly suggest Catalan as the Romance source language due to the importance of Catalan commerce, the Catalan language and the intensive language contact with Arabic speakers since the 14 th century.

4.3 Date of origin

The resemblance between both 14 th century Catalan as well as the extended Pidgin (varieties of the) Lingua Franca of later centuries suggests that the Lingua Franca could have existed in the 14th century as a fixed Pidgin. Continuing this line of thought would lead to the further conclusion that for the Lingua Franca to have reached this state development must have occurred earlier on. A Pidgin can develop very fast indeed. The amount of time it takes for such a language to be created, developed and used primarily in the context of trade only takes a few years to a decade. Thus, it is easily conceivable that the Lingua Franca originated, developed and spread within a

71 short amount of time. An ad-hoc Pidgin is established immediately if language contact is established - even through the most basic language contact (for a short amount of time) between two people with different mother tongues. The problem (here) is that an ad-hoc Pidgin originates when speakers of two different languages communicate (Holm, 1988: 4-7) and vanishes once two speakers part; however, if language contact is maintained, an ad-hoc Pidgin can remain in this state until it is either developed further into or replaced by a fixed Pidgin. Following this assumption, the 11 th century could be seen as the earliest possible starting point of the origin and subsequent development of a Catalan-Arabic ad-hoc Pidgin. It would be quite implausible that an ad-hoc Pidgin that originated when Catalan-Arabic contact was first established not only survived the subsequent centuries but also developed into the Lingua Franca itself. Although commerce most definitely established (Catalan-Arabic) language contact, it is questionable if this Catalan-Arabic language contact, which by definition of Pidgins and Creoles must have created an ad-hoc Pidgin (Holm, 1988: 4-7), can be seen as the earliest form of the Lingua Franca itself i.e. its origin. However, it could also be only a natural event caused by language contact which did not correlate with the Lingua Franca origin (at all). There is no proof that a Catalan-Arabic Pidgin from the early centuries developed further into a fixed Pidgin with the final result of being the same expanded Pidgin in the 14 th century as the Lingua Franca. The reason why an ad-hoc Pidgin from the earlier centuries can not be the same as the Lingua Franca is that, when commerce became re-established, language contact might not have been intensive enough to create an expanded or even a fixed Pidgin, and the Catalan language was not even fully developed yet. What can be said, however, is that most definitely an ad-hoc Pidgin arose and although it may not have been the Lingua Franca itself, this could be an important historical event which could have led to the development of the Lingua Franca or give a clue to its origin. To give an exact date for the origin of the Lingua Franca is probably not possible, as an exact date is usually connected to an (historic) event that led to its creation. As it is unclear which event truly led to this development, there are several possibilities. A most speculative date would be 1050, the re-establishment of commerce between Christians and Muslims which established contact and most likely an ad-hoc Pidgin, but only speculations can be made if this ad-hoc Pidgin developed further (or at all) or could have lasted long enough to become the Lingua Franca itself. Another possible event which facilitated trade and in this respect influenced or catalysed language contact was the conquest and one hundred year rule of Malta (1127) and Sicily (1282)

72 which were the earliest events that made the development and spread of both Catalan and the Lingua Franca within the Mediterranean possible. Via Sicily Algiers, Tunis and Tripoli, all of them very important commercial cities, lay within reach, which allowed a constant flow of goods that further ensured language contact, increasing knowledge of either Catalan or the Lingua Franca and its development. Therefore, the 14 th century, in which all important factors for the origin, development and spread of a Pidgin accumulate, presents a plausible theory. Unfortunately, this remains speculative, probably even just a vague guess, as the first (serious) reports and sources about the Lingua Franca can only be accredited to the 16 th century. Reports and sources before this date are far too imprecise, without examples and sometimes do not even match the geographical area the Lingua Franca occurred in (in later sources and centuries), cf. e.g. “reports of pilgrims blessed with the gift of tongues” (Wansbrough, 1996, 148) or “the grace of languages” (Vita Sancti Paterni quoted in Hancock, 1977: 287). Looking at the first reliable source from Haedo (1612 quoted in Cifoletti, 2004: 197-202), followed by La Condamine (1731 quoted in Cifoletti, 2004: 215-216), etc. to the last source during the active days of the Lingua Franca by Hugo Schuchardt (1909), creates the impression that the Lingua Franca was already established by the 16 th century. Judging from the sources, the Lingua Franca also must have reached Italy in the 16 th century before it became popular and part of everyday life in Tunis (and Tripoli) in the 19 th century. The conclusion drawn from the reports and sources alone would state that the Lingua Franca probably originated before the 16 th century and was the Spanish variety with the Italian variety following a century later. The possibility that Catalan (or, curiously enough, Arabic) could be the source language never occurs even remotely; it almost seems the Lingua Franca was noticed after Catalan had disappeared from the Mediterranean landscape and thus was never considered as a possible Lingua Franca source language. The curiosity about this circumstance is that this eliminates, or at least diminishes, the chance of discovery of the original source language considerably and also shifts the origin of the Lingua Franca by almost a century.

Much can be deduced from the documents, but contrary to this assumption is the fact that by the 16th century, the Lingua Franca is reported (by several sources) to be spoken within the whole Mediterranean in every port and every city, “on dit qu´on la parle dans tout le Levant et dans tous les ports de la Méditerranée” (La Condamine, 1731 quoted in Cifoletti, 2004: 216). This leads to the assumption that the Lingua Franca was (omni)present in the Mediterranean but restricted to its related groups, i.e. merchants and sailors who have acquired the Lingua Franca (first), but it

73 only spread within the cities among the local population once it had reached a certain point of popularity. Then, and only then, was it known by all people. Most likely Algiers was one of the first cities, as its use there was documented first (Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Foltys, 1984: 13-14 and Schuchardt, 1979: 35-37), although this could have been the case in all commercial centres. The possibility that the commercial centres were also centres of the Lingua Franca by the 16 th century would seem possible considering that the same commercial relations were being conducted by the Crown of Aragon in earlier decades, i.e. the 14 th and 15 th century. However, as the Lingua Franca rather settled because of slavery than commerce this was probably not the case. Unfortunately, due to missing reports concerning other cities of the time an accurate statement cannot be made. As no concrete evidence can be drawn from the sources alone, the most relatable theory that corresponds with the known documents of later decades is the spread of the Lingua Franca corresponding directly with the Catalan expansion i.e. the Crown of Aragon. This was an eastward (west-to-east) expansion (with temporal differences in its spread, i.e. reaching Algiers first and all other major commercial cities (Tunis and Tripoli) later on with Greece being the last in the 14 th century). If the spread of the Lingua Franca occurred parallel to the Catalan expansion, both empire and language, then it can be dated to the either the early or middle 14 th century depending on whether the Lingua Franca originated in the beginning of large-scale commerce or was at its peak during the Catalan Empire. Personally, I would argue that the Lingua Franca could have originated once commerce had reached a certain importance, in the mid-14 th century. Furthermore, the Lingua Franca must have also been spoken in the whole area in which Catalan was the official trade language, and language contact between Catalan and Arabic speakers was quite frequent. Consequently, commercial cities with ports and ships are the most plausible locations. Unfortunately, there really is no final evidence to fully prove this theory, but by analysing historical events further proof for such a claim might be found.

74 4.4 Historical and linguistic analysis

4.4.1 Earliest Catalan-Arabic language contact

Considering that the earliest Catalan-Arabic language contact may not allow the inference of an exact date of origin, it might still reveal further clues about the structure of the Lingua Franca and may lead to an explanation about some of its features and uniqueness; i.e. why Catalan or a Romance language in general became the superstratum language, which could also reveal (the motivation behind) the further history of its later development. Starting with Christian merchants gaining access to North Africa, later the Maghreb and Levant, and establishing continuous trade and contact with Arab merchants, these (Catalan) merchants used their own language and the Arabian merchants adapted to this situation and used the words of the Christian merchants (including some words of their own, but not in large numbers) thus creating the basis for an ad-hoc Pidgin. Ever since Catalan merchants were granted access to Arabian ports (Abulafia, 2014: 143 and Chaunu, 1979: 57), contact and, furthermore, communication was established which created Catalan language competence, although rather in a Second Language Acquisition scenario than a pidginized form. An important point here is that the definition of Pidgins (Holm, 1988: 4-7) suggests that after language contact was established and upheld an ad-hoc Pidgin must have been created. Perhaps even several ad-hoc Pidgins arose in different locations. Presumably the north-western part of Africa was the starting point, which was then expanded eastward with more ports becoming (inter)linked with this trading network. Presumably no ad-hoc Pidgin either lasted into the 13th /14th century or connected to the Lingua Franca, which might have originated once commerce had reached a high point centuries later. Language contact provided the basis for the Lingua Franca origin via increasing Catalan language competence among the Arabic speakers and also via the spreading the Catalan language itself with the outcome of a Pidgin that could be used primarily for trade. This ad-hoc Pidgin was a Catalan-Arabic Pidgin as, ever since the standardisation of Arabic as the standard written language in all of the Arabian Empire (Valérian, 2014: 83), Arabic, or a dialect of Arabic, was the language spoken in North Africa, and Catalan became the language spoken by the merchants that travelled to these North African ports due to the interest of the Crown of Aragon in trade. Italian and Occitan merchants were involved too, but with the success of the Crown of Aragon, both Occitan and Italian can be excluded from the North African coast

75 in the later development of trade, i.e. the 13 th -15th century where Catalan was the main language spoken in Mediterranean. At a later point in time, a similar process but in much larger quantity led to the creation of the Lingua Franca which, too, was a Catalan-Arabic Pidgin suitable for trade and useful for both Romance and Arabic speakers because Arabic speakers were using the Lingua Franca and Romance speakers could use their own language in order to achieve mutual understanding. Thus, technically, the Lingua Franca originated and expanded in North Africa because North African merchants used the Lingua Franca to communicate and trade with Catalan, Occitan, Italian, Spanish, Portuguese and Jewish merchants, who on the other hand used the Lingua Franca to be able to trade with other (Arabic speaking) merchants in different locations allowing the Lingua Franca to spread to other cities until the whole Mediterranean knew (about) it. The Lingua Franca then was used and adapted by merchants themselves, presumably in the 14 th (to 15 th ) century, although not by Catalan speaking merchants because the main idea behind the Lingua Franca, assuming that it is Catalan based, was to communicate primarily with Catalan (or Occitan) speakers, because it was a useful tool which made communication much easier and faster (they knew a few words in advance), with such great success and to such a degree that the Lingua Franca was spoken even after Catalan ceased to be spoken in commerce. Curiously, instead of declining during the 15 th century with the demise of the Crown of Aragon, the Lingua Franca continued to be spoken, presumably due to Catalan, as a language, reaching “a period of maximum splendor” in the 15 th century and, in the 16 th century “started to constitute the foundations of a future new diglossia…between Catalan and Spanish” 6. This might be the reason that the Lingua Franca neither declined nor became replaced by a Spanish Pidgin but, instead, became adapted to the new (lexical) situation and was overtaken, resulting in a shift of vocabulary towards Spanish and Italian as the Lingua Franca continued to be used as a means of communication. This was primarily as a language of commerce including piracy but there also was an increasing (Lingua Franca) speaker competence in the North African population (due to slavery) and even an extended use among Arab and Turkish diplomats. The Spanish relexification was conducted (and completed) in the 16 th century but started earlier

6 (http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710b0c0e1a0/en_GB/i ndex619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextchannel=77595c43da8962 10VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aR CRD&newLang=en_GB) 76 after Aragon and Catalan declined more and more in the (western) Mediterranean, because a marriage event led to the annexation of the Crown of Aragon with Spain (Chaytor, 1933). Supposedly, during the 15 th to 16 th century the central Mediterranean Lingua Franca was also relexified with Italian, and the eastern Mediterranean became more influenced through Arabic, Turkish and to some extent with Greek. Later during the 19 th century (1830) the Lingua Franca became relexified with French (Algiers), which can be interpreted as a certain dependence on the political influence the Lingua Franca, but basically any language, was exposed to. It is noticeable that the influence of (locally) dominant languages in the process of Pidgin creation and even more so during Pidgin development is prominent. As Guido Cifoletti (2004) states, the Lingua Franca has two (attested) varieties (Cifoletti, 2004: 32), which presumably developed during the 16th century, and these varieties were influenced not only by the local native languages but also by the (then and there, i.e. in time and place) dominant European languages. These insights are directly due to Guido Cifoletti (2004), who discovered that two different varieties of the Lingua Franca existed which developed during the 16 th century. This serves as an explanation for the existence of both Italianized and Hispanicized vocabulary in the source texts, e.g. Dio and Dios , parlar and hablar , testa and cabeza , and bono and bueno (Cifoletti, 2004: 202). Further expanding on his insights, it could be stated that the Lingua Franca of Algiers was influenced by Spanish and the Lingua Franca of Tunis was influenced by Italian (Cifoletti, 2004: 32), presumably including the Lingua Franca variety of Tripoli too. Any other Lingua Franca variety further east could have been influenced by Turkish and Arabic 7, even to some extent by Greek 8, and there also was a (short-lived) variety developing in the 19 th century which was influenced by French.

4.4.2 Linguistic interpretation of Catalan-Arabic historic contact

(In linguistic terms) The development of the earliest Catalan-Arabic language contact can be interpreted in terms of early Pidgin creation, i.e. usually a language other than the local language enters an area and is used or adapted by the local speakers to communicate with these other language speakers, while the local speakers only use the other language to communicate with the other speakers, not with local speakers, thus creating a context of when to use it, i.e. establishing

7 As is indicated by Desfontaines (1731, quoted in Cifoletti, 2004: 31, 216) 8 As is noted by La Condamine (1838, quoted in Cifoletti, 2004: 31, 220) 77 a certain limitation or boundary as to when the other language is used. Linguistically, the local people adapt to the new language as if learning new vocabulary which is (integrated into or) rearranged and restructured according to their own grammar. In this way, a new language is formed which exists only to form communication between locals and ´visitors´. As an ad-hoc Pidgin rather consists of single words and gestures, it is easy to acquire and, curiously, those that know such a Pidgin use it without hesitation, seeing that it works with speakers of the other language in general. Those that do not know such a Pidgin, either local or different language speakers, are, therefore, easily drawn in to participate in this form of communication and thus also learn the Pidgin quickly (but only in the context of its original social field of origin, i.e. creation environment). (Ad-hoc) Pidgins seem to follow the rule of taking vocabulary of the language entering a local language area which could lead to the assumption that when people come to another location they use their own language to communicate. However, this is rather due to them either not speaking any other language, meaning that people only use their own language because they have no other language at their disposal, or do not speak the local language, in which case they will use other languages that are available to them. A similar case occurs when locals speak the language of the guests as they will use the language that immediately establishes communication and therefore will indeed speak the language of the guests. Thus a Pidgin language will only arise in a situation in which mutual communication can not be established through shared knowledge of a language but when two different language speakers establish communication without having a common understanding of the other language. The assumption that can be made, however, is that rather local speakers adapt to the new language for communication purposes than the visitors, which could be a rule of Pidgin creation or common in a trade situation. One further observation is that acquiring a Pidgin, especially an ad-hoc Pidgin usually does not affect the local native language or a major part of the local population, as their mother tongue is used for communication among the other locals, not the Pidgin or words from the Pidgin. Even if the local population speaks unintelligible dialects, as is reported by Haedo (1612 quoted in Cifoletti, 2004: 197 and Foltys, 1984: 13) and Renaudot (1830 quoted in Cifoletti, 2004: 249), it does not lead to the adaption of a Pidgin by the local population itself but rather a full language which is spoken by a majority of the local population, i.e. morisca which is reported in Haedo (1612, quoted in Foltys, 1984: 13 and Cifoletti, 2004: 197) and also mentioned in Renaudot (1798, quoted in Cifoletti, 2004: 249). However, constant language contact between the local

78 population and speakers of a different language, i.e. Christians/Europeans, can lead to the origin of a Pidgin, which can become widely used as is shown in Haedo (1612 quoted in Foltys, 1984: 13) and Rehbinder (1798 quoted in Cifoletti, 2004: 228). Concluding, if both speakers have no knowledge about the other language and language contact continues, an ad-hoc Pidgin will be established which seems to follow the rule of taking the vocabulary of the non-local language, i.e. the ´visitors´, but the grammar of the locals, which is accomplished through the locals rearranging vocabulary according to their own grammatical system. Therefore, a vocabulary basis (i.e. Catalan) is almost immediately agreed upon, (and a consensus) quickly established and grammatically restructured, although grammar is seemingly avoided as only the simplest, most common and basic grammatical forms such as prepositions, present tense or infinitive are typically used. Only during the later development of Pidgin languages will grammatical traits be established or expanded through the existing grammar of the local language speaker. At the very beginning of language contact, however, only vocabulary (i.e. single words) is adapted, and if this process continues an ad-hoc Pidgin arises. The amount of vocabulary itself that is necessary to communicate and/or trade on a basilectal level is much less than could be assumed, as usually the general designation (or name of a product) and numerals are sufficient enough to express the desire to buy or sell a product. Furthermore, simple gestures such as pointing or shrugging are quite expressive and identifiable in their meaning. With only a few words in the repertoire that are remotely related to what one wants to express, a consensus or at least a rough understanding can easily and quickly be reached. What can be observed is that if contact remains constant, new words are integrated during the process of development and grammatical words are adapted too. At the beginning this is more a process of vocabulary expansion than a true Pidgin development, as it takes time to establish even an ad-hoc Pidgin. Once a certain intensity of contact is reached the process of grammatical acquisition starts and thus, after a short amount of time, an ad-hoc Pidgin is created. The result of language contact is an ad-hoc Pidgin which allows limited communication in a single context, in this specific case, trade. In commerce and situations of individual trading numerals, terms for basic products, verbs that express basic actions and prepositions are important as (all of) these are precise markers that express (and clarify) concepts such as ´what, how much and how many, where, from, who (etc)´. Furthermore, understanding what someone wants to express in a “closed context” i.e. trading, e.g. a merchant pointing out a product and saying a number, can then even be more easily achieved, especially if both participants are merchants with the desire to trade, and

79 as the participants know in advance what the interaction will be about. There are clear boundaries of what communication in such a context can or will be all about. Although, at the same time, this also shows the limitations of what can be expressed as a more sophisticated discussion (about the trade itself) which in such a situation can most likely not be conducted. In the specific case of the Catalan-Arabic language contact, this very specific feature of Pidgins may explain why Catalan was the primary vocabulary source. The question why Catalan became the lexifier language may not be answered via a linguistic analysis, but a socio-linguistic approach might offer an explanation. Many theories and explanations exist. One general conception, for instance, is that it was easier for an Arabic speaker to learn a European language than for a European to learn Arabic, which definitely applies to learning the Arabic language itself or any full language. This problem does not occur in Pidgins due to their simplicity and much less in ad-hoc Pidgins which only consist of the same few words that are being repeatedly used in different contexts. It could be stated that to learn an ad-hoc Pidgin would be similar to learning new vocabulary which should not be too difficult even if it is about a completely different language. Hugo Schuchardt also dismisses the idea that learning one (full) language is more difficult than learning another (Schuchardt, 1909: 447). Considering how much is known about Pidgin languages today, an alternative socio-linguistic explanation can be offered at this point. In my younger years I was able to witness the rise of a fixed Pidgin (myself), including its expanding vocabulary and its ad-hoc state. As Pidgins have a tendency to be very similar in their formation, this Pidgin can be compared to the early Lingua Franca development and thus I should be able to offer a plausible socio-linguistic explanation. This fixed Pidgin was an Austrian-Rumanian Pidgin that was created on something similar to a port, i.e. a (lively) gas-station where buses full of people would arrive constantly and people would all speak primarily Rumanian, but also Hungarian, Polish and other Slavic languages. Communication was really limited to buying supplies such as food, coffee, and fuel and the occasional attempt at small talk (“you like …”) usually ended unsuccessfully due to a lack of vocabulary as well as a lack of time. [Unfortunately, I do not speak Rumanian and as this Pidgin does not exist anymore, as almost 15 years have passed, with this Pidgin vanishing even more quickly than it originated, it is very difficult to recreate it completely. However, the most important aspects were still available to me due to my previous field research.]

80 Chapter 5: Socio-linguistic Excursus [An Austrian-Rumanian Pidgin]

This Austrian-Rumanian Pidgin was created through constant Austrian-Rumanian contact and language interaction, i.e. the repetition of it, but only existed as long as interaction between Austrian and Rumanian speakers existed (in this particular space), which was over a period of about 5 years. After that time the agreement that had led to this origin collapsed and the whole Pidgin vanished within three months, which is also the time it took for language contact to completely disappear. Curiously, nothing of this Pidgin is left and, therefore, one can not even go back to the place of origin and find remains of this language, as the conditions there have changed, completely erasing any trace of the former Pidgin. This Pidgin truly existed only at the very gas station it originated, which was also the place which stabilized it and kept it alive. The whole process started in 2001 with more and more Rumanian speakers visiting the gas station. Thus the acquisition of basic Rumanian vocabulary (coffee, milk, numerals) took place and an ad-hoc Pidgin formed in 2002. Its peak (of development) happened around 2004-5, reached its (final) fixed Pidgin state around 2005 and ended abruptly in 2007. Although during the last two years (2005-7) the intensity of contact remained on the same level, the Pidgin did not change or develop further. The Austrian-Rumanian Pidgin used Rumanian as its superstratum language, i.e. Rumanian vocabulary and Rumanian speakers saw it as a form of Rumanian, more an early Second Language Acquisition stage than a Pidgin. This is probably why they never adapted or further developed it and Austrian speakers never used it outside this specific location (as communication was not possible with other Austrian speakers). The truly fascinating aspect about Pidgins is how easy it is to work out a common means of communication between speakers of different languages and how well the whole process of origin and development operates. The development of even an ad-hoc Pidgin was a step by step bottom-up approach, one word (acquisition) at a time, with the amount of words and then grammatical rules and traits slowly but constantly growing. At first, before even an ad-hoc Pidgin arose, single words were acquired; however, the astonishing aspect about early vocabulary acquisition in Pidgin languages is that even before language contact creates an ad-hoc Pidgin the tendency towards one language as the superstratum (vocabulary) definitely exists depending on which language forms a (statistical) majority (in an enclosed space). The actual choice of words

81 is, however, more arbitrary and open (to exceptions) because any word that is comprehensible to a large number of people is preferred over a strictly language-based choice (or rule), although the probability of one language being picked over the other prevails for the majority of choices. As long as intelligibility is granted, vocabulary as well as phrases of any language can be adapted, as there is no strict deterministic rule from which language a word or phrase must be. This, however, slightly changes for the ad-hoc Pidgin state and changes drastically for the fixed Pidgin state, where very few words or phrases become adapted from the substratum language. Words become adapted due to their intelligibility in later states as well, but either if the word itself is very common and used by the superstratum language speakers (e.g. loanwords) or there is no other word available. Most of the words in use were Rumanianzed Austrian German and presumably Italian words which were used by truckers and bus drivers, i.e. /ka:fé/ (long unstressed vowels and short stressed vowels). When the number of Rumanian speakers and thus the Rumanian language element increased, Germanized Rumanian words were added i.e. lapte pronounced with an Austrian /b/ insead of a Rumanian /p/ or ceai and unu optzeci pronounced with an Austrian /tsch/ sound, which to some sounded like Rumanian in its early learning stages - curiously also to Rumanian speakers, who presumably assumed that one was trying to speak Rumanian, but making the notorious early mistakes or inherent errors. Although this Pidgin started with a rather Austrian German touch (in 2001) with many Austrian words including ´melange´ (/melánsch/), a true curiosity. However, once more and more Rumanian speakers were becoming present, and interacted with the Austrian speakers and basic grammatical traits started to appear, the ad-hoc Pidgin quickly shifted towards a Rumanian vocabulary in a few months, and more and more Rumanian words with an Austrian pronunciation became incorporated in this Pidgin simply because the majority of the customers knew Rumanian rather than German. The shift towards Rumanian vocabulary seems quite logical as when 200 people are speaking Rumanian and only 3-5 people are speaking Austrian the odds are that Rumanian will be used as it is, in this specific case and location, the language spoken by the majority (also referred to as the dominant language). To gain an approximate notion about the numbers involved, i.e. people visiting the gas station, it can be stated that on weekends a constant number of 200 people from 8 am to 10 pm and the rest of the week about 50 people were constantly at the gas station and communication was happening continuously. The reason for the high number was on the one hand that the gas station was situated on the highway and on the other that a contract with a Rumanian transport

82 company existed which brought people from Rumania to Spain (and back) who spent a couple of months there as field workers. Buses or vans full of people were constantly arriving, with weekends being more frequented than weekdays; however, at least every two hours one bus carrying 50 people arrived. Curiously the weekends were far too busy to really talk with people and thus communication became very short but efficient, settling only with the most effective expressions, that is expressions that were understood quickly by almost all Rumanian speakers. Simple grammar is essential as there was literally no time to communicate in any other way than in simple one word orders or two to three word phrases of which most were questions, both by personnel as well visitors, about coffee, prices and where to find something. As such, communication itself was limited to the basic needs (what people want/how to get there) that were available in this enclosed space. The weekdays were in some regard more relaxed concerning communication as leisure conversation was possible, although limited, as there were customers constantly requiring attention; however, it was useful in the sense that it allowed new vocabulary to be acquired and the further development of the Pidgin and the speakers´ language and communicative competence. The connection between weekdays and weekends concerning the Pidgin was that what was learned during the week was being tested and, if successful, stabilized during the weekend. However, other words were disregarded, as more than once a word or a phrase was discarded simply because not everyone understood what was meant with it. Confusing bits were replaced and alternatives were sought and, if they were available, these were tested too until the final result was established (i.e. a word, expression or phrase that immediately made sense). The weekends also served the purpose of correcting words and expressions, as the purpose of this Pidgin was quick but efficient communication. After all, it may have been possible to have a conversation with a single person during the week and work out some words and meanings.

5.1 Early developments

Once a steady amount of language contact was established, the Austrian-Rumanian Pidgin started to develop. Arguably, if this was not a completely different process, the shift from Austrian to Rumanian vocabulary could be regarded as one early vocabulary shift as almost the whole Austrian vocabulary was replaced with Rumanian words i.e. ´melánsch´ became (´Káfe´ and finally became) /káfé/, ´Milch´ became /lábte/ or ´latte´ (the Italian word was used because many

83 Rumanians seemed to understand it) etc. and although some typically Austrian words (´melange´) were replaced, the Austrian pronunciation did not became more close to the Rumanian pronunciation. Only some Austrian words remained but rather as doublets which were primarily used by Austrian speakers rather than being a Rumanian speaker choice. /Té/ existed alongside ´chai´ as did ´Mílch´ and ´labte´ and ´latte´. These are examples of dual terms as both of these words could be used in either variety and people would understand it. Some Austrian words would remain such as ´wirstel´ (´Wirschtl´ a dialect variant of Standard German Würstel), ´Schnitzel´ and the Italian ´pizza´. A curious fact is that when an Austrian or German word was used by an Austrian speaker the Rumanian speaker would repeat the same word in Rumanian for example Tee? - Chai! This is similar to the beginnings of the Pidgin. Another example concerns numerials, 1,80 was announced ´eins achtzig´ by the Austrian speaker and repeated as ´unu optzeci´ by the Rumanian speaker. The explanation for the process that only some words were Rumanianized German words while almost all others were Austrianized Rumanian words is that before this massive interaction started, Rumanian truck and bus drivers had been travelling frequently through Germany and picked up some common words which were adapted to Rumanian pronunciation and were used. As they proved successful in communication, they became established (also some Austrian words that were similar to Rumanian words). Such words, picked up elsewhere, were the starting point for the development of the Austrian-Rumanian Pidgin, as they served to facilitate basic communication, although many failures to establish communication were involved. However, through many interactions these words became established and more words could be added so that within a year an ad-hoc Pidgin stage could be reached. The curiosity about this (two-word) stage was that, just as in the German language, compounding was extensively used as in the curious expression of ´pízzamelánsch´ compounding both ´pizza´ and ´melange´ meaning “pizza and a coffee”, although the more common expression was ´pizzacola´ where cola was used as a general term (or hypernym) and could mean any kind of soft drink and was used far more often, ´Káfélapte´ meaning “coffee with milk” but at the three word stage could be become ´Kafelabte senza labte´. Presumably these expressions originated from two linguistic traits from German and Italian grammar. The first linguistic trait, which is probably the cause of this curiosity, is that the Italian language can combine words by removing the pause between them i.e. the Austrian-Rumanian ´Kafelabte´ most likely stems from the Italian pronunciation of ´ caffelatte´ , pronounced without a

84 pause, and was presumably picked up by bus and truck drivers in Italy itself. If picked up without knowledge about the Italian language, this could easily be interpreted as compounding. The second linguistic feature which made this a common linguistic trait in the Austrian-Rumanian Pidgin is that compounding as such is a typical process in the German language, in which almost any two or more nouns can be combined into one word. In fact, compounding can be brought to almost ridiculous levels i.e. ´Donaudampfschiffkapitänskajüte(nputzfrauendienst)´ 9. Thus, it does not seem unrealistic that this would be a common feature in a Pidgin language with Austrian grammar to be extensively used in many different kinds of (language) situations. Curiously, although (almost) all expressions were questions, they were stated like facts, instead of a low-high pitch it was a constant middle-low pitch intonation by the Austrian speaker which was confirmed in a low pitch by the Rumanian speaker, which, strangely enough, worked very well. One further interesting feature of the Austrian-Rumanian Pidgin was the titling of other speakers. Rumanian speakers had picked up the very typical Austrian expression of calling younger people /í:ju:nge/ ( Junge ; the vowels were always long and stretched out). In response, the younger Austrian speakers would start calling basically any Rumanian speaker /schefu/ on purpose which derived from the German word Chef , or Schef in Austrian German, meaning boss. This is common in the traditional employer-employee relation in which apprentice boys were called Junge and the superior was called Chef . One point to add here is that a barman is also addressed as Schef by Austrian speakers. Thus when one was addressing a customer as ´schefu´ one of the employees, including the employer, the actual boss himself, felt addressed. Due to this tradition the common expression of calling both a young man or generally any employee Junge was quickly accepted and adapted by the Rumanian speakers and with it ´schefu´, a sort of Rumanicized Austrian word, became used by Austrian speakers.

5.2 Superstratum language determination

Through a shift of conversation partners, from Rumanians speaking German to Rumanians speaking Rumanian (with Austrians), the main vocabulary for the Pidgin quickly shifted significantly in favour of Rumanian, which seems quite logical, otherwise communication could not have been established. It could further be assumed that this would show a power structure or hierarchy, often assumed in Pidgin theories based upon the slave-master relationship (Selbach,

9 The brackets indicate that this compound can be extend even further 85 2008: 30-32). In this situation, however, it simply (re)presents a statistical percentage of language speakers due to the amount of speakers of one particular language, i.e. the language spoken by a majority. It is the language that most people speak and the language that the majority of speakers understand in an enclosed space (i.e. not in a geographical area but instead in a demographic area) that turns into the superstratum language. After all, Rumanian is not really spoken in Austria, but in the enclosed space of this gas station Rumanian definitely was the majority language. At this point, it is important to note that this is true for any specific environment or enclosed space, especially in a trading situation where many people speaking a different language travel to distant places to trade with people speaking another language, i.e. many merchants from afar trading with a comparatively low number of local merchants. Curiously, the local language becomes the minority language in this specific scenario as more travelling merchants trade with a lesser number of local merchants. This sociolinguistic setup can be applied not only to modern Pidgin formation but should also be a valid process in earlier times and would serve as an explanation why Catalan was taken as a superstratum language. Although there were many more Arabic speakers than Catalan speakers in the geographical area, the enclosed space of trade created a surplus of Catalan speakers and a comparatively small number of Arabic speaking merchants trading with these Catalan merchants. This led to the result that Catalan became the majority language in this trade context, which turned into the the superstratum language while Arabic assumed the function of the substratum language. If Arabic speaking merchants were coming to Catalan merchants then it might as well have been the other way around, not because there were fewer Catalan speakers in the Catalan kingdom, but in an enclosed environment i.e. space of commerce there would have been fewer Catalan merchants that could trade with a larger number of Arabic speaking merchants and thus Arabic would have been the majority language spoken in this enclosed space. The main point in this comparison is that people who travel somewhere and also want to trade there use their own vocabulary if they do not know the local language, and the people at this location immediately respond to it by adapting the new words, meaning that, as long as communication is maintained, the creation of a Pidgin language is facilitated. This is a very likely explanation for the origin of the Lingua Franca. Thus Catalan merchants that came to an Arabic-speaking port to trade were using their language and the traders at this port simply adapted to it either as a sort of unofficial agreement or even because of one specific rule of Pidgin formation itself. Socio-linguistically, one could argue that social, i.e.

86 demographic, pressure would further facilitate the use of the majority language, especially in the enclosed situation of social and language contact in a small demographic area. The linguistic structure of the Lingua Franca suggests that Catalan is the superstratum language; therefore, it can be assumed that more Catalan speakers were involved than Arabic speakers in creating the Lingua Franca and in a trade relation this can only mean that a (large) number of Catalan speakers were communicating with a lesser number of Arabic speakers. Although it may seem illogical to assume that this would have happened on North African soil where a geographical majority would speak Arabic, it is the enclosed space of trade (and the context of Pidgin formation) which allowed such a language situation to arise with the result of a minority language, geographically speaking, becoming the superstratum language of a Pidgin. Assumingly, in another enclosed space or another environment of trade with different social and economic conditions this would not have led to the same result (rather the minority adapting the language of the majority or the minority staying secluded). For instance, the formation of some Pidgins in the South Pacific, e.g. Tok Pisin, where a minority of traders constituted the superstratum language i.e. English, definitely existed because of a difference in power relations which might actually not have been present in the case in the Catalan-Arabic situation of language contact. Tok Pisin was created due to English traders having both economic and military dominance while Catalan presumably only represented a majority in this enclosed space of trade. Furthermore, Catalan merchants were venturing for days, weeks even, without contact to an Arabic speaker. Arabic speaking merchants themselves had contact with (other) Catalan speakers much more frequently and thus were forced to adapt to the majority of Catalan speakers. It does make sense that the Arabic speaking merchants that were visited regularly by many Catalan speakers saw the necessity to create a Pidgin language to be able to communicate and establish trade relations. It seems to be rather the minority language that takes on the grammar while the majority language adapts the vocabulary. Thus it seems to be a demographic relation that determines the vocabulary and grammatical allocation.

87 5.3 Limits of language contact

Interestingly enough, the development of this Austrian-Rumanian Pidgin stopped as soon as it had reached the fixed Pidgin state, i.e. once everything trade-related could be expressed quickly and effectively. Another reason for its rudimentary fossilization can be found in the limited personal interactions and the restricted discussion topics that were explored during these conversations. Only when bus and truck drivers that had been coming to the gas station for years were telling a few tales and interaction became more personal did some more extended topics become part of the conversation. However, this rarely occurred (in fact, with less than five people and only at the end of the six years, of both bus and truck drivers visiting the gas station as well as the Pidgin itself existing). For the most part, the Austrian-Rumanian Pidgin remained oriented towards quick service and not towards establishing close personal relationships. The Pidgin remained a polite, but distanced, way of communication in which no deeper personal involvement was possible. After all, it only reached the three word stage resulting in ´Káfe con labte´ or ´Káfe senz(a) labte´, with the preposition ´con´ and ´senz(a)´ being one of very few prepositions used. Curiously, the Italian preposition senza , ´without´, instead of the Rumanian fârâ was used due to many Rumanians understanding this word. However, Rumanian speakers would sometimes leave out the final vowel resulting in ´senz´ rather than ´senzá´. Prepositions were also used by the Austrian speakers who would further intensify compounding in this Pidgin. The compounding pattern emerged because stressing the prepositions ´con´ or ´senza´ made it easier to distinguish if someone wanted something or not. Rumanian speakers, especially those that were not frequently visiting, used compounds and quickly started using these prepositions. At the same time they corrected themselves and frequently added ´no´ to the end of the preposition ´Káfe senz(a) no labte´ and others would even repeat the whole sentence ´Kafe senza labte, (Káfe) no labte´. Judging from this experience, personal interactions (including conversation topics) are essential for the development of a Pidgin into a Creole. Pidgins are functional in the sense that what needs to be expressed is developed through trial and error until a solution is found. Once communication needs are satisfied, there will be no further development, which seems logical as a problem stops existing once a solution is found. Furthermore, language usually expresses what people have established through communication. Thus if there is no need for a topic or specific action to be expressed, further communication will not develop. This is also the main point about different Pidgin states, i.e. the communication needs or range of expressiveness change 88 depending on the Pidgin states themselves. All Pidgin states seem to have a certain limitation as to how much can and needs to be expressed for communication to succeed, i.e. a need for specific topics to be expressed but nothing beyond that and only once this basic need (of expressiveness) arises through either an increase in language contact or a desire to express more complex concepts or the intention of using the Pidgin in a different (social) environment can the Pidgin state be expanded as well (by adapting to an increasing field of topics and ways to express actions and social status). That is the further development of a Pidgin state is only initiated if the need for it arises. The reason for this is that languages themselves are dynamic and flexible or, to use another term, they are a reactive medium in the sense that the development or the origin of a language occurs as a reaction to desired but failed communication which is set in motion after contact has been established. This has, for whichever reason, become a necessity as many people become involved and start using the language in all communicative domains on an everyday basis. The Austrian-Rumanian Pidgin would have evolved beyond the fixed Pidgin state if there had existed a need for an expansion of it in use, i.e. if more personal interaction had become more common in the environment. The context/setup of the gas station and the reason for the very existence of this Austrian-Rumanian fixed Pidgin made a further development impossible, as (the regular) communication (alone) was only focused on the most important aspects of functionality and not on establishing a variety of topics or expanding the limits of communication. Judging from the very few occasions when customers became regulars and expanded beyond the already established topics with the consequence of establishing new words and phrases and even grammatical traits, the possibility of developing the Pidgin further was too limited. Furthermore, from a geographical and economic point of view, one could hardly argue that this Pidgin could have reached the extended Pidgin state even if it had continued to exist. By definition, this would require that people outside the original social environment would take on the Pidgin, which seems very unlikely to have ever happened, as it would have been necessary for many more people or locations to get involved in this contact scenario. Curiously, the definition focuses more on or integrates more of the social factors into the definition of a Pidgin than linguistic attributes. It has to be mentioned that the Pidgin in its fixed Pidgin state was at the very beginning of this point of (even) being (called) a fixed Pidgin and therefore strongly limited. Expressing something in this Pidgin was quite difficult and always sounded clumsy, as it basically existed as a sequence of two and three word phrases (e.g. ´mit

89 Bus gefahren´, ´Spania gearbeitet´, ´mit Kollege (/kolega/) gefahren´). What is the most important aspect about this Pidgin state is that one was always dependent on the help of others, or, to be more precise, if others did not understand the general message and could not help out with words it was basically impossible to reach a mutual understanding. Even though, after some occasions, establishing communication worked more and more fluently, this was due to both speakers helping each other out, i.e. teamwork was automatized, and not due to the structure (or limitation) of the Pidgin itself (being fixed). Thus it could not truly evolve further or ever reach a further, i.e. extended, state.

5.4 The vanishing of the Austrian-Rumanian Pidgin

The collapse of the Austrian Rumanian Pidgin in 2007 was of an economic nature and caused due to a political decision. People from Rumania were hired as field workers in Spain and these Rumanian people were brought to Spain and back via buses, stopping at the gas station about halfway through the journey. This established a quite lucrative trade between the transport company and the gas station, which ensured that the buses would get cheap fuel and the gas station got a constant flow of customers, as people need supplies and facilities for a 20 hour drive. However, this contract only lasted for about 5 years after which the economy for hired workers in this form collapsed almost immediately, within three months. The reason for this was a change in EU politics (which forbade hiring migrant workers in this manner); therefore, the gas station was no longer visited by a large number of Rumanian people, i.e. no Rumanian was spoken, and with it the need for the Austrian-Rumanian Pidgin became non-existent. Most curious is the sudden disappearance of this Pidgin itself because it was already in the fixed Pidgin state. However, once language contact ceases, a Pidgin whose existence is based upon (continuous) language contact, immediately vanishes. Interestingly, after this Pidgin vanished, the occasional truck and bus drivers began to speak German again and later more and more English was used.

90 5.5 Limitations of the Austrian-Rumanian Pidgin

The Pidgin was never used outside of this location described and it was never used with other Austrian or German speakers due to it having Rumanian as a superstratum language. Thus, this Pidgin was less effective for communication in Austria, as not many Austrian people speak Rumanian and there were no other locations in which constant large scale communication between Austrian and Rumanian speakers existed. It was hardly possible to converse with this Pidgin in a small-talk manner as there was not enough vocabulary that would cover private or technical topics etc., making a lively chat with a Rumanian speaker almost impossible as it always remained an unsteady conversation with considerable searching for words and relying on other speakers to reach a consensus. Although this Pidgin could be seen as speaking some kind of basic Rumanian with Austrian grammar it still was not very effective to communicate with a German speaker. This is why the bus and truck drivers reverted back to using a Rumanianized German (and also Italian, presumably because the main bus and truck routes from Rumania went through Germany, Austria and Italy). Furthermore, it was impossible to converse with an Austrian speaker as the vocabulary was primarily Rumanian, despite the Austrianized pronunciation and the occasional Austrian or German word.

5.6 Conclusion

The reason this Austrian-Rumanian Pidgin offers insights for the Lingua Franca is that although this gas station was in Austria, the main source of vocabulary was Rumanian; furthermore, relating to the vocabulary shift aspect of Pidgin languages, on more than one occasion a succession of buses of Italian tourists visited during the course of a weekend and the basic vocabulary was switched immediately to Italian. This worked surprisingly well and happened immediately each time a similar situation with a different language occurred and although the vocabulary shifted back to Rumanian, a few very basic and similar words such as `labte`(milk), ´Káfé´(coffee) etc. and even some grammatical forms were adapted. It was the people that came to, not the people at, this gas station that offered the vocabulary but it was the people at this location that quickly adapted and used the new (Rumanian) vocabulary but not its (Rumanian) grammar and, instead, were reassembling the new words according to Austrian grammar (or rather the Carinthian dialect). Pronunciation, although also Rumanian, became Austrianized, meaning that Austrian speakers

91 were repeating Rumanian words but keeping the most common pronunciation features of the Austrian dialect i.e. /a/ -> /o/ as is (although not as strong as in the common Carinthian dialect, as imitating Rumanian would otherwise not have worked, i.e. caused too much confusion in the pronunciation). Using this Pidgin made trade much easier and faster, which truly was a necessity, as buses carried up to fifty people at a time and the people wanted to buy refreshments and snacks for their travel. Overall, this example shows the development of an ad-hoc Pidgin and that its vocabulary expansion can occur within a short amount of time, including the establishing of words and even phrases that are based on the immediate use of new vocabulary without abandoning established vocabulary when people with different languages enter this language situation. If the intensity of contact increases, the further development into a (fixed) Pidgin is one possible development for Pidgin languages.

Chapter 6: The nature of Pidgins

The experience with the formation of an ad-hoc Pidgin definitely helps the understanding of Pidgin language development at large because language contact and consequently Pidgin formation most definitely arose in similar contexts of commerce (in the 12 th as well as 21 st century). One important conclusion that can be drawn from the Austrian-Rumanian Pidgin is that communication needs to be constant, which is achieved either by a large or a stable number of people communicating. Once language contact is disrupted, a Pidgin, especially an ad-hoc one, but even a fixed Pidgin, vanishes. The key ingredient here is continuity of language contact which includes the repetition of vocabulary, phrases and grammar. Repetition is important for the development of a Pidgin and is a crucial aspect especially in the beginnings of Pidgin language formation, i.e. ad-hoc Pidgin state, as this is the point in time when a Pidgin can vanish without trace immediately if continuity (and with it repetition) is not upheld. This repetition process is responsible for both the use and/or disposal of words and their meanings including phrases as, in the process of repetition, the more often a word is used, i.e. repeated, the more likely it is taken over and used by others. The less a word is used, fewer or no speakers at all will use it thus preventing its spread and intake into the Pidgin corpus. Repetition, in this case, is a less general concept compared to continuity of contact itself and

92 instead is a more detailed aspect of Pidgin language (vocabulary) development and concerns meaning of words and phrases including prepositions. Through repetition meaning is established, but also ´meaning in context´, which is most important for a Pidgin language as, especially in the ad-hoc and fixed Pidgin state, the meanings of words are always bound to a specific context, (be it trade, tourism etc.) out of which phrases arise quite often, i.e. the repetitious use of a certain word order. Phrases are a sign of a fixed or the beginning of an extended Pidgin state, as whole phrases can only be established once certain words have become part of a basic repertoire. This can lead to phrases that are like the ones in the superstratum language. Phrases are also often taken over from the source language(s), but in most cases it is phrases that, even when rooted in phrases of a source language, are different to the original language, as these phrases either differ in meaning i.e. words have different meanings from the source language(s) or differ in structure, i.e. word order of the phrases used, grammatical features etc.. In any case, a Pidgin orients itself towards its original language(s), but phrases differing to the meaning of the source language constantly emerge. Concerning single words, it can happen that two terms are used for one word. Usually this leads to a local speaker having two words with the same meaning at their disposal which can further vary within the Pidgin. Frequently, terms stemming from the local or substratum language are specialities that only exist in that region or culture. For instance, in the Austrian-Rumanian Pidgin, words such as Melange (Viennese variety of coffee) and can replace or exist alongside general words such as /ka:fé/ (coffee) or /té/ alongside with /tschai/ (tea) or latte alongside /labte/ (milk). Such a scenario is very likely for the Lingua Franca as well, due to the fact that merchants from other nations travelled to the same North African ports, establishing contact and communicating with the (same) local merchants. This most likely enriched the vocabulary and extended the expressive means of the Lingua Franca. The establishment of words must have occurred on a regular basis (as it did in the Austrian- Rumanian Pidgin), with some words eventually becoming fixed (or learned), maybe even widely used. Constant language contact, in both Pidgin cases caused through commerce, with same language speakers using one specific language caused a set of fixed vocabulary to become established. Furthermore, language speakers used the same general terms for the things they wanted to trade for. The same process seems to occur in the expanded Pidgin state.

93 6.1 Catalan-Arabic language contact

In the scenario of Catalan-Arabic contact, facilitated through commerce, communication between Catalan and Arabic speakers in the Southern Mediterranean resulted in the adaptation of the Catalan language on part of the Arabic speaker. Local Arab merchants adapted to the foreign language while the Catalan travelling merchant would use his own vocabulary and language. The adaption of the Arabic speakers meant that they acquired Romance/Catalan vocabulary and applied that in conversations with other Catalan merchants. This is an important aspect of vocabulary acquisition (and consequently ad-hoc Pidgin development), as for the participant that adapts to the foreign vocabulary it does not matter if communication is established with the same person but communication needs to be constant with the same language speaker. It is sufficient if some speaker of the same language communicates with the person adapting the vocabulary with the result of both adaptation as well as use of the newly acquired vocabulary on the participant side. As the new vocabulary is tested or used upon speakers of the same language, soon a realisation process will occur presenting the possibility of mutual communication among any speakers of the same particular language, or, in the best case scenario, communication with speakers of other similar languages, i.e. Romance. This will further lead to the acceptance of this newly adapted vocabulary as well as the intake of other vocabulary and the consequent origin, development and expansion of the Pidgin and its vocabulary. When a North African trader realises that they can communicate with any Catalan speaker, they will readily adapt to this new vocabulary, also realising that with such an ad-hoc Pidgin one is able to communicate much more effectively and successfully. After all, being able to communicate is a big motivation especially if it also allows communicating with other language speakers that do not speak the adapted but a similar language, i.e. another Romance language such as Occitan. Communication with speakers of similar languages is possible due to language kinship and the existing similarities in vocabulary shared by such a language group, for instance Romance languages. While the Arabic speaker acquires Catalan words and uses them with speakers of other Romance languages they will be able to understand the basic vocabulary due to this kinship. The Arabic speaker himself is able to reassemble the newly acquired words due to an already existing linguistic system of his own language which allows the Arabic speaker to reassemble the newly acquired words according to an already predetermined grammar. Thus a certain continuity is guaranteed, including the use of the (own) intonation system to emulate the foreign vocabulary pronunciation. This whole process is upheld as long as continuity of 94 communication exists and an ad-hoc Pidgin will consequently arise, which is established by itself and basically determines the use of certain vocabulary and the intensive/contextual use of words. The repetition of communication, which is enabled through this, is quite an important factor that this process is dependent on.

6.2 Important aspects of Pidgin languages

At this point, some assumptions about the nature of Pidgins must be made. The first is that Pidgins orient themselves towards the languages that lead to their creation and thus remain similar to them. The second is that Pidgins do not easily change their grammar although they easily change their vocabulary. The third is that Pidgins usually stay both in the geographical area and surroundings but also in the social environment in which they were created, including those in which they became used. It may not be an irreversible rule, but it is true for the Austrian- Rumanian Pidgin and it also seems to be true for the Lingua Franca. When the first assumption is applied to the Austrian-Rumanian Pidgin, it becomes apparent (and can be confirmed that) the superstratum language was Rumanian due to Rumanian speakers being the travelling agent while Austrian speakers were the local agent adapting Rumanian vocabulary by applying their own grammar. This is parallel to the contact between Catalan and Arabic speakers with Catalan speakers being the travelling agent and the Arabic speakers the local one. According to the previous assumption, it could be suggested that the primarily used vocabulary must have been the Catalan language, reassembled according to Arabic grammar by the local North African merchants therefore being closely tied to both source languages with the superstratum language determining the Romance affiliation, i.e. the Pidgin appears to be a Romance language. The second assumption is that vocabulary can easily be replaced, however, vocabulary itself does not change, i.e. shift, as long as the main input from the superstratum language itself is upheld. Only when influences of other languages become present does a shift occur. Grammar seems to change less easily than vocabulary, but also changes if the degree of a different language influence increases and, at the same time, the superstratum source language input becomes less present. This process is, of course, not exclusive, meaning that it can not be completely ruled out that grammar features are taken from the superstratum language, although it appears to be less common. After all, some minor adaptations primarily affecting the basic grammatical traits (e.g.

95 infinitive, prepositions and negations) can occur especially when these changes are barely noticeable ( andar and andare , como and come or no and non ). A major grammatical shift only comes about if both the superstratum influence is weakened and becomes replaced by another language which gains more importance and thus influence on the Pidgin.

Grammatical features in Pidgins stay stable due to people using these grammatical traits without seeing them as grammar but more as words (that are understood by other people) and not due to their arguable simplicity. After all, it is much easier to use one word and apply grammatical meaning(s) to it as using several words to differentiate the real grammatical meaning(s), e.g. personal pronouns. In the case of the Lingua Franca mi and ti10 were the only ones in use but could be applied as any person or form of them (reflexive, possessive, i.e. ´me´, ´my´, ´mine´, etc.), differentiating only between first person and third person (although ti in Old Catalan is used for the second person (Monk, J. 2013)). If something is understood by others the main purpose of communication is fulfilled; therefore, grammatical traits might be the ones that last the longest because they are either shared by other languages, which applies in the case of Romance languages, or because they are used predominantly (as seen in later instances of relexification). After all, grammar is an essential part of the structure of language. It is most likely that these are among the first words to become fixed in a Pidgin and, following the principle that first words learned are forgotten last, it could also serve as an explanation as to why grammatical forms in Pidgins (or (incompletely) learned languages) do not (easily) change. However, the claim that grammatical forms are most stable can not be entirely true, as not only vocabulary but also grammatical features can be replaced if the dominant language is exchanged. Although grammatical traits tend to be the most stable and static features about Pidgin languages, they change eventually through the adaptation of another language when the input language becomes interchanged by another without the intensity of input decreasing. In the case of both Spanish as well as Italian takeovers, which grammatical words were affected and depended solely on what is commonly used in a language and what is not. Only if a language insists on using a different grammatical expression for a grammatical trait is this word exchanged. For example, in the case of Petit Mauresque (and Sabir) ne …pas11 was adapted for expressing negations, which seems to be due to French employing this form of negation compared to the Lingua Franca

10 Cifoletti, 2004: 26-27, 43 and Schuchardt, 1979: 29, 34 11 Broughton, 1839 quoted in Cifoletti, 2004: 239; Waille, 1884 quoted in Cifoletti, 2004: 274 and Mayet, 1887 quoted in Cifoletti, 2004: 276 96 negation particle non12 (and the less common no). Changes in grammar are still rather rare as most grammatical features, at least in Romance languages, are so similar that there is no necessity for them as they can be understood easily by other Romance speakers. However, due to a definite but only slight change of Lingua Franca grammar in its varieties an alternative explanation can be given for this slight change of grammatical traits in Pidgins found in the example of the Lingua Franca. Most grammatical features are very similar in Romance languages and thus it would be very easy for a different Romance language Lingua Franca variety to arise, as only a slight change in basic grammatical traits would be necessary. For instance, the adaptation of the inflectional infinitive form of Lingua Franca ar with the Italian infinitive form are or the adaptation of the past participle form of the (Italianized) Lingua Franca – to (sometimes - ato, - ito) and the Hispanicized form - ado (e.g. mandado ) could easily be achieved (Cifoletti, 2004: 43). For the most part there is no necessity for a grammatical shift as basic grammatical traits can be understood even by speakers of other languages (as in the curious case of the Austrian-Rumanian Pidgin preposition con (instead of fârâ), which was used by Austrians but was understood by Rumanians). Therefore, it can be concluded that a grammatical shift is a rare occurrence and is caused by language change and considerable input. However, if the conditions are met that languages are quite similar, this could be a much easier process than anticipated because very minor changes in basic grammatical traits can easily be made. What determines the stabilisation of such a variety is if language input from a similar language is upheld and the use of such a variety continues to exist.

6.3 Vocabulary and grammatical shifts in Pidgins

Clearly, a so-called shift is not bound to vocabulary. The whole process of shifting (or, more appropriately, adapting) vocabulary, which in this intensity is unique to Pidgins, seems to have a remarkable flexibility concerning “fixed” expressions. What can be observed is that there are more changes in vocabulary, which seems to be nothing unusual for any Pidgin language (and even more so in a Creole language), due to the interchangeability of vocabulary itself. However, there are also changes in grammar or grammatical shifts, although to a lesser degree. Both these common vocabulary and rare grammatical shifts (or their occurrence in the Lingua Franca) can

12 Rehbinder, 1798 quoted in Cifoletti, 2004: 44, Schuchardt, 1979: 29 and Haedo, 1612 quoted in Cifoletti, 2004: 37 97 fairly easy be explained through a linguistic as well as a geo-political analysis (although not the reason why these shifts occur in general). In linguistic terms, shifts occur because of a change in language input and linguistic set-up. Vocabulary shifts can occur depending on which language is present (or dominant). If one language is replaced by another and the input is consistent then a shift will most likely occur. If the geographical area is huge and the linguistic set-up, i.e. the language input, changes in some regions but not in others, several different vocabulary shifts can occur with several different varieties arising. However, there is also the possibility that the Pidgin vanishes or is replaced through the different language input. In general, a change in language input is not a guarantee for a linguistic shift, more a probability or possibility of what can happen. Unfortunately, this does not truly give an explanation as to why these shifts occur in the first place in Pidgin languages, or why it has the effect of changing a language at all when it could also create another new language. One logical reasoning for shifts would be that Pidgins are not fixed and stable languages, which means that a shift can easily and quickly happen as soon as a change in the linguistic set-up becomes noticeable. Although this is not a substantial explanation, at least the shift phenomenon itself is a recognizable reoccurring pattern which can easily be seen through both a linguistic as well as a geo-political analysis. While the linguistic analysis shows which languages were involved and which changes occurred, the geo-political analysis allows to retrace where, when and how these shifts occurred during the development of the Lingua Franca. A combination of these techniques proves very useful for the study of the Lingua Franca. The proof for the assumption that shifts occur easily and quickly in the Lingua Franca is due to a report by Haedo (1612) about Portuguese captives that were brought in large numbers to Algiers as prisoners after a lost battle (in 1579). This created a newly found Portuguese vocabulary among the speakers of the Lingua Franca, i.e. a shift towards Portuguese in the Lingua Franca vocabulary, within a very short period of time (Haedo 1612 quoted in Foltys, 1984: 13 and Cifoletti, 2004: 198) 13 . Therefore, if an estimated number of 10,000-15,000 captives can cause an expansion or a shift in the Lingua Franca vocabulary then strong political and economic influence (or a politically motivated situation) over the period of a whole century most definitely can cause a shift in the Lingua Franca structure and vocabulary. Depending on the intention or the politics of a political power, influence can go in both directions. An active political power (but also a policy) can cause political, economic and

13 Unfortunately there are no examples given by Haedo 98 language changes, and supposedly influences the development of a different language variety, just as the politically motivated preservation of the same circumstances or the preservation of an economic model in a different political environment can also prevent linguistic change itself. Presumably, this was the case with the Catalan Lingua Franca, which had no vocabulary shift within its first century of existence, even after the taking over of the Crown of Aragon by the Spanish Crown. Only after Catalan declined and lost its influence after almost a century did Spanish influence start to become noticeable via a vocabulary and slight grammatical shift. In fact, two separate vocabulary shifts occurred due to this change of power in favour of Spanish and Italian but at the expense of Catalan. As the analysis of the vocabulary development has shown it can be assumed that Catalan must be the source, i.e. superstratum language, of the Lingua Franca (besides the Arabic language). Politically the Catalan kingdom spread in the whole Mediterranean until Catalan was omnipresent in several different political, economic and scientific domains from the 13 th to 15 th century 14 . However, this all changed around the 15 th (to 16 th) century when several conflicts and political problems led to the end of the Catalan kingdom and the demise of Catalan itself. Through the Spanish takeover of territories belonging to the Crown of Aragon, Catalan became slowly replaced (by the 16 th century) by Spanish in the western but also the central Mediterranean. Sicily, formerly a Catalan colony, also became Spanish for a century (Chaytor, 1933). The importance of Catalan started to decline in the 15 th century with (the) Spanish (language) slowly taking over, but its decline became noticeable only in the 16 th century. Furthermore, through the disappearance of Catalan, Italian also became very influential as in the previous centuries ´competition between Catalan and Italian was predominant´ (Valérian, 2014: 85) and, presumably parallel to Spanish, Italian could establish itself in the central Mediterranean as well. As a result Italian then became dominant but also accepted and popular in the central Mediterranean and, after all, it was the time of the Italian renaissance. Both Spanish and Italian influence diminished in the 17 th century once the Mediterranean trade became less important than the Atlantic and Pacific trade and the Lingua Franca thus became more international. From a linguistic viewpoint, the political development of the 15 th century in which the fading of

14 (http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710b0c0e1a0/en_GB/i ndex619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextchannel=77595c43da8962 10VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aR CRD&newLang=en_GB) 99 the Crown of Aragon occurred had no immediate impact on the Lingua Franca because Catalan remained influential as well as spoken until the 16 th century, as was the Lingua Franca in the Catalan variety, before it finally and completely declined until the 19 th century. Therefore, the Catalan variety only gradually underwent a vocabulary shift, which was a slow process that was completed only by the 16 th century due to the Catalan language being spoken in the Mediterranean (and the Catalan Lingua Franca still being used by Arabic speakers). Furthermore, Spanish speakers had gotten used to speaking the Lingua Franca in the Catalan variety. Thus, there was only a minor shift, comparatively late i.e. in the 16 th century, within the most basic vocabulary and hardly any grammatical shifts (except the most basic occurred). Therefore, no replacement of the Lingua Franca with a Spanish Pidgin occurred, instead a Spanish Lingua Franca variety developed. A quite similar situation occurred with the extension of the influence of Italian. Consequently, a vocabulary shift occurred early on followed by a grammatical shift later on. This shift is attested for instance in Caronni (1805 quoted in Cifoletti, 2004: 229). One major shift, bigger than the Spanish one, occurred in which grammatical forms (i.e. the infinitive) seemed to have changed: andar becoming andare , and bisogno being used as a tense auxiliary to form the future and the past form etc. This shift was more intense than the Spanish one because Catalan had already declined in the central Mediterranean. Although the Catalan variety of the Lingua Franca was still spoken by many Arab merchants, thus Italian merchants also used the Lingua Franca at first in its (more or less) original Catalan variety, the Lingua Franca eventually became Italianized and was well known (but not spoken) in the Italian cities. What is noteworthy is that the political changes did not cause the Lingua Franca to be replaced by a new Italian Pidgin but caused a vocabulary as well as grammatical shift (as politically there does not seem to have been any interest in colonisation). Curiously, both the Italian as well as the Spanish varieties only evolved once Catalan was fully gone as illustrated by reports about the Lingua Franca in the 15 th century i.e. 1420 by Dopp (Dopp, 1958 quoted in Foltys, 1984: 8), but the Spanish influence only becomes visible in sources from the 16 th century i.e. the poem by Encina mentioned in Foltys (Foltys, 1984: 10). Therefore, Catalan must have played an enormous role as a language, otherwise the vocabulary shift towards Spanish and Italian would most definitely have happened a whole century earlier. In any case, both shifts, Spanish and Italian, caused a vocabulary shift and even the replacement of some grammatical traits, but to a lesser extend in the Spanish variety. The most important or

100 rather most influential shift which differed immensely from the previous shifts occurred in the 19th century with French. The French influence was enormous as it did cause a vocabulary shift (at first), but almost immediately after the development of a Frenchified Lingua Franca variety the Lingua Franca became replaced by a French Pidgin. This was presumably due to the weakening of the Lingua Franca prestige due to the French military occupation, settlement and growing importance of the French language in Mediterranean commerce. Most interesting is that ´the conquest of Algiers in 1830 not only ended (legal) piracy but also saw the manifestation of the French dominion in the realms of culture as Algiers was rebuilt or re-imagined as a clone of Marseilles´ (Abulafia, 2014:147), which most perfectly explains the great influence of French. The conquest of Algiers in 1830 had not only the immediate effect of the Lingua Franca shifting in vocabulary (within five to ten years) but also greatly affected its grammar over time, which can be seen in Sabir. This is rather unusual compared to the previous (Lingua Franca) shifts. Starting in the 16 th century but especially since the 18 th century, French gained commercial dominance in the Mediterranean, for the most part over the Ottomans (Greene, 2014: 101). A political change occurred with French becoming the official language of commerce in the 19 th century, which strengthened the French (language) input and caused the final replacement of the Frenchified Lingua Franca. Although similar changes happened in the 16 th century, the consequences of the 19th century change was both a vocabulary and grammatical shift as well as the final replacement of the Frenchified Lingua Franca (so-called Sabir) with a French Pidgin. The formation of Petit and Grand Mauresque, which, according to the sources, are supposed to be the Pidgin and the Creole stage of the same Pidgin, could be explained by a French occupation including settlement, colonisation and Frenchification of Algiers itself. The aspect of colonisation, however, was not the case in the 16 th century where political but predominantly economic dominance was present in the seeming colonies. There was no colonisation process in the common sense which would have included complete political, economic and language dominance because “what was “colonial” was the system of exploitation of the soil” (Abulafia, 2014: 150). It can be concluded that a vocabulary shift is closely tied to a political shift. Whether such a shift may have any consequences on the language is another question. For instance, the shift from Catalan to Spanish did not have an immediate effect regarding language use as the Catalan Lingua Franca was still used without any difference. Even once Hispanicization took place only the most basic vocabulary and hardly any grammar was truly affected, which was similar to Italianization, which also did not have an immediate effect on the Lingua Franca itself. This is in

101 direct contrast to the Frenchification caused by political occupation which did have major political and linguistic consequences, i.e. vocabulary shift, grammatical shift and even the disregarding (or abolishing) of the Lingua Franca. Political changes only cause a shift if language is affected as well, which seems obvious given that the Lingua Franca is a Pidgin which is easily affected by changing language input. It is difficult to predict the development of a political entity or the political motivation behind it. In other words, the policies carried out may not have the desired result and thus the consequences for a language are difficult to foresee linguistically. The final political(ly motivated) change that truly affected the Lingua Franca was the introduction of English as the official language of commerce and transportation which could have caused a shift in Lingua Franca vocabulary had the Lingua Franca not been weakened, i.e. been replaced (as the language of commerce), almost become extinct and even become irrelevant for commerce of the (middle) 20 th century. If the Lingua Franca had still been present in Mediterranean commerce, then it would not have been unlikely that the Lingua Franca would have adapted to this final political change of the 20 th century. This is also mere speculation as there is the problem of the Lingua Franca working splendidly with Romance languages, especially earlier forms which were more close to each other, but could probably cause communication errors with non-Romance language. In the process of relexification, the Lingua Franca would still have underlying Catalan (vocabulary) as well as Arabic grammar which could cause linguistic complications (the transition process may not work fluently) or almost completely alter the form of the Lingua Franca itself (for instance through English vocabulary). One further (small) point to note is that the Lingua Franca adapted much more easily to older forms of Romance languages and rather became replaced by the newer ones. These older (Romance) forms were far more close to each other in the 16 th and especially in the 12 th century than they were in the 19 th or 20 th century. This could be the reason why such a shift would occur easily and quickly in the 16 th century without breaking the flow of commerce, i.e. an immediate occurring vocabulary shift and also why the Frenchified Lingua Franca would not uphold and could not withstand the replacement by another French Pidgin. Concluding, it seems to rather be the linguistic aspect (or linguistic correlation) which is the determining factor, as it may not matter to the Arab merchants which language they use to communicate with the European merchants, but it may be a lot easier for European merchants to pick up a language that more closely resembles their own, even if this is another Romance or other related language. However, considering that Pidgin languages are not dependent on the affinity of languages to

102 successfully alter their vocabulary, i.e. for a vocabulary shift to occur, consequently, English could have become the source language of the Lingua Franca, even if this scenario seems rather unlikely, as there always exists the possibility of replacement. It was the case with French and it could have been the case with English. In fact, English has replaced the Lingua Franca; however, not by being its superstratum successor (i.e. a vocabulary shift) nor directly taking over and replacing the Lingua Franca but by becoming a lingua franca itself. Being a so-called Kontaktsprache despite being a fully developed language seems to be unusual as a lingua franca usually rather consists of two languages interacting with each other and thus forming a Pidgin. However, this is achieved by speakers from all over the world adapting English to their own mother tongue, thus creating (dialectal) English which allows speakers from different mother tongues to communicate quite successfully with each other in lingua franca English. The third assumption above of Pidgin languages not leaving either their place of origin or their social field is a more complex topic. The Lingua Franca (in particular) seems to defy the theory of close social as well as geographical vicinity due to being spread across the Mediterranean and even being introduced to other social fields. Although always remaining close to and involved in the field and vicinity of economics (geographically), admittedly, not leaving the communicative domains can not be fully applied to the Lingua Franca since an abundance of social, political and economic factors have contributed to its spread through Mediterranean society. The Lingua Franca, if not the exception to the rule, is indeed unique in many aspects. However, if historical and social developments are considered (and analysed), logical causality in its spread can be discovered as the Lingua Franca was a Pidgin language established, presumably, by travelling merchants. Hence this Pidgin is bound to become diffused in far spread and seemingly disjointed, but connected in the commercial sense, parts all across the Mediterranean. As travelling merchants had different locations in which they traded and established constant contact, strictly speaking, the Lingua Franca, as such, did not change the main reason for its spread although it did leave or expand its geographical area of origin due to travelling merchants establishing contact from the time of the expansion of the Crown of Aragon in the western Mediterranean. This makes the third assumption still apply but only to a minor degree. While the Lingua Franca place of origin itself may be seen as not bound to one distinct geographical location but to a whole trading network, concerning the social field, there is a discrepancy here, as Pidgins usually do not easily become introduced to a new demographic context. The Lingua Franca, however, despite this tendency, landed in several different kinds of social areas and

103 communities presumably due to its social flexibility and rising political relevance (although formed without political motivation), but also due to the economic rise of piracy and slavery, i.e. political-economic factors. One reason for this leap into several different demographic contexts is that the Lingua Franca, an originally merchant trading language, quite possibly became used on (all) commercial ships at some point in time, presumably once Catalan had an enormous importance in commerce and was the official language of the Mediterranean. Considering that the Consulate of the Sea was the sea law and written in Catalan, the assumption of Catalan being used on all non-Catalan merchant ships seems quite possible. Although there is no explicit evidence by other sources, the ´Generalitat de Catalunya´ makes a more general claim about Catalan being used by the navy of the Crown of Aragon and, of course, commerce 15 . It seems logical that this could have had a great influence on naval travelling in general due to Catalan being spoken on many ships. It also became the language learned by sailors, merchants (travelling merchants were also sailors (Chaunu, 1979: 261; Wansbrough, 1996: 6)) but also Arab pirates and slavers and (converted) renegades. The establishment of Catalan as the Mediterranean language also makes such a claim quite plausible. The circumstance of many sailors, merchants and also pirates acquiring the Lingua Franca to be able to communicate with European merchants was based upon commerce, but also piracy and slavery, as many European merchants were abducted which led to a massive accumulation of captives in one of many slave strongholds, namely Algiers, Tunis and Tripoli. Many European languages were present in these cities and, consequently, to be able to communicate with these captives many Arabic speakers came into close contact with the Lingua Franca. The consequence was the introduction of the Lingua Franca to another communicative domain, i.e. the local population of commercial cities. The last transition into another communicative domain is connected to slavery introducing the Lingua Franca to the local population. This had the effect of spreading Lingua Franca popularity across a major part of the Arabic and Turkish speaking population. Due to such large numbers of Lingua Franca speakers this Pidgin became known and used even in the circles of Southern Mediterranean diplomats, causing also awareness among the Northern Mediterranean European population, although not

15 (http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710b0c0e1a0/en_GB/i ndex619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextchannel=77595c43da8962 10VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aR CRD&newLang=en_GB) 104 with the same result of adoption but, to a surprisingly large degree, knowledge about this Pidgin language. The socio-demographic expansion seemingly is determined by the number of people involved, but also, in the case of the Lingua Franca, by a(n) (existing) development into an expanded Pidgin. It does make sense for a Pidgin to be used in the exact same location where the source languages established contact. After all, this is exactly where a Pidgin will arise and be used, i.e. in both the geographical as well as social vicinity of the source languages, establishing and maintaining contact. People that do not share the same mother tongue and come into close contact with each other usually also remain (in contact) in the geographical vicinity, which is especially true for the ad-hoc and fixed Pidgin state. However, this can change considerably in the expanded Pidgin state, as it also makes sense that heavy use as well as the popularity that contributes to an increased use of a Pidgin language, i.e. repetition, not only causes a Pidgin to develop but also strengthens the likeliness for people with different mother tongues to learn the Pidgin. Therefore, the more highly evolved a Pidgin language becomes and the more people are involved in language contact, the more likely it becomes for a Pidgin to expand its social and, with it, its geographical field. This can even include speakers of languages different from the source languages. Furthermore, a Pidgin could be seen as a variety of one language or rather the essence or core of one language that is meant to communicate with the linguistic means of the speaker of another language so that speakers of both different languages can communicate. Through an increase in possibility to communicate with more people, an extension not only in the geographical but also in the social field can occur. Arguably, a Pidgin language in its fixed or extended form can be seen in terms of Second Language Acquisition, with the difference that a Pidgin integrates new words into an already established system keeping the basic grammatical traits, instead of fully adopting the traits of another. The same strategy occurs at the beginning of Second Language Acquisition where language strategies are applied, i.e. in learning new words, pronunciation and grammar etc. The difference, however, is that while Second Language Acquisition involves a correction process as the acquisition of the full language is the aim, the essence of a Pidgin is not to acquire a new language but to create a means of communication based upon (two) already existing languages as quickly and effectively as possible. Therefore, correction does not occur in the strict sense, but rather serves as a stabilizer of terms as (through repetition) it is determined what achieves communication successfully and what does not. Curiously enough, this happens automatically.

105 Chapter 7: political analysis

Political changes can cause linguistic developments such as vocabulary and grammatical shifts and there were several major political changes which affected both politics as well as the socio- economic development of the Mediterranean and the Lingua Franca. These occurred since the demise of the political and economic dominance of the Crown of Aragon, which can become visible and be analysed through a geo-political analysis. Since the fading of the Catalan kingdom, Spanish and Italian influence in the (western and central) Mediterranean was constantly growing (since the 15 th century). Furthermore, piracy from the southern Mediterranean shore was on the rise and reached an alarming level in the 16 th century, i.e. the Ottoman period (Backman, 2014: 175). The result was more Spanish-speaking merchants being present in the (western) Mediterranean, thus increasing commerce, but also an increased presence of pirates in the whole Mediterranean, especially the western and central part. This increased slavery led to a circumstance that greatly affected the Lingua Franca, which was closely, almost intrinsically, tied to piracy and astonishingly had a much greater economic effect than a political (or social) one. Not only were these Spanish merchants (linguistically) influencing the Lingua Franca through their increased numbers and language presence, resulting in more Spanish input to the Arab-speaking merchants, but also pirates were capturing and enslaving more and more Spanish merchants and bringing them to their headquarters in Algiers. This resulted in an increasing Spanish influence on both the pirates/slavers and the local Arabic population. Similar to the political development of Spanish, Italian was also quite present in the (central) Mediterranean including, of course, Italian-speaking merchants (Chaunu, 1979: 305) as well as pirates from Tunis and Tripoli capturing Italian merchants and adventurers. Thus both Italian merchants as well as Italian captives/slaves gathered in Tunis and Tripoli caused not only a shift in vocabulary, as the Italian language became quite present in the central Mediterranean, but also the stabilisation of an Italian variety of the Lingua Franca which lasted until the 19 th century.

7.1 Piracy and slavery

Basically the Lingua Franca benefitted in two ways from the changing political scene of commerce to a piracy and slavery based economy: first by merchants, predominantly Spanish and Italian, trading with Algiers and other commercial cities such as Tripoli or Tunis, and second by 106 pirates bringing captives to the same Algiers and the other pirate city strongholds of Tripoli and Tunis. This caused a shift of vocabulary in both the western and central Mediterranean, creating a Spanish and Italian variety of the Lingua Franca as well as further increasing the numbers of speakers of the Lingua Franca within the commercial cities. Not only did other merchants adapt the Lingua Franca but pirates, slaves and even parts of the local population spoke the Lingua Franca, which created a massive stability in Lingua Franca use motivated by a political agenda through piracy or, as its legalized form was called, corsairing, which had become an official state affair as corsairing was a legal form of warfare (Kahane & Tietze, 1958: 12-13). The whole piracy and slavery system was motivated politically, as the political power, the Ottomans (and before the Arabian Empire), had legalized piracy and slavery (although it was called corsairing), which had been practised not only by Ottoman pirates but also by Arab as well as European pirates. It dated back even to the Catalan era, although not in the intensity of the 16 th century when slavery had almost become large-scale commerce itself. What is different to the Catalan era is that the 16 th century differs in its growing importance of piracy and consequently slavery during the Spanish reign. Although commerce was still the major aspect, slavery, which had existed long before, but not in this dimension, had by the 16 th century become a main part of commerce itself, as slaves were seen more as goods that could be sold or held for ransom or as human resources (even though the northern Mediterranean made captives as well). This whole situation caused a massive politically motivated development in the Lingua Franca which had several different consequences for it linguistically, socially and even politically, as based in the southern Mediterranean, but being influenced by the economic (and political) dominance of different Romance languages in the northern Mediterranean. What happened in the main pirate cities was that captives were put into so-called bagnos , prisons, and were forced to work or beg during the day thus were scattered all over the city and had to learn the Lingua Franca to communicate (Selbach, 2008: 33). The local population, who bought slaves who then stayed at their houses, also used the Lingua Franca to communicate with these captives. However, most curiously, the local population only acquired the fixed Pidgin instead of the expanded Pidgin as did the pirates, and thus this excludes a Creole development in the 16 th century initiated by the local population. Pirates spoke the (fixed Pidgin) Lingua Franca to communicate with the people they captured. This emanates from reports about the Lingua Franca. A fixed Pidgin variety but never an expanded one was used by pirates, slavers and also sailors. The fixed Pidgin state Lingua Franca was used by pirates and slavers to give the slaves orders

107 and force them to work, as slaves had to either pay a ransom fee or work for about five years to be freed (Selbach, 2008: 33). Both pirates and slavers themselves presumably used the Lingua Franca to communicate among each other, especially later on when pirates and slavers consisted of both Turkish and Arab but also European crews (Haedo, 1612 quoted in Kahane & Tietze, 1958: 20). The reason for ´pirates consisting of international crews is due to the monotheistic cultures prohibiting the enslavement of their own people and, more specifically, the law prohibited Christians to buy or sell Christian slaves as was the case with Jewish and Muslim slaves, therefore, pirates consisted of different nationalities to be able to stand as the appropriate seller and be able to sidestep religious scruples and laws´ (Backman, 2014: 180-181).

7.2 Varying Lingua Franca competence

The 16 th century with pirates, slaves, diplomats and even the local population speaking the Lingua Franca shows both a fixed Pidgin and an extended Pidgin state which is very informative about the use of the Lingua Franca. Depending on the speakers, the Lingua Franca was used either for sophisticated trade and diplomacy as an expanded Pidgin (but only in the groups of merchants and diplomats), or for simple or basic communication in the fixed Pidgin state by pirates, slavers and the local population as a means of (limited) communication. Looking at the available reports, this assumption can be confirmed. Contact with pirates and slavers is very basic, impersonal and only exists in the form of simple commands consisting of where to go or what to do, followed by or, in fact, primarily consisting of, insults of almost the same length. This form of impersonal communication remained as such within slavery until the end of legal piracy itself, in the 19 th century (Selbach, 2008: 33). Due to pirates collecting a considerable number of captives and bringing them to pirate cities, these captives, or slaves, were forced into slave labour in the cities as workers or craftsmen, waterboys or beggars (Selbach, 2008: 33) and had to communicate not only with slavers but also with the local population, and thus the Lingua Franca was a necessity. However, communication only occurred in the most necessary or basic manner and neither idle interaction nor close contact between pirates and captives or between slaves and the local population seems to have existed. Among the slaves themselves ´Christian languages (and religion) were not forbidden by slavery´ (Selbach, 2008: 34), presumably because even though there may have been many different languages present among the captives, which would be a reason to speak the Lingua Franca for

108 communication purposes, there were also, due to the substantial number of captives alone, captives with the same mother tongue establishing contact among each other. Therefore, communication in the native language must have occurred when slaves communicated with each other rather than the Lingua Franca, which could explain why the slaves did not learn/use the Lingua Franca among themselves. It would be possible that slaves used the Lingua Franca for rather basic communication, presumably with speakers of other languages. One source states that speaking Romance languages among slaves was prohibited. However, this was an exception and limited to galley slaves (Marott (1677) quoted in Selbach, 2008: 54). Most unexpectedly as well as startling is that the same impersonal communication used by slavers was also reported from the local population towards the slaves. As reports point out, the local population commanded and insulted the captives in just the same way as the slavers. Consequently, it comes as no surprise that contact remained on a low personal level and the Lingua Franca itself remained in its state and never developed further. In fact, the Lingua Franca presented in the source texts, judging from the 16 th to 18 th century sources reporting about piracy, slavery and captivity but also renegades (Schuchardt, 1909 and Cifoletti, 2004 and Foltys, 1984), primarily show a fixed Pidgin Lingua Franca and hardly any extended Pidgin Lingua Franca, as reported for instance in Haedo (Haedo, 1612 quoted in Cifoletti, 2004: 197-202) or Aranda (Aranda, 1656 quoted in Cifoletti, 2004: 206-207). Exceptions exist, although the use of the expanded Pidgin Lingua Franca was reported concerning an Algerian Rai (a ruler in Arabic states), who seems to have enslaved a priest (Tamayo, 1905 quoted in Cifoletti, 2004: 209). Also a marabout (holy man) is reported keeping slaves in his house who used a very advanced form of the Lingua Franca, i.e. most definitely an expanded Pidgin state, but also rather used insults (Haedo, 1612 quoted in Cifoletti, 2004: 200-202). Further reports mention one doctor who also had a quite high Lingua Franca competence (Haedo, 1612 quoted in Cifoletti, 2004: 198), and another doctor using the expanded Pidgin state (Caronni, 1805 quoted in Cifoletti, 2004: 230). In addition there are reports of presumably a nobleman (Basci Amba) keeping slaves (Caronni, 1805 quoted in Cifolett, 2004: 230), a Pascha who, by far showed the highest Lingua Franca competence, which almost appears to be a Creole-like state (Pananti, 1817 quoted in Cifoletti, 2004: 244). Curiously, one slaver (seemingly in a higher position) used a slightly more developed form of the fixed Pidgin Lingua Franca. This, however, did not have the same level as the expanded Lingua Franca (Haedo, 1612 quoted in Cifoletti, 2004: 198). Further, judging from the available sources, the conclusion seems to be that there might have been a considerable

109 difference in Lingua Franca competence itself, with on the one hand people in higher positions and, presumably, more intensive language contact with European speakers (merchants and diplomats or other officials), and on the other between pirates and slavers including the local population.

7.3 Slavery not responsible for extended Pidgin state

When looking at most of the Lingua Franca documents, not only does a fixed Pidgin state, which is present in most documents, become visible but also an extended Pidgin state, predominantly in the later documents, but already present in Haedo (Haedo, 1612 quoted in Cifoletti, 2004: 197- 202 and Foltys, 1984: 13 and Schuchardt, 1979: 35-37). This might suggest that commerce developed the Lingua Franca into an extended Pidgin, but unfortunately, this does not seem to be the case, as when looking at documents that relate to either pirates or slaves, then only a fixed Pidgin is used. However, outside this particular group other groups such as ambassadors, doctors and people such as Haedo (Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Foltys, 1984: 13- 14 and Schuchardt, 1979: 35-37), who spent a great amount of time in the Southern Mediterranean, and travellers and adventurers always report an extended Lingua Franca. As most of these reports cover the same time span, this fact suggests that there were substantial differences between the Lingua Franca used for commerce and the Lingua Franca used for more personal and interactive purposes. This would suggest that there were two separate ´lects´ of the Lingua Franca, the basilect as well as the mesolect, but also that group-dependent and group- distinctive competences of one and the same language existed parallel to each other. With high probability this would also explain why the Lingua Franca is mentioned only in the major locations to be spoken by everyone, but everywhere else only reported to be spoken by specific groups of people, (simply) because the Lingua Franca was strictly bound to commerce. It only reached, even less involved, the local population in major cities, and even less in small ports, and only some selected few outside these locations truly had access to this language. Therefore, the development of the Lingua Franca remained unnoticed, at least by early sources. However, it is also present in later sources and due to the intensity of trade the assumption can be drawn that this development may have occurred early on. Furthermore, different Lingua Franca competence between groups but also classes seems to have existed, with merchants and state officials having a greater understanding of, or competence in, the Lingua Franca than pirates,

110 slavers or the local population itself. The fact that the local population employed a lower level of the Lingua Franca could also indicate that the local population was not the main group to use the Lingua Franca - just like pirates, slavers and sailors, who also displayed a lower Lingua Franca competence and thus appear to be a secondary group of speakers. Furthermore, this could be an indication that the local population also was not the first group to acquire the Lingua Franca. Therefore, (Arab) merchants, but also diplomats, can be considered as the primary group of Lingua Franca speakers. Judging from the Lingua Franca competence among pirates and sailors, the Lingua Franca probably did not develop on ships either, as commands on ships are always fixed. One could even speak of a special mentality on ships regarding language communication, which is minimal and (almost) military due to the short and precise commands which are necessary in navigating a vessel. There are also no reports about sailors using a more sophisticated Lingua Franca, and the Lingua Franca certainly was used on ships. Assumptions as such are difficult to prove, as the specific group of merchants is hardly mentioned in the reports and then only by reference and although there are very detailed reports about diplomats, other state officials and doctors, these reports are few in number. Thus the level of (personal interaction and) Lingua Franca competence is difficult to measure in a broader perspective. However, it definitely seems possible for the Lingua Franca to have prevailed given that large-scale commerce would allow merchants more language contact and with it the potential for a higher Lingua Franca competence, although the extended Pidgin state represented its furthest development, also due to the Lingua Franca having found its way into diplomacy.

7.3.1 Reasons for Lingua Franca downgrade

Even though the Lingua Franca clearly shows signs of being in an extended Pidgin state in the 16th century the captivity reports hardly show the use of this Lingua Franca state and instead only display the fixed Pidgin state. This seems to be due to the extended Lingua Franca being more than would be necessary for the purposes of slavery, as a fixed Pidgin would serve the purpose of insulting or commanding someone to go somewhere or do something. However, at the same time, both piracy and slavery caused the necessity to increase the number of Lingua Franca speakers, and with it its spread in the Mediterranean. Judging from the use of the lower Pidgin state, piracy, but especially slavery, did not encourage the further development of the Lingua Franca but

111 actually made the Lingua Franca less inventive concerning the means of expression (terms and phrases), i.e. it diminished (or downgraded) the (expanded) Lingua Franca (state). Through solitary imprisonment, communication and interaction became restrictive, while domestic slavery also only led to unilateral commands and insults and not to mutual personal interactions. In fact, instead of keeping the expanded Pidgin state, both piracy and slavery degraded the Lingua Franca state, as communication was held in a lower lect or on a lower level than what the extended Pidgin state would have been capable of. This is most likely the reason why the Lingua Franca not only did not develop beyond the extended Pidgin state but deteriorated instead and remained in a fixed Pidgin state, because there was no close personal relation or exchange between Lingua Franca speakers and slaves and thus no necessity for a more complex (Pidgin) language (development) to prevail.

7.4 The expanded Pidgin state

As slavery presumably was not the reason for the extended Pidgin state or further developed the Lingua Franca, but rather used the already developed and expanded Lingua Franca in its previous (fixed Pidgin) state, the expanded Pidgin state presumably must have been established before this event, i.e. before the 16 th century through commerce. In fact, the Lingua Franca could have reached the extended Pidgin state in the 14 th century at the peak of (Catalan) commerce, caused by the sunstantial volume of commerce and communication between people. Although the Lingua Franca may have reached the fixed Pidgin state previously due to the prevailing Catalan- and European-Arab language contact, communication was neither that intense nor required more than a few fixed words or phrases for commerce before the end of the 13 th century. The existence of an expanded Pidgin state of the Lingua Franca around the 14 th century could be explained through the complexity and intensity of trade (i.e. the massive involvement of people) and the constant input of Catalan, which only grew in the 13/14 th century with large-scale commerce. This would indicate a strong connection between large-scale commerce and the development of an expanded Pidgin state, which was picked up by the Arab merchants but did not lead to an adaption of the Lingua Franca by Catalan merchants, only by travelling merchants of other nationalities, i.e. other languages. The (obvious) conclusion that can be drawn here is that there was a development of the Lingua Franca into an expanded Pidgin although hardly was there any personal interaction among Catalan and Arabic speakers. Presumably there were

112 exceptions on the personal level but not on a public level, probably because particular religious (and political) attitudes prevailed among the people. Instead, this development was initiated by Arabic speakers because of the high input of Catalan causing the Lingua Franca to develop into an expanded Pidgin. Cross-checking these assumptions with the case of the Austrian-Rumanian Pidgin it can be assumed that sporadic or individual language contact alone does not create a Pidgin at all, but the learning of some words. More intense and group-oriented language contact leads to a specific routine repeating only specific content nouns and verbs which, of course, establishes a certain Pidgin state but it, curiously, always remains very stable at the specific point of the fixed Pidgin state (as seen with the Austrian Rumanian Pidgin). Large scale communication (for instance through large-scale commerce), on the other hand, almost immediately creates an extended Pidgin, which is quickly reached but never develops into a Creole state without individual variation. Concerning merchants, a lower Pidgin state would have hindered communication and thus deterioration would have been counter-productive. However, there also was no need for anything beyond the extended Pidgin state in commerce, as close personal relations would have been rare, and most definitely not a mass occurrence. Trade could function perfectly with the expanded Pidgin. In the case of the Turkish-Arab diplomats, deducing from the reports and source texts, this group never expressed or showed a desire to speak any (European) language at a higher competence than necessary. These diplomats would actually use the Lingua Franca because it was capable of expressing everything necessary for diplomatic relations, but at the same time they regarded European languages as being beyond their dignity and never intended to have more contact than necessary with them. Such attitudes serve as a plausible explanation why the Lingua Franca remained as the extended Pidgin until the 19 th century, because merchants and diplomats never required or wished and, therefore, never initiated a further development of the Lingua Franca. In addition there was no encouragement from piracy, slavery or the local (Arab) population to even remain with the extended Pidgin state (in fact rather a discouraging), much less for any further development.

7.5 The Creole(-esque) state

The expanded Pidgin state seems to have sufficed even for diplomacy, and although high Lingua

113 Franca competence among the diplomatic group can not be denied from the sources, this Creole(- esque) state never appears or is even mentioned before the 19 th century. Therefore, it can be assumed that the Creole-like state of the Lingua Franca reported in a source text of the 19 th century (Cifoletti, 2004: 244) cannot have already been initiated in the 16 th century, as the Lingua Franca does not seem to have been used in the expanded Pidgin state by the majority of Lingua Franca speakers (the local population, slavers, pirates etc.). Furthermore, if the creolization process had started in the 16 th century the Creole state itself would have been reached at least by the 17 th century, as the development process of a Pidgin does not take long, a few decades at most. However, such a process needs to be initiated and only once the social need or political- economic necessity becomes imminent will it start. However, as it definitely was the local population which initiated the Creole development of the Lingua Franca, which becomes visible from the available source texts of the 19 th century (Renaudot, 1830 quoted in Cifoletti, 2004: 249 and Ferrari, 1912 quoted in Cifoletti, 2004: 246 and Calligaris, 1834 quoted in Cifoletti, 2004: 258), this process must have been a very late development, starting not earlier than the 19 th century. The reason why the Lingua Franca developed into a Creole-like state was because it had become a lineament of its users themselves, which simply was not the case in the 16 th century, where not development but rather deterioration occurred. Only in the 19 th century did the Lingua Franca became popular enough among the local population for its speakers to use it among themselves in personal interactions, in fact in all situations, presumably due to the close contact established within European culture, which consequently initiated the development process into a Creole. The only report about Lingua Franca competence beyond the extended Pidgin state is from the 20th century when the Lingua Franca, or rather a Lingua Franca(-esque) Pidgin, seems to have reached the Creole state, spoken in Tunis by a young Turk (Nitram, 1931 quoted in Cifoletti, 2004: 288-292). From the source text it can (only) be assumed that there were more people with the same Lingua Franca competence as the person who was also described as a typical young Turk prototype. Unfortunately, this being the only report, how many people possessed the same language competence is unclear and debatable, as it could have been only the beginning of a youth trend of acquiring a high Pidgin language competence. It could also have been very common among young people to have such high (“European”) language competence. However, it would seem unlikely that a Creole competence should be an exception and acquired only by one person given the great popularity of European culture at that time, which presumably also

114 benefitted the use of Romance Pidgins in general, especially among (young) Turkish and Arabic speakers in the later centuries. What must be added here is that, unfortunately, it is debatable if this Creole state really is the Lingua Franca itself or another French Pidgin highly influenced by the Lingua Franca, as many forms do not resemble the Lingua Franca and appear to be French, although similarities to the Lingua Franca do exist. In any case, by the 19 th century, the Lingua Franca had developed considerably and had left the extended Pidgin state and was in the process of creolization, i.e. in a Creole(-esque) state. It was thus close to becoming a fully fledged language based on Romance vocabulary and Arabic grammar, with whole complex sentences including ´punctuation´ (subordinate clauses), tenses, prepositions, i.e. a (fully) developed grammar that showed enormous regularities and advanced complexity (Pananti, 1817 quoted in Cifoletti, 2004: 244).

7.5.1 Reasons for the Creole and Creole(-esque) state development

It must be added that this development was originated from the Arabic speaker side that had developed and spoken an almost Creole-like Lingua Franca but which almost suddenly appeared in the beginning of the 19 th century. The reason for this development was that the eastern Mediterranean (Levant) opened up its culture towards western culture, presumably resulting in acquiring western languages, which also affected the Lingua Franca in developing further into a Creole. Unfortunately, only a few paragraphs of what the Lingua Franca looked like in this state have survived, but these fragments clearly show how far advanced the Lingua Franca was at that time. If this development had continued the Lingua Franca would definitively have become a Creole. In fact, the Lingua Franca was in the process of creolization when these few examples were written down and it seems as if the creolization process was in progress and almost finished. If the Lingua Franca had not become outdated, or rather been replaced, and thus vanished, it would even have developed further into a full language (within a short time). Due to there being no Creole(-esque) Lingua Franca before the 19 th century, there definitely would have been reports about such a Lingua Franca state, and slavery, but also piracy, as well as commerce can be excluded as the cause of this development. What led to the Creole(-esque) development appears to be diplomacy. The only group, however, which truly initiated a Creole development, although not with the Lingua Franca itself but a similar Lingua Franca(-esque) French Pidgin, was the local population

115 themselves. Due to the surprisingly low opinion about slaves and their consequent prevailing mistreatment among the local population of the 16 th century, such a (Creole) development can only have occurred in later centuries, presumably the 19/20 th century, as no source text beforehand even mentions a Creole state of the Lingua Franca or a similar Pidgin in the Mediterranean region. Therefore, the reason for the development of a (French) Creole was due to a more positive trend or at least an opening up towards the European culture on the local Arab population side which initiated this development and not the impact of commerce or diplomacy, and especially not the system of piracy or slavery. The only aspect that did change because of slavery (and out of necessity) was the expansion of the Lingua Franca vocabulary, the occurrence of doublets (Selbach, 2008: 45-46), i.e. more than one Lingua Franca word for the same term, which was an adaptation to the multi-nationality of the captives.

7.6 Vocabulary expansion through multi-language input (through piracy and slavery)

Another important fact is that although piracy and slavery caused a change in the Lingua Franca, in fact a deterioration of the Pidgin from the expanded to fixed Pidgin state, at the same time an extension of vocabulary occurred. This happened via the occasional replacing of the most common and most similar (Romance) words (including grammatical words), leading to a huge increase in intelligibility as speakers of many different European languages became involved in piracy and slavery at that time, both as captives as well as captors. At the end of the 16 th century Latin domination of the Mediterranean became a European domination through the joining of the French, Dutch and British into Mediterranean commerce, resulting in the increased presence of several different European nationalities and languages in the early 17 th century (Greene, 2014: 94, 97). Consequently, much (international) vocabulary entered the Lingua Franca, but intelligibility was maintained, leading to an increase in vocabulary choice without change of the basic vocabulary. An increased number of different words occurred with one and the same meaning. When more and more Europeans with different languages such as French, English, Dutch, Greek, etc. became present in the Mediterranean during the 16/17 th century (Haedo quoted in Kahane & Tietze, 1958: 20) they were captured by pirates, became slaves in the major pirate cities and sometimes they even deserted to become so-called renegados (renegades), serving among the Ottomans mostly as pirates (Kahane & Tietze, 1958: 20). Due to such a number of different

116 (European) languages entering into contact with the Lingua Franca and contributing to it, not only did many different language speakers acquire the Lingua Franca but they also influenced the Lingua Franca owing to the languages involved. In any case, many different languages entered the Mediterranean area in use, thus consequently influencing the Lingua Franca, and especially the pirate strongholds of Algiers, Tunis and Tripoli became most “international” (due to the various nationalities present). For the Lingua Franca this meant a considerable change in its pronunciation and vocabulary, but in a rather unexpected manner. Instead of adapting (grammatically) further to the many new languages which seem possible for the Lingua Franca, it (presumably) simply became open to new vocabulary, thus extending but not structurally altering it in the process. Many different pronunciations of this vocabulary, occurred causing various different spellings in the source texts. Depending on nationalities and the mother tongue of a speaker, the Lingua Franca became open to more or less any (“spelling” and) pronunciation (Cifoletti, 2004: 33-34) without changing its grammatical core at all or much of its basic vocabulary 16 . Of course, minor changes in previously fixed vocabulary and a general extension of vocabulary occurred, but not to a truly considerable or expected degree. Previous vocabulary was maintained, and only some very common words and a few loanwords were added. This reaction to increasing linguistic influence is successful as it avoids the rise of many diverse Lingua Franca varieties as well as any major complications in establishing which is the correct pronunciation (or spelling, if the Lingua Franca had any) or the correct word, as simply any general Romance word could serve the purpose of conveying a particular concept.

7.7 Variations in pronunciation

The best example is the word bono which exists in almost any possible form regarding European spellings and pronunciations, as is demonstrated in Selbach (2008) in which various spellings of bono occurring in the source texts are listed i.e. bon , bono , bonou , bouno , bouonou and bueno (Selbach, 2008: 46). Another good example would be mucho i.e. mucho , mouchu , mouchou , mouchuos , mu-mucho (Cifoletti, 2004: 32, 55, 84, 95), and the reason why this is a perfect example is due to the argument that the pronunciation is based upon the mother tongue of a

16 This can be perceived if examples from all known sources gathered in Schuchardt, 1909; Foltys, 1984; Lang, 1992; Cifoletti, 2004 and Selbach, 2008 are considered 117 speaker and not on the correct pronunciation of the Lingua Franca. However, as all of these reports about the Lingua Franca from the southern Mediterranean in the 17/18 th century present the Lingua Franca as being pronounced differently, i.e. as spelled by a speaker’s own spelling rules, it would not seem far fetched to assume that there was no generally agreed pronunciation. It was up to each individual speaker to pronounce the Lingua Franca as they saw fit, at least on the side of Romance speakers. Furthermore, this probably was also the case on the Arabic speaker side. According to Cifoletti (2004) ´every dialect of the Maghreb does have its very own system and during the reign of the Ottomans there was only a stronger tendency towards particularism´ (2004: 34). Thus many Arabic speakers had their distinctive way of pronunciation which was often represented in the Lingua Franca sources, and it may be assumed that there were further differences between Turkish and Berber Lingua Franca speakers. What is also suggested through multi-pronunciation is that there was a great variety of speakers that used the Lingua Franca, correlating to the multi-nationality of speakers present in the Mediterranean at that time. It can therefore be taken (and it is even reported) that the Lingua Franca, presumably in any variety, was understood by every other speaker of the Lingua Franca (Cifoletti, 2004: 32). As intelligibility was still guaranteed this leads to the plausible conclusion that the Lingua Franca itself did not change (as such) in the 17 th century, and definitely not in its grammar. Instead extended its vocabulary as well as its pronunciation, which allowed speakers of all nationalities to pronounce the Lingua Franca according to their own language system and still be intelligible to other speakers, which also increased the number of Lingua Franca speakers (as well as its stability itself).

7.8 Positive aspects of slavery

The only positive aspect that can be mentioned about piracy and slavery is that both might have been the reason why the Lingua Franca had become known in the whole Mediterranean during the 16 th century by people not involved in commerce or diplomacy and caused the spread in several Mediterranean cities as well as the vocabulary expansion. Piracy and slavery might even have caused the Lingua Franca to be known and spoken until the 19 th century, as piracy had made use of the Lingua Franca and slavery forced it to be spoken and even resulted in the local population of cities in the whole Mediterranean basin coming into contact with this language. Once the Lingua Franca was widespread, people would continue to use it even outside its main

118 field of communication. Furthermore, piracy and slavery continued onwards until the 18/19 th century (Selbach, 2008: 33) which could also be a reason why the Lingua Franca did not change its innermost core structure, only its vocabulary (and some grammatical forms). The whole (political/social) process stayed the same or at least similar during this time, as piracy was similarly conducted as commerce and commerce in the 16 th century in any case consisted to a great extent of slavery.

7.9 Major political changes and final development

Unfortunately, three major political changes that greatly affected the Lingua Franca (in a negative way that led to its extinction) occurred within the last one and a half centuries of its existence. As referred to above, one such political change was the French conquest of Algiers in 1830 which led to French colonisation which had tremendous consequences for the Lingua Franca. Not only did this cause a shift in vocabulary resulting in a French Lingua Franca variety (Sabir), although according to Faidherbe (1884) ´the Lingua Franca was simply called Sabir since the conquest of Algiers´ (Faidherbe, 1884 quoted in Foltys, 1984: 31). However, the influence of the French language continued and was immense and relexified the Lingua Franca and then replaced the French variety of the Lingua Franca with a French-Arabic Pidgin (Petit/Grand Mauresque), which had (probably) developed out of Sabir itself. This occurred all within a relatively short amount of time (i.e. a few decades). The French influence was quite different compared to all other previous influences - both the Spanish and Italian influence as well as the (unusual) multi-lingual influence of the 17 th century - as French appears to have had the strongest influence among all other ones because it caused both a shift in the Lingua Franca vocabulary followed by a (almost immediate) replacement of the Lingua Franca with a French Pidgin. Curiously enough, a French Lingua Franca variety evolved and the adaption of commerce to French occupation changing led to the shifting the Lingua Franca towards French. However, due to colonisation, this did not lead to the adaption of the French Lingua Franca variety, although such a (similar) process had always occurred centuries earlier. Its replacement was with a French Pidgin called Petit Mauresque, which was the adaption of the local population to French being the official language of Algiers. Several reasons seem to be the case for the replacement of the Lingua Franca: one is that French became the official (political) language and was being spoken in Algiers since the conquest and

119 not the Lingua Franca, which was spoken in Algiers before by the local merchants and presumably the local population. The Dictionnaire (1830) itself even states that the Lingua Franca was spoken by everyone in Algiers and Tunis ( Dictionnaire De La Langue Franque Ou Petit Mauresque , 1830 quoted in Cifoletti, 2004: 89-90), at least at that time (19 th century), otherwise there would not have been a shift at all to a French Lingua Franca variety. Comparing the source text of Haedo (1612) and the Dictionnaire (1830), the impression could easily be reached that the Lingua Franca had always been strongly and, most importantly, continuously present in Algiers. Furthermore, one could further conclude that due to the success of the Dictionnaire , which gained a 2 nd (printed) edition, a need for it existed, presumably because the Lingua Franca was spoken by a majority of the population. However, as the French occupiers did not know the Lingua Franca other than from the Dictionnaire and definitely did not speak either the Lingua Franca or, presumably, the French variety, as otherwise the French Lingua Franca variety would have prevailed, the Lingua Franca soon became less used after the French occupation. It was weakened by the circumstance that French and not the Lingua Franca was politically preferred and thus spoken, but also by the existing lack of sympathy towards the French Lingua Franca variety. One quite substantial explanation for the emergence of a new French-Arabic Pidgin is given by Schuchardt (1979), who states that “knowledge of Arabic increased among the French, while knowledge of French increased among the Arabs (and Berbers)”, which led to the result that “the scope of Lingua Franca was increasingly narrowed” (Schuchardt, 1979: 43). The only approach to the Lingua Franca by the occupiers was by the compiling of Lingua Franca terms and phrases in the Dictionnaire , which was undertaken only to gain a better understanding of the area to be conquered (Cifoletti, 2004: 83). Otherwise there seemed to be relatively little contact with the Lingua Franca itself (in Algiers). Furthermore, the Dictionnaire was only an attempt to generally prepare occupying soldiers for the situation of two different mother tongues being present to be able to establish basic communication with the local population. Algiers was conquered by soldiers and settled by civilians, not by merchants, and therefore, it would have been unlikely that many soldiers would have known the Lingua Franca before or used it after the colonisation of Algiers. The settlers that arrived later, i.e. civilians, would also not have known the Lingua Franca other than from the Dictionnaire or strove to adapt to it, and would rather have used French or the French Pidgin to communicate with the local population (as French became the official language). The Dictionnaire was the only point of reference to the Lingua Franca for

120 many French speakers, and presumably used only to introduce French soldiers, later settlers, to the Lingua Franca before it became replaced by a French Pidgin. Another reason for the replacement of the Lingua Franca was that, although popular in Algiers, not many French people acquired the Lingua Franca but instead used French as a means of communication and only used the Dictionnaire as a short guide for communication. Due to French being the official language and also spoken primarily, the extent of spoken Lingua Franca decreased considerably as the local population would adapt to French and abandon the Lingua Franca. Over time fewer people would know the Lingua Franca and in the meantime the French Pidgin would become more popular by 1852, which is the first appearance of the designation Sabir (Mac-Carthy and Varnier, 1852 quoted in Foltys, 1984: 30). The name itself, however, is said to have been in use since 1830, i.e. since the conquest of Algiers (Faidherbe, 1884 quoted in Foltys, 1984: 31). The Lingua Franca in Algiers was far less in use than a century previously and almost became extinct, even though in the beginning of French-Arabic language contact the population speaking the Lingua Franca adapted to this new language situation, causing a vocabulary shift towards French. Parallel to this development, the Lingua Franca was also used in its Creole(-esque) state reaching a point (of development) unsurpassed before. Some Lingua Franca speakers, Turkish and Arabic diplomats, had developed the Lingua Franca further, curiously without French influence, and most curiously this Creole(-esque) Lingua Franca state developed in the same location, i.e. Algiers (Pananti, 1817 quoted in Cifoletti, 2004: 244), when European political dominance was hardly felt or was not there at all. Unfortunately, this development did not prevail from a Mediterranean perspective as both the Creole(-esque) and French Lingua Franca became extinct. The second major political change and also the end of the Lingua Franca as a Fixposten , i.e. an established entity, was the end of legal piracy through the French conquest leading to “the final extinction of the danger from Barbary corsairs” (Abulafia, 2014: 147). This was one of the main reasons for the Lingua Franca to (still) exist until the 19 th century, as piracy as well as slavery were the main promoters of the Lingua Franca in the sense that pirates and slavers primarily and exclusively used the Lingua Franca as a means of communication, ensuring that Lingua Franca competence was steady and available at all times and also (forcefully) introduced European captives to the Lingua Franca itself. Although not every single captive would adopt the Lingua Franca, the forced introduction to the Lingua Franca by piracy and slavery left a mark on the whole Mediterranean and affected many individuals, and definitely made the Lingua Franca

121 known, or presumably infamous, at both the Northern and Southern Mediterranean shore. However, once piracy and especially slavery ended officially in the 19 th century (Selbach, 2008: 33), people were not captured, enslaved and locked up in such massive numbers in the piracy strongholds of Algiers, Tripoli and Tunis any more. Less of the population, both European captives as well as the local population (that were keeping slaves), came into (forced) contact with the Lingua Franca and therefore the number of Lingua Franca speakers steadily decreased. With the main (life)source, i.e the established entity gone, if not to say, the main reason for its existence, at least among the local population from the 16 th to 19 th century, for the Lingua Franca, the number of speakers was reduced among the European but especially among the Arab population. The Lingua Franca could have continued onwards and survived in its original merchant group due to the language purpose this Pidgin language always had - being a trade language spoken by (Arab) merchants - if not yet another major political change had prevented the Lingua Franca from surviving. The third and final major political change that affected or actually ended the Lingua Franca presence as the Mediterranean lingua franca was the replacement of the official language of trade with French and (later) English. As has been shown, the Lingua Franca had been weakened before, in the 19 th century, when French became the official language of trade and was, therefore, not adapted but replaced by another French Pidgin. It was exactly due to French (later English) replacing the Lingua Franca in its function as a (Mediterranean) lingua franca that the Lingua Franca lost all its purpose for existing, even for the merchants themselves. Not much was left of language competence in the Lingua Franca (language competence) by the end of the 19 th century and this trend continued onwards into the 20 th century with English becoming a global lingua franca as well as the international language of commerce (transport and science), thus also the official language of trade for the Mediterranean. This made it impossible for the weakened and already declining Lingua Franca to develop further or to continue its existence, causing it to finally vanish from the Mediterranean basin and become extinct. The Lingua Franca had finally become irrelevant for the Mediterranean in the 20 th century.

7.10 Political and economic connection

Although political and economic developments lie close together and seem to have shaped the Lingua Franca, commerce and economic relations seem to predominate rather. However, both

122 commerce (including piracy and slavery) and politics (including diplomacy) were independently performing social systems that had their own dynamics, which in the aspect of functioning existed separately but not totally isolated from each other, as one was constantly influencing the other. However, both commerce and politics existed independently of the Lingua Franca, which just happened to be a very small part in a much more complex communication system. Nevertheless, the Lingua Franca did become an important integral part of both the economic system as well as the political system, i.e. in trade and diplomacy, which is not to say that not any other language could have taken its place. In fact, any other Pidgin or language would have been suitable, maybe even more suitable than the Lingua Franca itself. However, the circumstances were beneficial for the Lingua Franca and thus it was used (without being fully noticed linguistically). In the history of the Mediterranean, there were several economic systems that existed independently from each other as well as functioning apart spatially but were still connected through Mediterranean commerce. Among the first European commercial systems since the re- establishment of commerce was the Crown of Aragon, Italy, which had always been present in the region, although primarily in the Black Sea (Kahane & Tietze, 1958: 9), and the Spanish Crown, which followed a century later. In the later centuries, from the 18 th century onwards (Greene, 2014: 84), France became dominant in commerce. In the Southern Mediterranean the Ottomans developed an extensive commercial system and interacted with the commercial systems of the Northern Mediterranean, i.e. with Europe, as had the Arabian Empire before. These political systems were spatially separated but still corresponded commercially and thus interacted with each other. Politically this meant that Algiers was dominated by Catalonia, then by Spain and finally by France. Tunis was dominated by the Catalan language followed by Italian, and Tripoli was dominated by Catalan then the Ottoman (and the Arab) influence dominated. However, the (rest of the) Southern Mediterranean was controlled by the Arabian Empire which later became replaced by the Ottomans. Most importantly, every single one of these political systems employed the Lingua Franca, and depending on the prevailing political influence, the Lingua Franca adapted its vocabulary to fit the individual politico-economic system. The only exception concerning the utilisation of the Lingua Franca was the French politico-economic system, which first used the Lingua Franca, then Frenchifed it and finally replaced the Lingua Franca through a French Pidgin. Concerning Lingua Franca competence, it has to be noted that although the Lingua Franca was in use, only

123 the population of the Southern Mediterranean really spoke it. Furthermore, only one system overtly allowed the political use of the Lingua Franca, or at least did not seem to prevent it from being used, i.e. the Ottoman political system, which does not come as a surprise, as the Ottoman Empire has always been very liberal and open towards (different) languages (Selbach, 2008: 34). Wherever commercial (and, to some extent cultural) contact between these systems occurred, the Lingua Franca was part of it. Consequently, the important point to note is that the Lingua Franca existed, survived and was sustained because of the (close) contact and interaction between these systems, which also provided close language contact and interaction between speakers of different mother tongues, thus facilitating the further necessity of a common means of communication. After all, the Lingua Franca was (presumably) created through the exchange of two of such politico-economic systems, (possibly) the Catalan and Arabian one, and existed for as long as exchange and trade between these two systems was upheld due to a mutual need for communication. Through the exchange of commercial systems, language contact could be established and maintained. This even allowed further growth in both the commercial as well as language contact sense until it had reached such a high point that the common means of communication could spread even outside its natural language borders, as the whole Mediterranean became involved in mutual commerce (and thus became aware of the Lingua Franca). Of course, there were locations more involved in this commercial system exchange than others and these locations seem to be located for the most part in the Southern Mediterranean, i.e. Algiers, Tunis, Tripoli as well as Morocco, Malta, Crete and Egypt with the (few) northern Mediterranean cities being Barcelona, Marseilles and the Italian city states. The Lingua Franca, however, was only locally settled in very few locations, i.e. Algiers, Tripoli and Tunis in the Southern Mediterranean, but seems to be known in the Northern Mediterranean, especially in Italian cities. However, from the documents one can only deduce that this was the case in the 18 th century and nowhere to the same extent as in the Southern Mediterranean, because the Lingua Franca does not appear to be spoken by the local populations (Cifoletti, 2000: 14). Presumably, the Lingua Franca was spoken in other cities as well. However, depending on the volume of commerce and consequently language contact, its use was either limited to the groups involved with commerce or to a majority of the (Arab) local population. During the course of history, the commercial systems changed due to political changes, i.e.

124 through the decline of the Crown of Aragon, the rise of the Spanish Crown and the political importance of Italy, which created new commercial systems. These replaced the old commercial system of the Crown of Aragon, but not the Arabic commercial system, which was replaced by the commercial system of the Ottomans in the 16 th century. Technically, the Lingua Franca should have been replaced since the 16 th century with the Spanish and Ottoman acquisition. However, Arabic remained to be spoken in the North African region and thus one very important source language input for the Lingua Franca continued to be used even during the Ottoman reign. Concerning the replaced source language, Catalan, even after the takeover of the Crown of Aragon, it remained to be spoken until the 16 th century, thus providing much stability for the Lingua Franca itself.

7.11 Politico-economic aspects and unique linguistic traits of relexification

Even though the commercial systems were replaced during the course of the 15 th to 16 th century, the languages responsible for the origin of the Lingua Franca remained the same (for quite a long time). Thus during the 15 th to 16 th century the Lingua Franca had almost a century to adapt to the increasing Spanish and Italian influence, which is more than enough time for a Pidgin to adapt to the new language situation and to relexify, and consequently be adapted into the new commercial system. Furthermore, it seems logical for the Lingua Franca to be adapted into another commercial system considering that the language situation and language contact did not change rapidly but only slowly and thus the Lingua Franca had the possibility to be established, resulting in a new Lingua Franca variety. In fact, it does make a lot of sense not to replace but to relexify a Pidgin that operates smoothly throughout the exchange of two political and economic systems, even if one system becomes (slowly) replaced by another. The reason therefore is that instead of replacing a whole Pidgin language a simple relexification process has the same effect but with less duration, allowing commerce to flow continuously and without disruption. What truly supports the whole process of vocabulary shift, i.e. relexification, is the very fact that the relexification process takes considerably less time than the creation and especially the development of a new Pidgin suitable for the same purpose, as only vocabulary has to be replaced in a grammatically established language. The advantage of a vocabulary shift is that it is a fluid process that takes place during the use of

125 an already established Pidgin, which allowed commerce to proceed in the same intensity as before. Otherwise, through the creation and development of a new Pidgin, commerce would have suffered a disruption which is not desired. The criterial point (a comparison with the Austrian- Rumanian Pidgin would be futile because it never left the fixed Pidgin state) is the development of the Frenchified Lingua Franca called Sabir into a French Creole, which was also called Sabir. The relexification into Sabir was completed within 20 years, but the development into a French Creole (also called Sabir) was established within 30-40 years (Foltys, 1984: 32-33), clearly showing that relexification is by far a less complex and thus more easily accomplished process. This, of course, is only possible with Pidgin languages, as Creoles lose the ability to easily change their vocabulary. This might be a reasonable explanation why commerce only allowed the Lingua Franca to develop into an extended Pidgin state but not into a Creole, as the exchange between commercial systems requires both stability in expression (i.e. terms and phrases for (trans-)actions) but also flexibility in (the forms or means of) expression to establish a common means of communication with all involved groups (even if multiple languages are present). This can not have been provided by any fully fledged language, as such languages are both too time- consuming to learn and to express coherently and without ambiguity as well as too inflexible in the expressiveness of their simple grammar to communicate effectively with speakers of different languages. This is because if the grammar of a fully fledged language is simplified it can not convey enough meaning (in a mutual language situation) to both language speakers and establish mutual communication (at least not over a prolonged amount of time or a more complex context). Easy acquisition and vocabulary flexibility can only be provided by a Pidgin language and its unique linguistic traits which seem to fit the commercial needs perfectly.

7.11.1 Economic facilitation of Lingua Franca use

The theory about the Lingua Franca being a product of the interconnectedness of politically influenced commercial systems might (just) be the most plausible explanation about why the Lingua Franca survived and relexified. In fact, it might even explain the reason for its origin and spread with the exchange between commercial systems being the cause, and both the creation and stability of the Lingua Franca being the effect (provided through language contact between Catalan and Arabic speakers with language contact rather being an intermediate step). This might be a daring theory, although it does explain the curiosities of the Lingua Franca history and

126 matches the known facts. Presumably, the Lingua Franca must be seen from a more general or economic point of view. After all, it seems to have originated in and from this economic environment and thus it seems logical that its development would rather be determined (and thus be stabilized, developed and maintained) by macro (economic) factors, i.e. large-scale commerce, but also influenced by micro (economic) factors, i.e. groups such as merchants and the local population. Therefore, relexification might be seen as the integration (process) into a new commercial system and the final demise of the Lingua Franca as a rejection of integration. The creation itself of such a Pidgin can be seen either as the inevitable secondary product of such an exchange but also as the mutual requirement of finding a common means of communication between two commercial systems.

7.12 Political apprehensions

Curiously enough, the Lingua Franca was in existence throughout centuries of conflict, from political warfare to religious crusades and corsairing (Backman, 2014 and Valérian, 2014: 77). This constant state of warfare and conflict in the Mediterranean led to a fascinating contradiction between a politically motivated belief about the Lingua Franca, from considering it as “bastardized” or “broken” or disregarding it completely on the European side, to the denying of its status as a language and the denial of its very existence, i.e. by Hyde Clark (quoted in Lang, 1992: 64), and the actual use or lived reality of the Lingua Franca itself (on the Arabic speaker side). Primarily visible on the European speaker side, this ´lingual racism´, however, existed on the Arabic speaker side as well as some diplomats were not willing to speak a Romance language, which they considered as inferior to their own language, but used ´dragomen i.e. translators who, consequently, were fluent in Lingua Franca´ (Lang, 1992: 74). Rehbinder (1798) also reports about the dey , i.e. another term for pasha or ruler, definitely speaking the Lingua Franca but using dragomen. However, he adds that it is the aspect concerning speaking to Christians which made him use dragomen, and it is not explicitly stated that it is because he refused to speak the Lingua Franca. But, the Lingua Franca being associated with Christians, there is a possibility of the dey also refusing to speak the Lingua Franca due to racist or (perhaps) religious beliefs (Rehbinder, 1798 quoted in Cifoletti, 2004: 226). Schuchardt (1979) on this matter states that the dey definitely “understood and spoke the Lingua Franca, but considered it beneath his dignity to use it with free Christians” (1979: 39) and even while

127 downplaying the racist aspect, it leaves this quote still ambiguous. Most curiously, European ambassadors were fluent in Lingua Franca, as seen with Chevallier d´Arvieux, and as conversations often occurred through dragomen, translators must have spoken the Lingua Franca as well (Lang, 1992: 74). Therefore, the conclusion can be reached that the Lingua Franca was employed in diplomacy or had “at least accessory diplomatic use” (Lang, 1992: 74). Considering that European ambassadors spoke the Lingua Franca, it could also be a possibility that while Turkish diplomats refused to speak European languages but did speak the Lingua Franca in dialogue with European diplomats themselves. This seems to be the case in some reports, i.e. Renaudot (1830) in which the dey does use the Lingua Franca (Renaudot, 1830 quoted in Cifoletti, 2004: 251) and there are many other instances of consuls (Ferrari, 1916 quoted in Cifoletti, 2004: 246), rais (Tamayo, 1905 quoted in Cifoletti, 2004: 209) and pashas (Pananti, 1817 quoted in Cifoletti, 2004: 244) using the Lingua Franca (but not a European language). One further speculative assumption would be that it could also be the other way around with Europeans refusing to speak Arabic or Turkish due to their racist beliefs considering their tongue to be the superior one. On the other hand, Arabic speakers such as merchants, sailors and the local population would readily use the Lingua Franca without hesitation. With some exceptions, the Lingua Franca appears to have been used extensively and it could be because the Lingua Franca was close to the European languages and, consequently, European culture and therefore some diplomats and officials refused to speak the Lingua Franca. It could also be that the Lingua Franca was not used by diplomats because it would be the language used either by the common people, not the aristocracy, or the language associated with pirates and slaves. One very interesting point is made by Cifoletti (2004) about the use of the Lingua Franca in diplomacy, which is that in later centuries Turkish/Arab ambassadors were rather soldiers than aristocrats due to a period of demise. Thus many ambassadors would speak the Lingua Franca but not contribute to its further development (Cifoletti, 2004: 48). This attitude as such does not fully explain the Lingua Franca as a language choice. It could, however, be explained with Schuchardt’s observation about the choice of the infinitive in the Lingua Franca (Schuchardt, 1979: 28), since it was the Europeans themselves who, due to political opinions, were rejecting the Lingua Franca as an impure form of their own language and thus refusing to speak it. This could be the reason why (Arab and Turkish) diplomats chose to speak the Lingua Franca because of the European’s (own) contempt towards it. The choice of the Lingua Franca by Arab and Turkish diplomats was thus due to a socio-politically motivated

128 opinion of the Europeans about the Lingua Franca. The actual (lived) reality of the use of the Lingua Franca itself was completely different. The Lingua Franca was a useful, popular language spoken among the people that came into close contact with each other, i.e. merchants, sailors, locals etc. - a situation which rather facilitated than restricted language contact. Therefore, the Lingua Franca probably served as a means of communication but not as a cultural or political obligation, because Arabic speakers rather wanted to communicate and trade in a language that was understood by many Romance speakers. Also many Romance speakers were able to communicate with Arabic speakers precisely because of them speaking the Lingua Franca, allowing the conclusion that the Lingua Franca was very present in the Arabic speaking parts of the Southern Mediterranean. The political opinion created of the Lingua Franca was one of disdain, yet it did not diminish its usefulness, or stop its use and notoriety. The only aspect that was affected was the attitude of people who had little to no contact with the Lingua Franca (directly) and were, therefore, condemning it and writing articles about it being an abomination or a farce with respect to pure languages. However, this was a political opinion formed out of ignorance and an obscure idea about (im)purity in languages which in no way represented the reality of the Lingua Franca itself. If there had not been such (negative) consensus about the Lingua Franca, there probably would have been much more Lingua Franca presence in the European cities and the Lingua Franca could have gained the same popularity in Europe, and maybe even acquired local European speakers, as it did with Arabic, Berber and Turkish speakers in the Southern Mediterranean basin. One interesting example is Italy, specifically Venice in the 18 th century, as it demonstrates how (political) beliefs and reality collided and in what form this occurred. The Lingua Franca was heard everywhere in the ports via Arab Lingua Franca speakers and seems to have become enormously popular or rather notorious in that it was even known, although not spoken, by the local Italian population despite an opposing attitude towards it (Cifoletti, 2000: 14). The astonishing aspect about this is that the Lingua Franca was present during the high point of (political) hostility against Pidgin languages, yet the Lingua Franca was still present in a European city (at least the port), which definitely points to the considerable nominal recognition and spread of the Lingua Franca at that time. As has been indicated, politically, the Lingua Franca seems to have been perceived quite differently depending on the Northern or Southern Mediterranean shore, while Europeans, for the most part, had a strong aversion against the Lingua Franca. Ambassadors and state officials of

129 Algiers, Tunis and Tripoli are reported to have used the Lingua Franca instead of the European languages, although they definitely spoke both, to communicate with European ambassadors. The Lingua Franca may have gained more distribution in the Northern Mediterranean but not recognition, as it was met with hostility by the Europeans. Groups that had a political motivation to use the Lingua Franca include Arab, Turkish and European ambassadors, poets or dramaturgs who used the Lingua Franca to ridicule other nations via the conception of other languages being underdeveloped. This suggests that politically motivated linguists wanted to prove the supremacy of European languages. Schuchardt (1909) in particular was exceptional in seeing the Lingua Franca as a language of its own, and even though his work was composed after the peak of the linguistic intolerance of previous centuries, it was still written at a time in which many racist perceptions prevailed, making his work not only ground-breaking and taboo-breaking but also allowed insights which might not have been possible to reach (only) a century later without his essay.

7.13 The Lingua Franca development and politics

The Lingua Franca in itself had no political motivation, as it simply was (used as) a means of communication, at least in its original Catalan form, which however might have changed ever since piracy used it intensively. This also decreased the reputation and status the Lingua Franca might have had, at least with Europeans. It is, however, fascinating that the Lingua Franca was used as a diplomatic language. The reason given by the known documents state that since many ambassadors considered it beneath their dignity to use a European language, therefore, the Lingua Franca was chosen as a substitute, as has been indicated above. A (most) fascinating aspect is that while the Lingua Franca was used by a considerable number of persons involved with piracy and slavery (in some form or other) the Lingua Franca still managed to become a language spoken by diplomats, which required a higher (mesolect) level of competence than the Lingua Franca had acquired ever since the rise of piracy. According to documents, the Lingua Franca seems to have evolved into a Creole(-esque) state primarily spoken by diplomats and high state officials on the Arabic/Turkish side in the 18 th (to 19 th) century; i.e. cf. “ Guarda per ti, et non andar mirar mugeros de los Moros, nous autros pillar molto phantaia de quaesto conto ” Rehbinder (1798 quoted in Cifoletti, 2004: 227), “ Oh si vi starei pur a meraviglia, perché io tener gia donne assai, tener belli figliuoli, danaro molto, e

130 molta riputazione in Tunisi ” Caronni (1805 quoted in Cifoletti, 2004: 232), “ ma chi salire al potere, non star piu schiavo: vedere anzi schiavi al suo piede; servire a uno per comandare a mille: ti star buona cavezza ” Pananti (1817 quoted in Cifoletti, 2004: 244) and “ Ti aver comandate per mi de fasir guardia a la porta de fortizza ” Calligaris (1929 quoted in Cifoletti, 2004: 258). The Lingua Franca seems to have evolved substantially compared to previous reports and sources, and thus it seems logical that it developed further through political motivation, i.e. the use of diplomacy (and not through piracy). After all, diplomatic conversations, even when reduced to their essence require a much greater vocabulary, number of topics and expressions than a simple order-insult monologue by slavers or pirates. Thus the Lingua Franca used in the 16-17 th century was degraded by piracy. If diplomacy (i.e. political reasons) truly was the cause for the Lingua Franca development into a Creole(-esque) state or not cannot be answered conclusively, although the documents would strongly suggest this. Presumably, through being used in diplomacy the Lingua Franca also gained more prestige among the local population and thus more and more people became involved in learning and speaking this language causing it to evolve even further. Curiously, the massive spread among the local population seems to have been the initiation point for diplomacy to adapt the Lingua Franca, to then further develop it and pass the developed form back to the local population. In fact, politics instead of economic relations may have been the determining input for the Lingua Franca Creole(-esque) development. The course of a Pidgin becoming a diplomatic language must have been (not only unheard of before, but also been) one of the most rare and curious cases to occur in Mediterranean language history. The political facilitation of Lingua Franca development does not seem to have increased the numbers of Lingua Franca speakers, as commerce did, but instead caused a linguistic development of the language. Thus one might speak of a quality improvement by political factors and a quantitative improvement by economic factors. The economic development of the Lingua Franca seems to have been stable and continuous but stopped at, or maybe was even limited to, a very specific point in Pidgin language development, i.e. the extended Pidgin state which presumably lasted from the earliest beginnings, documented (only) from the 16 th to the 19 th century. It is difficult to argue whether the Catalan Lingua Franca was at a fixed or an extended or even a further state of development before the 16 th century (if it existed at all), as hardly any documents about that time have prevailed (Zerbitana, 1353 quoted in Collier, 1977: 295).

131 Presumably the Lingua Franca must have reached the extended Pidgin state at least at a time as the intensity of trade was at its peak with the Catalan Empire and certainly at the same level as with Spain, later in the 16 th century. Furthermore, presumably pirates simply adapted the Lingua Franca and most likely did not change it themselves, but only adapted it to their admittedly less complex communication needs. Thus piracy and slavery only seem to have brought about a more internationally shaped vocabulary.

7.13.1 Economic and political interdependencies

It was economic reasons which defined the existence, stability, use and consistency of the Lingua Franca, i.e. commerce was a main motivation for quantitative language contact. In the meantime, political influences initiated the further development and the rising popularity or prominence of the Lingua Franca and were thus the main motivation for qualitative language contact. As a result, quantitative development seems to be limited (or bound) to the economic aspect of languages and qualitative development, at least in the case of the Lingua Franca, seems to be bound to politics. Political motivation can have effects such as the (qualitative) change of Pidgin states, linguistic improvement and further development, but also the complete opposite, as seen on the European side, with opposition to qualitative use. A Pidgin language is degraded through political rejection of non-Europeans speaking a Romance language, with the result of a Pidgin perceived as an “inferior” and “mixed” (European) language. As a result, Lingua Franca interaction decreased and even its existence was completely ignored with the same effect of decreasing the quality of the Pidgin.

There are a few rare cases in the course of the centuries when political factors directly or indirectly affected the (perception of the) Lingua Franca itself and these instances had a qualitative influence on it. One such case was diplomacy facilitating the Lingua Franca (development); another case was the political attitude of aversion against impure forms of European languages (by Europeans) among which the Lingua Franca was counted, which caused a rejection of speaking (or even attending to) the Lingua Franca. Thus no further development was possible on the European side, but instead the development of the Lingua Franca occurred on the Arabic speaker side, although or maybe even because there also was an aversion against other non-European national languages by Arabs and Turks. If this is true, then economic factors may

132 support the Lingua Franca but limit its development to a very specific point, while political factors would allow the Lingua Franca to develop further but can also limit or deny its development due to an aversion and/or rejection of the very same language. The 19 th century shows the Lingua Franca developing beyond the extended Pidgin state which occurred when it became politically accepted in North Africa and the Southern Mediterranean. The Lingua Franca was spoken before by the population but it never evolved. Only after the politically motivated development occurred can a steady increase of the local population being involved in speaking an evolved Lingua Franca be noticed, i.e. by the Dictionnaire (1830 quoted in Cifoletti, 2004: 89-192). The creolization process took considerably longer than for the Lingua Franca to reach the extended Pidgin state through economic exchange, because this process started in the late 18 th to early 19 th century and only showed results later on in the 19 th century. At this point it must be mentioned that it can only be deduced from the documentation collected in Schuchardt (1909), Cifoletti (2004) and Foltys (1984) that the results of politics, but especially diplomacy, became visible in the 19 th century. It would also be possible that this whole process started much earlier and maybe even did not take as long as the documents suggest; i.e. a steady increase in Lingua Franca competence over the course of the centuries from a solid expanded Pidgin state in the 16 th century as reported by Haedo 17 (1612 quoted in Cifoletti, 2004: 199), to a prevailing decrease into a fixed Pidgin state until the 19 th century mentioned in the sources as in Dan18 (1637 quoted in Cifoletti, 2004: 205), de Breves (1628 quoted in Cifoletti, 2004: 202) and Poiret (1789 quoted in Cifoletti, 2004: 221). This continued with an almost sudden reappearance of the expanded Pidgin stage in Rehbinder 19 (1798 quoted in Cifoletti, 2004: 225) and then to a very advanced expanded Pidgin to an almost Creole(-esque) state in Pananti 20 (1817 quoted in Cifoletti, 2004: 244) and, finally, a (non-Lingua Franca) Creole state in the 20 th century as in Nitram 21 (1931 quoted in Cifoletti, 2004: 288). Judging from the sources alone, the development from the expanded Pidgin to Creole(-esque) state took more than a century. While it seems plausible enough to assume that evolving a Pidgin from an extended state to a Creole state takes longer than from a fixed to an extended state and

17 “Si cane dezir dole cabeca, tener febre no poder trabajar, mi saber como curar, a Fe de Dio abrusar vivo, trabajar, no parlar wue estar malato” 18 “No pillar fantasia; Dios grande, mundo cosi, cosi, si venira ventura ira á casa tua: ...cane, perro, Iudaeo traditor” 19 “si e vero que star inferno, securo papasoso de vos autros non poter chappar de venir d´entro” 20 “tutti dipender dai principi, dai piu forti, dalle circostanze, dall necessitá; tutti stare schiavi degli usi, delle convenienze, delle passioni, delle malattie, della morte” 21 “Ji si pas borquoi vos en avi bisoan ti casse ton coumbasse bor trovi l´moyan di crier li p´tites z´histoires qui ti racontes, à mon blace, dans l´boublic tounisian dipouis plous di vangt ans” 133 requires much more time than creating a Pidgin or an ad-hoc Pidgin, the exact time necessary for such a development can not be determined in detail. This process could have lasted a century but could also have been limited to a decade or two, judging from the development from the Lingua Franca to Sabir which took from 1830 to 1852 (Foltys, 1984: 30). Presumably then, the development of a Creole(-esque) state was a longer development, although diplomats would not have taken on an insufficient undeveloped Pidgin (from piracy or slavery), but rather an already extended Pidgin and adapted it to their needs, which definitely accelerated this process (of development). However, it probably would still have taken much time. The transition from diplomacy to take effect on the population may have been a further cause for this temporal delay of development and its visibility, as the Lingua Franca might have been developed through diplomacy but only have been adapted by the population afterwards once the Lingua Franca had reached its almost Creole-like state. Within another century it developed into a Creole. As there was a strong cultural tendency towards European culture in the 20 th century, consequently there was a strong European presence in the Mediterranean and much language contact with European speakers. Therefore it can be assumed that the Lingua Franca would definitely be influenced by European languages and thus had the opportunity to develop into and maintain its state as a Creole, i.e. as an independent and distinct language. Most curiously, in fact, a slight state of decreolisation or depidginisation can be noticed in Lingua Franca competence in the later centuries as noted by Nitram (1931 quoted in Cifoletti, 2004: 288-292), in which a convergence with French becomes clearly noticeable. The reports from the 20 th century quoted in Cifoletti (2004: 287-292) and what examples are given of the Lingua Franca, could be interpreted to support the claim of it being the result of a decreolisation or depidginisation process. Unfortunately, the examples recorded are not written by linguists and thus are linguistically inaccurate and can consequently be interpreted as evidencing both, a pidginised French language or an imprecise description of a decreolisation or depidginisation process of the Lingua Franca itself. However, the examples themselves do not truly prove the decreolisation of the Lingua Franca, as some examples can be described rather as a French Creole. However, there is a slim chance that this could be due to the imperfect (written) reproduction of the relevant author and the Lingua Franca being in a Creole state would, at least, fit into the general tendency of uniqueness of this Pidgin language as well as the history of the Lingua Franca development itself. After all, there are examples of the Lingua Franca having reached not only the Creole(-esque)

134 state, but at least one example shows the progress of depidginisation (Nitram, 1917 quoted in Cifoletti, 2004: 288-292) in the 20 th century. This suggests that the local population itself acquired a Pidgin, although it is unlikely to be the Creole(-esque) Lingua Franca state, and developed it further into a Creole. However, even though a very intense relexification process occurred due to the strong political influence of French on the Lingua Franca, it is most implausible that this Creole truly has a connection to the Lingua Franca itself as, linguistically, hardly any Lingua Franca trait can be analysed.

7.14 Political-economic historic analysis

The close link between political power and the economy creates a strong politically controlled economic influence on language itself, i.e. whichever political power is economically dominant is also culturally and especially linguistically dominant. Consequently, to know the dominant languages in the Mediterranean and their influence and development over the course of history also means to know the Lingua Franca. An analysis of the dominant and thus influential language can lead to the Lingua Franca origin itself as well as its development and spread, because the language that was dominant and therefore spoken in the Mediterranean was also the language that created and/or influenced the Lingua Franca in its linguistic course. After the Roman Empire collapsed in the 5/6 th century, the Arab Empire thrived (Valérian, 2014: 77-78). Meanwhile the European powers were scattered and constantly involved in conquest and warfare within (and outside) their own borders and from the 7/8 th to 14/16 th century numerous kingdoms were defending and reconquering the Iberian peninsula (through the Spanish Mark), among them the Catalan and Occitan kingdoms (Chaytor, 1933). These kingdoms were constantly expanding and constantly in a state of war. However, only certain of them had access to the Mediterranean, i.e. the Catalan and the Occitan kingdom (both of them became united at some point), and were expanding their influence into the Mediterranean (Valérian, 2014: 84). By contrast, the Basque kingdoms, however, expanded towards the north and the northern Iberian coast (Chaunu, 1979: 305) and there were the Italian city states on the southern European shore, although these were not involved in war and rather expanded towards the Black Sea (Kahane & Tietze, 1958: 9). Geographically, only the Catalan, Occitan and Italian kingdoms or city states had direct access to the Mediterranean basin. The Occitan kingdom, for instance, did have a large port, Marseilles,

135 but the Occitan kingdom as such was not a thalassocracy. Italy continued to be active in commerce, but its interest shifted more towards the east, especially the Black Sea but also into the Aegean. Only the Catalan kingdom became strong in the western and central Mediterranean. It was also among the first to venture to the southern Mediterranean shore, which is shown through the expansion within the Mediterranean with Catalan kings occupying Malta, Sicily, Sardinia and later even parts of Greece, and colonising the Mediterranean with great success. However, the newly conquered territories were basically just a means to fund more wars (Abulafia, 2014: 145, 146-147). Through the investment of Catalan kings in commerce, even though the primary concern was territorial expansion, first efforts to establish large scale trade were nonetheless made. Thus commerce flourished and contact as well as communication with Arabic speakers was facilitated and (slowly) increased. Curiously enough, most Catalan kings before James I, but also many after him, were not interested in seafaring and maritime ventures or even maritime law. Only James I and a few after him were active in seafaring. However, they all were always involved in conflicts, wars and expeditions. More importantly, none of the kings themselves had any interest in commerce itself and left this very profitable business in the hands of merchants. Thus a very wealthy and influential merchants class developed, which had lodgings all over the Mediterranean, the so-called merchants´ quarters. Nonetheless, the Catalan kingdom became a vast empire called the Crown of Aragon, and Catalan itself became the official language of the empire and consequently the language used in the whole (Western and Central) Mediterranean basin, as it became the (official) language of commerce and (unofficial) language used on ships (in fact quite similar to English today being used for commerce but also on ships, trains, planes, i.e. all means of international transportation). Catalan even became popular among people of different mother tongues, i.e. in colonies such as Sicily or Sardinia, which seems quite plausible as Catalan was widely accepted in the Mediterranean, but also due to Catalan settlers successfully replacing the local population (Abulafia, 2014: 146). For commerce to reach such a large spread and intensity by the 14 th century, within a relatively short period of time in such a very wide area, long-term investment on a large scale was unavoidable. One must keep in mind that in the Middle Ages time was an essential factor, as any progress and any development was very slow. To simply connect the Mediterranean through a commercial network in only a few centuries was an impressive achievement (also for Barcelona

136 to become such a great and influential port). Quite a frequency of trade was needed, which could not have been accomplished by (individual) commerce led by the public (merchants) alone, i.e. backup was required. Although the (Catalan) kings invested in commerce early on, commerce remained at a comparatively low level even in the 12 th century. The commerce practiced by the Italian city states, Genoa, Venice and Pisa being the most active, was far from the dimensions that Mediterranean commerce would reach in the 14 th century or had reached during the Roman Empire. The most likely historic event that initiated this frequency of large-scale commerce was the (political) conquest of Malta (1127) and Sicily (1282), which created a setup that facilitated the economic dominance of the Crown of Aragon and consequently increased the flow of commerce as well as the spread of the Catalan language, which led to the creation and increase of language contact. Through these strategic locations, merchants could venture to all major cities in the Mediterranean in a relatively short period of time as well as return from a larger venture into the Mediterranean and restock their supplies. Such a central location controlled by the Catalan kingdom allowed commerce to flourish. Although Catalan commercial and political dominance (itself) lasted from the 13 th until the 15 th century, it also ended abruptly as, via marriage, the Crown of Aragon became Spanish, and with it the influence of Catalan faded until it was replaced by Spanish in the 16 th century (Chaunu, 1979: 62). At this point, it is important to stress that this process took about a century until Spanish had fully replaced Catalan, which means that Catalan was strongly present even until the 16 th century. The most likely reason was the fact that people outside the Crown of Aragon still used Catalan until the Spanish presence became overwhelming and it was more useful to use Spanish. About the same time that Catalan faded, Italy had reached such a prominent status in the Mediterranean (Chaunu, 1979: 305) that Italian became the most aspired to language in its (central) area. At the same time the Ottoman Empire became the driving force in the eastern and southern Mediterranean since the conquest of Constantinople in 1453 (Valérian, 2014: 77). Meanwhile the presence of Italian could be felt very clearly, while Turkish itself as a language was not strongly present in North Africa (Kahane & Tietze, 1958: 20) and therefore also not in the Lingua Franca itself. This can be explained by the curious circumstance that the Turkish Empire built their sea empire and their success on recruiting and enslaving Europeans, i.e. so- called renegades and slaves or captives, to upgrade their technological and maritime knowledge and be able to compete with European powers (Kahane & Tietze, 1958: 15). Consequently, Romance languages entered the Ottoman Empire and, furthermore, most people in the Southern

137 Mediterranean rather spoke Arabic than Turkish. Turkish itself was reserved for the higher positions and present in the eastern Mediterranean (Kahane & Tietze, 1958: 20), diminishing the influence of Italian in that part and limiting it to the central Mediterranean. Meanwhile Spanish dominated the western Mediterranean during the 16 th century. The final chapter in the considerable political and economic changes in the Mediterranean (which are relevant for the Lingua Franca) occurred in the 19 th century when many political realities were removed there with a shift of power towards French through the occupation of Algiers in 1830. As also noted above, the conquest of Algiers by the French and the shift of the language of commerce to French also caused a shift of the dominant language to French and thus this gained massive importance as a language (even) to such an extent that it actually replaced the most common (second) languages in Algiers. Arguably this was rather a political reason (i.e. territorial expansion) than a purely economic one. However, language contact and influence in the end depend on political decisions, especially if one such decision results in French becoming the language of commerce as well as introducing French into the new colony and not tolerating any other language. Concerning the Lingua Franca itself, as long as the influence or dominance of Catalan was present, the Lingua Franca remained Catalan-based. Starting in the 15 th century but only succeeding in the 16 th century, the Lingua Franca shifted to Spanish, where especially the western part of the southern Mediterranean became heavily influenced by Spain. Additionally Italy gained both political and economic dominance in the central Mediterranean and thus the Lingua Franca in the central Mediterranean became heavily influenced by Italian. The east, always being strongly influenced by Arabic, became annexed by the Ottomans and thus Turkish influence started to become present in these eastern parts of the Mediterranean in the 16 th century, which resulted in Turkish vocabulary being taken in by the Lingua Franca. However, as Arabic remained the language spoken by most of the population, the Lingua Franca remained very close to its previous form. As has been noted above, in the 19 th century, through the occupation of Algiers, French influence on the Lingua Franca increased and extinguished the Spanish Lingua Franca variety. In fact, this seems to have been a particularly strong influence, due to it being a culturally dominated colonialization process (Abulafia, 2014:147). It can be assumed that the linguistic support of the local population, as usual, first showed a tendency towards the already established Pidgin but was then politically influenced to shift towards French. Furthermore, the Lingua Franca was a

138 language that originated from large-scale commerce, or to be more precise, the exchange between the European commercial system and the Arab/Turkish commercial system, and as long as this situation was upheld the Lingua Franca could continue to exist. Finally, the end of the 19 th century also marked the final end of a commercial era and with it the inevitable end of the Lingua Franca itself. The Lingua Franca had become to be replaced by a French Pidgin, which then developed into a Creole in the 20 th century.

7.15 Conclusion

These are the most important political changes in the Mediterranean that also had an economic impact and were (thus) relevant for the Lingua Franca development. Of course, there were several other changes (as well as inventions in maritime technology) which were important for political reasons, such as political agendas, religious movements, wars or alliances and commercial treaties, etc. However, not all of these changes were relevant for, caused changes in, or affected the development of the Lingua Franca. After all, the Lingua Franca was a Pidgin, where minimal to more general changes could easily be reversed, if influence either did not last long or reverted back to a previous state through, for example, the expiration of a previously accepted peace treaty, old commercial networks or routes being strengthened again or completely discontinued. The most interesting aspect is that the development of the Lingua Franca is closely tied to the political and economic development(s) of the Mediterranean; beginning with the Catalan expansion across the Mediterranean along with the Catalan language becoming the official language as well as accepted, even adopted, in many parts of the Mediterranean, to the rising of Spanish influence from the 15 th century onward with a big shift of vocabulary noticeable in the Lingua Franca in the 16 th century. To this is further added then the Italian influence, although present in Mediterranean commerce since the 6/8th th century (Chaunu, 1979), which relexified and further shaped the Lingua Franca grammatically, especially in Tunis and Tripoli. Subsequently there was the strong presence of piracy in Algiers, which gave the Lingua Franca the most unique but at the same time internationally intelligible shape, leading to the creoliziation of the Lingua Franca into a Creole(-esque) state. It was the French occupation that had the biggest influence on the Lingua Franca linguistically, as it both shaped the Lingua Franca into a French variety and then also replaced the Frenchified Lingua Franca with a French Pidgin, which

139 even underwent a state of decreolization in the 20 th century. Finally then the introduction of English as the international trade language was a development that ultimately led to the subsequent demise of the Lingua Franca and replaced the trade languages used in the Mediterranean and with it even the last scattered remains of the Lingua Franca. By following the development of Catalan and its spread in the Mediterranean, the origin, development and spread of the Lingua Franca itself can be reconstructed in considerable detail, just as the consecutive development of the Lingua Franca over the course of history until the 20 th century can be retraced. Judging from the close political and economic connection, the constant economic exchange in the Mediterranean, facilitated (or hindered) through political decisions, was responsible for the regular need and use of a Pidgin language which made communication between people that had no common language possible. As long as this established form of (commercial) contact was kept alive the Lingua Franca could continue to exist until the very end of this type of commerce in the 19 th century. All in all, the Lingua Franca had a long history (of existing) and survived many changes. However, with the large-scale political and economic change of the 19 th century it finally vanished and may only exist sporadically in separate locations in the (Southern) Mediterranean, if at all.

140 Chapter 8: The development of the Lingua Franca The starting point for the development that led to the origin of the Lingua Franca is definitely the establishment of Latin-Arabic contact. All started with (the historic event of) the re-establishment of commerce, when Latin, i.e. Italian, Catalan and Occitan speakers came into close contact with Arabic speakers. This is unlikely to have happened (in any situation) on the Iberian mainland as the kingdoms of the Hispanic mark were constantly in a state of war with the Arabian Empire and Arabic speakers in Iberia itself were not only oppressed with few to no rights but also a minority. Language contact between Latin and Arabic speakers on the Iberian Peninsula was scarce and Arabic speakers had to learn a Latin language, and in the end most of them were even expelled. Furthermore, there was no trade, communication or any other real exchange between Arabic and Latin speakers in any form that could have facilitated even a Pidgin on the Iberian Peninsula itself (Chaytor, 1933). The only possibility for Latin-Arabic contact to arise under such sparse language contact circumstances was the re-establishment of trade between Christians and Muslims, in this case between Catalan and Arabic speakers. This was achieved through the opening up of Muslim territory for Christians (Abulafia, 2014: 183 and Chaytor, 1933), which is the only historical event (the Reconquista surely did not facilitate communication) that allowed contact between a broad(er) public of Catalan and Arabic people, i.e. merchants. This historic event dates to 1050 when commerce was re-established after almost 300 years of seclusion, although “it would be imprudent to overestimate the extent of these exchanges, trade and navigation never completely ceased” (Valérian, 2014: 80). Before this event, commerce would have rather been conducted by Jewish traders than by Christians, as commerce of the 10 th century was at a very low level due to the restriction of trade and proscription of contact between Muslims and Christians. This is also the reason why “Jews dominated long-distance Mediterranean commerce in the West during the early medieval economic depression when the volume of trade was much reduced from the previous highs of the late Roman Empire” (Pirenne, 1939 quoted in Astren, 2014: 399). As soon as commerce and the North African shores were opened up for trading, Latin merchants, among them Catalan merchants (but also Italian and Occitan), ventured to various ports in the south-western Mediterranean and started to trade with the local merchants (unlikely the population itself). This created language contact, i.e. a language situation in which speakers of two different languages, Catalan and Arabic, had to find a means of communication. Presumably, the local merchants adapted to the new (Latin/Catalan) vocabulary, i.e. words became adapted by 141 Arabic speakers. However, due to the irregularity of encounters, a low number of merchants coming to North Africa and low intensity of contact, no Pidgin was or could have been created (only ad-hoc Pidgins) and this remained unchanged for quite some time. The same is true a century later when commerce had slightly increased and reached a small scale level and thus marginally increased the intensity of language contact, although still not enough to create a fixed Pidgin. However, there must have been some form of development, as some new words and possibly phrases must have been acquired (concluding from the Austrian-Rumanian Pidgin), which is a common development process during (the beginning of) continuous language contact. During the course of the 12 th century, through a slow but steady rise of commerce in the Mediterranean, merchants interacted more with each other and thus more language situations occurred. Language contact became less irregular, however, and there were still seasonal difficulties which interrupted contact between Latin and Arabic speakers. Although this irregularity would not allow more than the acquisition of words and phrases that were helpful for communication, it must have created a (Romance) language competence among merchants (which established some form of fixed communication). However, the typical Pidgin feature of blending two languages did not occur as there appears to have been no visible grammatical adjustments which would allow the assumption that this was a Pidgin. However, this commercially caused language competence definitely provided the basis for the later development of a (fixed) Pidgin.

8.1 Practical effects of language contact

Switching from the theoretical aspect to the practical aspect, what happened in terms of language contact was that vocabulary acquisition and language competence development occurred but not the creation of a (fixed) Pidgin. The reason for this assumption of mere vocabulary acquisition is that commerce (or rather trade, i.e. single transactions) does not need a large number of words or expressions to be successful. Even numerals and the basic vocabulary for the available trade goods can suffice to conduct business. However, this is not very effective for more complex commercial demands, which are bound to fail because one simply cannot express everything with certainty to make complex transactions clear. For instance, contracts that require various shipments over a certain extended period of time including the responsibility for the cargo can unlikely be expressed via a limited vocabulary. The more complex a contract or transaction

142 becomes the less appropriately a few acquired words can suffice for the situation. Therefore, (these) single and simple transactions (of directly buying or selling wares) might not have required a Pidgin. Small scale trade also only required a few more words, maybe even phrases, as some wares or transactions were only available or possible if some messages could be expressed clearly and without ambiguity. As long as people did comprehend the meaning of some basic (trans)actions and commercial vocabulary, small scale commerce could flourish via the further acquisition of vocabulary. However, more complex transactions that operated on a large scale required a much more sophisticated language to express more complex concepts, i.e. business transactions that needed to be negotiated, agreed upon and realized by both parties. Therefore, depending on the intensity of commerce, a more complex means of communication eventually became inevitable. This could only be provided by a language capable of conveying meaning at least on a basilectal level, e.g. a Pidgin. Concluding, the transition to large-scale commerce certainly could have facilitated the Pidgin origin process.

8.1.1 Catalan historical comparison

Cross-checking historical events and the development of the Crown of Aragon confirm this thesis (of practical effects). Until 1127, i.e. the annexing of Malta, only very few Catalan merchants ventured to North Africa, most likely Fez, Amalfia and then Algiers 22 . Consequently, communication between Catalan and Arab merchants existed but was far too inconsequent and infrequent (generally on too low a level) to facilitate the creation of a (fixed) Pidgin. Within the next century, commerce became a little more established and thus communication intensified until a point was reached when contact was (annually) regular but still very infrequent due to its dependency on seasonal winds, the disruption through climatic peculiarities and political occurrences. Communication between Catalan and Arabic speakers could only have been established to a point where the learning of a few words would occur. However, the Catalan language competence would increase on the Arabic speaker side through the further intensification of commercial relations leading onto a better understanding of basic Catalan. Presumably, Catalan could also have became pidginized by Arabic (or established an ad-hoc Arabic Pidgin), but a stable (fixed) Pidgin did not arise.

22 This is a conclusion drawn from various sources: Gertwagen, 2014: 159, Chaunu, 1979: 68, Catlos, 2014: 368 and Wansbrough, 1996 143 An increase in Catalan language competence (and dominance) is most likely to be found in the course of the 12 th to 13 th century, but was a long process that only became noticeable when commerce started to intensify and definitely culminated in the 14 th century when commerce had reached a large scale in the Crown of Aragon. One could argue that it was a very fluent transition. The volume of commerce only reached a suitable level for a sustaining (Catalan) language competence during the late 13 th to 14 th century and it rather seems to have provided the basic conditions leading to the development of the Lingua Franca than the date of Lingua Franca origin itself.

8.2 The impact of frequent language contact

Although trade existed before the 13/14 th century, its frequency was rather low and language contact occurred on a very inconsistent basis and thus the probability of the development of any stable means of communication would be very low as it was easier to accommodate oneself in every individual encounter than establish a fixed language. After all, to create a Pidgin a certain extent of frequency and with it language contact is necessary. Otherwise, when only a few speakers are present, neither the need nor the attempt to learn or create a new language exists. Furthermore, the extent or frequency of language contact necessary to create a Pidgin was unlikely to be reached in the early stages of the re-establishment. In the case of the Lingua Franca, a certain number of people and a certain degree of language contact made the creation of this Pidgin necessary, which is unlikely to have happened before the 14th century. The number of merchants venturing to the southern Mediterranean was simply too low and most Romance languages were neither fully developed before the 12 th nor established before the 14 th century. A (fixed) Pidgin could not have originated under such circumstances, and most definitely not the Lingua Franca. The only set-up in this time and place (i.e. the 14 th century western Mediterranean) which would have allowed the origin of a (fixed) Pidgin was the peak of Catalan political and economic dominance. In fact, the existence of any fixed Pidgin before this peak in the 14 th century is very unlikely, and the existence of a Romance-Arabic Pidgin from 1050 to the 13 th century is even less plausible and before 1050 it is almost impossible as the Arabian Empire did not allow Christians into its territory, i.e. its ports and cities. The only form of contact (with Christian Europe) that was allowed was trade with Jewish merchants. However, this group would rather trade with their own

144 people or would rather use Arabic or Judeo-Arabic to communicate with Arabic-speaking merchants (Catlos, 2014). Consequently, in this situation Jewish merchants could not have created the Lingua Franca before the 11 th century, although it might have been a different case with Iberian Jewish merchants from the 16 th century, when the expulsion of the Jewish population from the Iberian Pensinsua was completed, but commercial relations with Romance speakers still occurred. The most curious aspect about Pidgin creation is that for any Pidgin to arise, the necessary amount of language contact is of the essence as a (deep) level of interaction is needed for the origin of a Pidgin. Only hearing or listening to a language does not (in most cases) lead to the acquisition of a language or even a Pidgin language: only repeated interaction and active participation can facilitate the process of creating a Pidgin as well as learning a (Pidgin) language. This process is to be found in active participation, linguistic involvement and regularity of (language) contact, i.e. the constant repetition of interacting with speakers of a different mother tongue in a language situation requiring all participants to be active and having to find solutions for communication problems. Only then, with repeated effort involving a high level of participation and interaction, can a Pidgin originate. The Austrian-Rumanian Pidgin, for instance, only originated once a considerable number of Rumanian speakers frequently visited the gas station, which created both the necessity and basis for a Pidgin. Beforehand, when only very few speakers of Rumanian or other languages appeared, language contact (even in the Pidgin sense) was not even taken into consideration: learning a new language or even words from it did not come to mind at all because pointing and showing served the purpose (of communicating) and a shrug of the shoulders could solve any communication problem (i.e. not my problem). Furthermore, as long as there were co-workers that had knowledge about the needed language (Rumanian, Italian, Spanish etc.) that could be asked, communication was (considered as) established. The worst case scenario was a communication failure. The point this example should demonstrate is that as long as there were measurable numbers of customers and a low frequency of language contact creating (language) situations that could be handled via signs or be ignored altogether, there existed neither the necessity nor even the willingness or attempt to learn new words, not to mention the formation of a Pidgin. This assumingly represents the first 50 to 100 years since the re-establishment of trade, where no vocabulary adaptation or the adaption of grammatical words occurred. The first encounters presumably rather consisted of sign language and the repetition of simple

145 words which, as in the Austrian-Rumanian Pidgin case, led to the curious situation that when making a transaction, i.e. paying for a product, both parties would state and repeat the required sum in their own language (A: einsachzig, R: unu optzeci , A: einsachzig , R: unu optzeci etc., with the occasional pointing to the display of the cash register by both parties to presumably remove doubt). No new words would be adapted, much less pidginized, as each speaker would remain in their own mother tongue. This must have occurred at the very beginning of (language) contact in the Mediterranean between 1050 and 1100 to 1150 (as a rough estimate), where such encounters were rare (Chaunu, 1979: 57, Gertwagen, 2014: 159 and Wansbrough, 1996: 139) and never demanded interactive language contact or created the necessity of communicating with a speaker in another language. Arguably, however, after some time of increasing frequency of language contact an ad-hoc Pidgin must have been created due to an increasing number of interactions as communicating with Romance speakers became inevitable. This is not a unique situation because any circumstance that creates sporadic contact between speakers of different languages should also have similar results, especially in the (limited) spectrum of Pidgin (and Creole) creation. Limited contact does not lead to the creation of a Pidgin (intensive language contact does): it rather created language situations that were solved differently via sign language or the repetition of words while each speaker would use their own language.

8.2.1 Attempt at an explanation

An (speculative) explanation would be that communication among merchants (between 1050 and 1300) consisted of people learning single words and phrases of another language without lingually approaching this other language. Presumanbly, it rather was the acquired requisite of vocabulary of the most common items, and actions and eventually phrases emerged. This formed an elementary language basis for restricted communication (in the sociolinguistic sense sense of topics and expressions) among speakers of different mother tongues. The acquisition of basic vocabulary represented the smallest common denominator of (the beginning of) successful communication which, however, established itself further over the course of the 12-13 th century, leading to an increased (commercially dominant) language competence and (after quite some time) a language basis was established for the later development of a Pidgin. Catalan and Arabic speakers came into contact with each other and this contact caused words and

146 phrases to be learned, which seems unavoidable and probably happened regardless if a Pidgin had existed or not. At first, different language speakers came together and through a certain regularity, or at least the desire to continue communication (and trade relations), merchants, presumably mostly on the Arabic-speaking side, learned words and phrases from the Catalan language. Presumably, after this event, language competence increased parallel to commerce and then remained on the same level until commerce further increased.

8.3 Pidgin development based on frequency

Although Pidgins can develop quite easily, language contact or rather the intensity of language contact is essential and determines the initiation of this development. Only once a certain amount or frequency of language contact is reached will a Pidgin originate and develop. Considering the first decades after the re-establishment of trade, contact within these decades must have been sporadic and presumably showed little results because language contact could not have reached a large enough scale for any development within language competence. Commerce hardly increased after the re-establishment of trade until a point of low but stable language contact was reached during the 12 th century. Therefore, any (language) development during the 12 th century can not have been more than the acquisition of some words (and presumably resulted in not even that before this time). However, it can be assumed that contact remained sporadic, but at least did not vanish completely and (Catalan) language competence must have developed, although slowly and on a very basic level. Language contact definitely existed, and even if contact remained sporadic and was reduced to rather learning single words than truly establishing close and interactive language contact, this kind of contact must have changed, or rather intensified, at some point in time due to the (rather late) increase of commerce and its beginning development into large-scale commerce. Presumably, the continuation and increase of commercial contact during the 12-13 th century (Chaunu, 1979: 57 and Wansbrough, 1996: 139) must have resulted in increased language contact, i.e. (Arabic-speaking) merchants learning an increasing number of words (and perhaps phrases) to be able to trade more effectively, and through the sheer continuity of this (form of) contact, Catalan language competence must have developed. This would also be an important basis for the development of any Pidgin (and could probably be interpreted as a social approximation as well).

147 Language contact, alongside with commerce among Catalan and Arabic speakers seems to have increased in the 13 th century and due to the “repetition of action”, i.e. repeated trade encounters, a certain kind of language routine must have occurred and phrases must have been acquired, considering that language contact existed for over a whole century. Furthermore, the increase in economic contact since the 13 th century was an initiation point for the further development of the Catalan language competence of Arabic speakers. A further increase in Catalan proficiency must have occurred, although this might not have resulted in the formation of a Pidgin. Consequently, language contact and Catalan language competence must have developed until the necessary conditions were met to form a Pidgin, which would speculatively be possible from mid 14 th century or later.

8.3.1 Austrian-Rumanian Pidgin comparison

Further parallels can be drawn to the Austrian-Rumanian Pidgin where a certain pattern of development directly proportional to the amount of trade can be observed. As long as language contact was sporadic there never existed the necessity to learn even a few words of Rumanian as the truck and bus drivers used words that (most) Austrian speakers would understand or had heard before in this business. Consequently, there never arose a situation in which an active participation or any form of interaction with a different language was needed, thus no Rumanian words were learned. Language contact with Rumanian in itself was very rare as almost all truck and bus drivers knew a little bit of a Romance language, mostly Italian or French but also a little English or German, most likely picked up on their tours through Europe. Most importantly, all knew how to order coffee which basically was the main consumption at this gas station. However, due to economic reasons more truck and bus drivers changed their trade routes and went through Austria, thus the number of Rumanian speakers slightly increased to what could be considered as a very low but constant language contact. The result was that some useful Rumanian words were learned. This situation of vocabulary acquisition existed with sporadic contact and as long as contact remained haphazard (but regular over an extended period of time, i.e. 20 encounters a week or about 2-3 short encounters a day for a year), only a few Rumanian as well as a few Austrian words were used. These few words conveyed the required meaning necessary for trading and even if not everyone picked it up immediately, their meaning was soon realised after a short reflection (a few seconds) about it. The most important vocabulary included

148 basic words or concepts like Rumanian numerals, products that were most frequently consumed such as coffee, milk and tea ( cafea, lapte, ceai ), different fuel types ( diesel , benzinâ ) etc., (no alcohol was consumed). Curiously, the English expression to go /tugo/ for “takeaway” was understood by almost all Rumanian speakers as well as the names for several credit cards and services which were used by truck and bus drivers ( TDS -Trucker Diesel Service, Go-Box , Vignette etc.). These expressions were the first to be both adapted and used due to them referring to items which were frequently consumed and constantly needed. This form of language contact changed once Rumanian speakers were arriving at this location in large numbers. The economic background is that Spain hired field-workers from Rumania who were transported via buses through Austria and this gas station had made a contract with the transport company. Consequently, buses would make a stop at this very gas station when going in any direction between Rumania and Spain. The advantages were cheaper gas for the transport company and more customers for the gas station. Unlike the bus drivers, who all either knew a Romance language or English, the hired workers from Rumania knew Slavic languages but not German or Romance languages and thus primarily used Rumanian to communicate. These persons were very interactive and as they did not know much about Austria, in contrast to the bus and truck drivers, they were constantly asking for something, e.g. where to find a certain product, commenting on something (usually complaining about the prices), etc. Consequently, many language situations arose that required active participation and the acquisition of vocabulary and phrases, as one had to explain repeatedly to many different people where to find things, where to pay, where to get things (that these were the correct prices), etc. This created a very interactive atmosphere, by far not unpleasant, as social interaction was of a very friendly nature, but it did require the knowledge of and, more importantly, increase of Rumanian vocabulary as otherwise, communication would not have been successful. Finding a linguistic solution to the problem of facing a different language was the main objective (as always in a Pidgin language situation), as communication would have collapsed otherwise. Unlike before with bus and truck drivers, one could not easily end communication when one did not know what a person wanted or ask another co-worker to deal with this problem as all of them were involved in language situations themselves and there were simply too many customers present to ignore. Most definitively it was the large number of people that became involved through (large-scale) trade that facilitated such a necessity of communication and through this circumstance the Austrian-Rumanian Pidgin originated.

149 The important point here is that (high frequency) trade facilitated (a great amount of) interaction between speakers of different mother tongues. The perpetuation of such interaction is essential for a Pidgin to originate, continue existing and develop further because without persisting communication the Pidgin (as well as the very communication itself) simply breaks down and vanishes. A Pidgin, especially an ad-hoc but also a fixed Pidgin, is more of a “Provisorium”, i.e. a temporary solution or temporary arrangement than a (fixed or) stable language and can thus not sustain itself over time without communication input.

Chapter 9: A question of definition

Concerning earliest language contact and the further acquisition of vocabulary, by definition (Holm, 1988: 4-7) this must have immediately led to the creation of an ad-hoc Pidgin that had no lasting effect, as it was disregarded as soon as the trade was finished. However, one might argue whether this truly was the creation of a Pidgin. Vocabulary is not necessarily bound to grammar in language contact and can thus be acquired (immediately) and used separately (from grammar), while grammar rather has to be acquired over time. As all development states of a Pidgin, or languages in general, are rather tied to a grammatical development than mere vocabulary acquisition, one could argue that the creation of a Pidgin is the result of a long(er) process of development, starting with the acquisition of new vocabulary which predates the Pidgin origin but continues in the later development states (as new words have to enter). Due to grammar itself being less flexible and only developing in larger intervals (than vocabulary) and itself represented in (or simply called) ´Pidgin states´ (even though the Pidgin definition is also closely tied to social aspects), development is divided into several Pidgin states of which each one requires a certain extent of language input to proceed. Only once language contact increases to a specific amount or volume will a Pidgin develop to a further state. The involvement or establishment of grammar can only be found following the fixed Pidgin state. The so-called ad-hoc Pidgin state arguably does use trace amounts of basic grammar of the substratum language, but the grammar itself is neither fixed nor used unitarily by any speaker, as it has not yet been established. Therefore, the earliest developments of language contact could rather be classified as vocabulary acquisition than creating a Pidgin. Historically, a (true) Pidgin development could only have happened since the late 13 th or, more

150 likely, in the 14 th century. Commerce had reached an extended enough scale for regular language contact to be established and this would be the point where a (fixed) Pidgin could start to originate. Interestingly, by definition, a fixed Pidgin probably did not originate before the 14 th century, as the requirements for a fixed Pidgin could, technically, not be reached due to the strong Catalan input which provided a Catalan language competence very close to its standardized form. However, by definition, language contact must have created an ad-hoc Pidgin. Although it may only be a minor detail whether contact immediately caused the origin of an ad-hoc Pidgin or if contact (which was too low for that and) only caused a vocabulary exchange, it is an important detail for uncovering the origin of the Lingua Franca and should, at least, be considered. The most common definition of an ad-hoc Pidgin actually is language contact of any kind, meaning that as long as two people with different languages are communicating, an ad-hoc Pidgin is formed 23 . Personally, I would not define this as a Pidgin in the strict sense, as this constitutes mere language contact without establishing either vocabulary or grammar, i.e. the most basic traits of a (full) language. Technically, an ad-hoc Pidgin does not mean that a (fixed) language is established, although the definition itself states that an ad-hoc Pidgin only remains for as long as contact is upheld and then vanishes as soon as both speakers leave. If, however, a Pidgin is defined as at least a steady language spoken by minimally a small number of people and/or a specific restricted group, the fixed Pidgin state could be (re-)classified as the origin of a Pidgin. The reason why the assumption that only in the fixed Pidgin state can a Pidgin be regarded as true could be reached is because it describes regularity: it is a language that in its basic features is consistent and at least has a stabilized basic grammar. This does not occur in the ad-hoc Pidgin state or through the acquisition of vocabulary. Language contact between Catalan and Arabic speakers until the 13 th century, I would argue, can not be classified as a Pidgin state as contact was too restricted to establish a fixed Pidgin. What did increase was language contact and with it the basis for vocabulary acquisition (in the 11 th/12 th century), which in turn led to an increasing (Catalan) language competence later on (in the 13 th century).

9.1 Different processes

The confusing issue here is that even the attempt at communication between speakers of different

23 This is the definition by Holm, 1989: 4-7 (mentioned above) but also represents a common view on ad-hoc Pidgins 151 mother tongues is regarded as an ad-hoc Pidgin, and therefore the learning of a few words and phrases of another language could be confused with the formation of a Pidgin. However, the attempt to learn vocabulary is not the same as the formation of a Pidgin itself, although it can lead to it. Pidgins, I believe, are determined by the combining or rather reassembling of two languages in vocabulary and grammar, while language competence is the acquisition of vocabulary (and grammar) without any integration of (the mother tongue) grammar into the learned language. The major difference between learning single words (and phrases) and creating a Pidgin for communication purposes is that one (language) process represents language competence and is the effort of one language speaker to learn another language (in part). This includes the typical and unintentional mistakes when learning a new language, i.e. mispronunciation of vocabulary, mixing up grammatical forms etc., and although grammatical features of the mother tongue are used (to compensate a lack of knowledge of correct secondary language traits), this can be seen rather as early errors in second language acquisition. The languages involved, however, remain separate from each other. The pidginisation process represents the combination of two separate languages where both vocabulary and grammar (of these languages) become simplified, reassembled and intertwined with each other. This language process represents a creation of a language (based upon two separate languages). Although these two linguistic processes seem to be quite similar to each other (especially if the early Pidgin state and the very early stage of (secondary) language acquisition are compared), they clearly differentiate themselves through the inclusion of grammar, in the sense that a new grammar is created in Pidgin languages. Consequently, language contact does not inevitably lead to the establishment of a Pidgin language, it may instead result in language acquisition and competence.

9.2 A misleading theory

Taking this (somewhat confusing) definition at face value and applying it to the situation of the Lingua Franca could form a biased opinion about the origin of the Lingua Franca. The definition of the ad-hoc Pidgin state is slightly misleading, as it not only states that a Pidgin always emerges through language contact between different languages, but could also be interpreted as language contact always intentionally forms a Pidgin (which is not the case). Taking into consideration

152 that, by definition, continuing language contact is (always) bound to develop further, this could lead to the assumption that the ad-hoc Pidgin formed at the very first language encounter must also represent the earliest form of a (later) Pidgin language (although this can only be determined in retrospect). By definition, an ad-hoc Pidgin must have existed ever since the re-establishment of commerce as an ad-hoc Pidgin basically is defined as language contact itself between speakers of different mother tongues. Therefore, the first language contact among Catalan and Arabic speakers would represent the origin of the Pidgin and every other following event must have been a further development of the same Pidgin due to the continuation of this kind of (Catalan-Arabic) language contact. The assumption would thus be that the Lingua Franca originated in 1050, remained in its ad-hoc Pidgin state throughout the 11 th century, slightly developed during the 12 th century due to commerce as well as language contact (somewhat) intensifying, and probably reached the fixed Pidgin state in the 13 th century, due to commerce becoming more important for the Mediterranean until it finally reached the expanded Pidgin state in the 14 th century because of Catalan reaching its maximum importance and spread in the middle of the 14 th century. Furthermore, due to the definition of the ad-hoc Pidgin and through the continuation of Catalan- Arabic contact, it could also be assumed that the Catalan adapted by Arabic speakers was in fact pidginized Catalan. Depending on the intensity of contact, pidginized Catalan language competence would have started on the level of an ad-hoc Pidgin state and remained as such until the increase of commerce occurred and thus the Catalan language competence would have reached the development level of a fixed Pidgin state. It could also be assumed that pidginized Catalan, i.e. the Catalan spoken by Arabic speakers was, in fact, the Lingua Franca. Taken even further, the assumption would inevitably arise that many ad-hoc Pidgins and consequently not only one but rather many Lingua Francas must have been created in several different locations due to language contact being always established anew in many different locations through the eastward expansion of the Crown of Aragon since the beginnings of commerce. Curiously, as by definition an ad-hoc Pidgin would definitely originate through small scale commerce, the origin of the Lingua Franca could be seen in small commercial ports of the North African coast far before any major commercial city was reached, although the (further) development of the Lingua Franca into a fixed or expanded Pidgin must have occurred where large-scale commerce was present and the most plausible location would have been a commercial city (e.g. Algiers). Once the fixed Pidgin had originated, it would also have been carried back to the small ports which ensured the Lingua Franca to become updated as merchants had to take the

153 same port stops on their way back to their home bases. However, as these ad-hoc Pidgins were spatially separated they would have never spread into the rest of the Mediterranean or outside their own location.

9.2.1 Disproving this theory

Personally, I would argue against the assumption that the Lingua Franca itself existed before the 14th century, as several reasons definitely disprove this assumption. On the grounds of the Lingua Franca being a Catalan-Arabic Pidgin alone, but also due to political and economic developments in the medieval Mediterranean, it would be possible to narrow down the date of the Lingua Franca origin between the 11 th and 16 th century; these dates correspond to the earliest beginning of Catalan as a language on the rise and the official end of the Crown of Aragon in the 15 th century, including the final demise of Catalan itself in the 16 th century. However, further a consideration of social and linguistic factors allows a narrowing down of the origin of the Lingua Franca even further to a much more realistic date of origin of between the 12 th and 15 th century due to Catalan gaining its very own linguistic features as a language in the 12 th century, due to Catalan becoming predominant in the field of commerce in the 14 th century (also due to the close resemblance of the Lingua Franca to 15 th century Catalan texts 24 ). Finally, contact to Arabic speakers, although established since 1050, only occurred on a more regular basis since the late 13th century, but reached new heights in the 14 th century through the political and economic dominance of Catalan. Therefore, the 14 th to 15 th century would be a plausible date of origin.

9.2.1.1 Linguistic proof

Regarding language itself, Catalan was not fully developed as a language as such before the end of the 10 th century, as were all Romance languages at this point in time. Basically, any Romance language in the time before the 11 th century still closely resembled Latin and only slowly and with great effort developed into an independent language. For instance, Catalan only started its development around the 8/9 th century and remained more close to Latin until the 10 th century (orbilat.com/Languages/Catalan/catalan.html). Italian, for instance, did not establish its own

24 Cf. Alfajarín, J. R. 2000, Fischer, S. 2003, Igual, V. 2003, Mateu. J. 2009, Soler Llopart, A. 2003, Rialto.it(website), Trobar.org(website) and Lingweenie.org(website). For further reading see the Webliography and the Further Reading section on p.239-243. 154 identity until the 14 th century and Occitan reached its state around the 12 th century (orbilat.com/Languages/Occitan/index.html). If the Lingua Franca had originated before (or even during) the 11 th century, it would have been much more close to Latin in more regards than the sources would suggest and a relexification process from Vernacular Latin to a Romance language could not have affected the grammar to such a strong degree that no distinct traces of Latin would be left. Thus this scenario would not only seem unlikely but rather implausible. Despite the relexification of vocabulary, grammatical traits do not change to such a high degree which could prevent their identification in later varieties of a Pidgin. After all, Pidgin languages tend to remain with their original grammar, and by any linguistic analysis Latin can be excluded as a source language of the Lingua Franca. Therefore, the Lingua Franca can not have originated before the 11 th century and also unrealistically before the 12 th century. In regard to language, contact to Arabic speaking merchants had yet to be established, because communication did not start off fluently immediately, but was a long process that required constant interaction, with the result of a Pidgin after centuries of creating the necessary conditions for its origin. This can not have been the case within the first centuries since the re- establishment of trade. The establishing of close and constant language contact and concomitantly communication itself, started on a very irregular basis and only slowly developed until commerce had reached a considerably high level that allowed frequent language contact in the late 13 th to 14 th century. Only then could communication between Catalan and Arabic speakers be established to such a degree and on such a (consistent) basis that both a high Catalan language competence including the learning and stabilising of words and phrases and a general necessity for a means of communication was created. This could have been the point of origin of the Lingua Franca. With this step, and not beforehand, could a (fixed) Pidgin have originated as communication was established but only to such a degree that allowed speakers to expand on (a few) regularly used vocabulary and phrases due to a certain repetitive pattern or regularity (i.e. repetition) of language contact that had become established by many speakers. Vocabulary, in this case, was a learned requisite of words to be able to name the most common items, which in itself does not necessarily represent a (fixed) Pidgin, but if expanded upon might lead to its creation and this point of development existed the earliest by the (late 13 th to) 14 th century. Linguistic evidence further strongly suggests that later Catalan texts from the 14 th, but especially

155 15th century are far closer to the Lingua Franca than Catalan texts of before this date 25 . In fact, especially later Catalan texts are almost identical with the Lingua Franca and can with ease be adapted, or “translated”, into Lingua Franca. Most curiously, 12 th century Catalan still rather resembles Latin and does not strongly approximate the Lingua Franca in grammar or vocabulary, presumably due to Catalan only appearing in written documents in the mid 12 th century (orbilat.com/Languages/Catalan/catalan.html). Furthermore, the Lingua Franca also shows fewer similarities the earlier the Catalan texts are dated, thus the conclusion can be drawn that the Lingua Franca, at least in the form that is known from the source texts, can not have reached its (known) Pidgin form before the 14 th century. Furthermore, the assumption that it could have been one and the same language and, consequently, one continuous process of development from the 11 th to 14 th century would appear to be rather unlikely, if not impossible, as no Pidgin language could have remained in its ad-hoc Pidgin state for several centuries (without either vanishing or developing further). A further possibility would be that the input Arabic speakers received from Catalan speakers due to Catalan being the official language was actually preventing the pidginization of Catalan owing to its strong presence, official status and the massive input from Catalan merchants. Therefore, no (fixed) Pidgin as such could have arisen before the 14 th century (and presumably only in a commercial city due to the intensity of language contact and amount of Catalan speakers).

9.2.1.2 Documentation proof

Considering the (beginning of the) 14 th century as the earliest plausible date for the Lingua Franca to originate in the form that it is known from the documents would date the origin of the Lingua Franca at a time when both Catalan commerce gained Mediterranean importance (Chaunu, 1979: 83) and the Catalan language was fully developed, which is about two centuries before the first source, Haedo (1612 quoted in Cifoletti, 2004: 197-202), explicitly states its existence. One curious point to note here is that not any evidence of the Lingua Franca exists before the 16 th century, unless one counts the poem of La Zerbitana , which is dated 1353 but seems to have been written earlier in the 14 th century (Lang, 1992: 67, 68). This poem has been disregarded as the Lingua Franca and it would be difficult to truly prove that this is, in fact, the

25 As mentioned above, cf. Alfajarín, J. R. 2000, Fischer, S. 2003, Igual, V. 2003, Mateu. J. 2009, Soler Llopart, A. 2003, Rialto.it(website), Trobar.org(website) and Lingweenie.org(website) 156 earliest form of the Lingua Franca itself. Most curiously, this poem appears to be similar to Catalan. However, it also shows forms used in the later documented Lingua Franca such as “barra fuor ” ´get out´ (Schuchardt, 1979: 33). Curiously, it uses both the Italian infinitive form but also the (presumably) Catalan or Lingua Franca form, i.e. both ´parlar´ and ´parlare´ appear with the Italian form always at the end of the line. What would be an indication that this was primarily a poem and not a linguisticly accurate description is the metre (of the poem) and the fact that in the poem itself the infinitives at the end of the line always end with the Italian infinitive forms such as parlare and conciare , but use the Catalan (or Lingua Franca) infinitive form in the beginning e.g. parlar and voler (Zerbitana, 1353 quoted in Collier, 1977: 295) 26 . However, if Haedo (1612) is considered, the evidence presented is from the 16 th century but it is stated that the Lingua Franca had existed there before. Thus one could argue that although no written sources exist before the 16 th century, there is suggestive evidence that the Lingua Franca had existed before this date. The inevitable question would be if this could possibly date back to the 14 th century. Unfortunately, there is no hard evidence of the Lingua Franca before the 16 th century, the only possible exception being the poem of La Zerbitana , and there is hardly any evidence referring to a language that would resemble the Lingua Franca. Hence it is very difficult to support this claim via documents. Linguistic analysis of the earliest Lingua Franca form that has explicitly been documented suggests that the Lingua Franca was an expanded Pidgin. Although the Lingua Franca is only reported in the 16 th century, it would be rather unrealistic to assume that it suddenly appeared without undergoing the usual stages of Pidgin development. Therefore, as the Lingua Franca is already reported as an expanded Pidgin, it seems only logical that it must have reached its documented state earlier on. However, how far back the date of origin can actually be dated is up for debate, as it could simply be a few decades. Suffice it to say that the plausibility for the origin of a Catalan-Arabic Pidgin in the 14/15 th century is (at least) theoretically granted, as it would match with historic and economic factors. Although documentation and source texts never state the existence of a previous Pidgin state, one might (still) argue that the Lingua Franca originated beforehand as a recognizable (fixed) Pidgin and view the time before its documented proof as a constant development process into the

26 O la Zerbitana retica? Oi Zerbitana retica, Il parlar ch´ella mi dicia ! : come ti voler parlare ? “Per tutto lo mondo fendoto, se per li capelli prendoto, E barra fuor casa mia”. Come ti voler conciare ! 157 expanded Pidgin state. A development process must have occurred, as the Lingua Franca would very unlikely have originated as an expanded Pidgin, but presumably also did not remain in its ad-hoc or fixed Pidgin state for long. Hypothetically, if the Lingua Franca originated earlier than the 16 th century, there should be sources showing or, at least, indicating this circumstance. However, this does not seem to be the case. Technically, as Pidgins develop rather quickly, the Lingua Franca presumably reached its documented form, the expanded Pidgin state, within a short amount of time due to the politico- economic stability of Catalan and its economy (Chaunu, 1979: 62, 83), regulated Catalan-Arabic contact, the continuously growing intensity of language contact within the Mediterranean and a pre-existing Catalan language competence. The development of the Lingua Franca into the expanded Pidgin state (as it is known), therefore, could have been accomplished within a short space of time (perhaps less than a few decades). For instance, the Austrian-Rumanian Pidgin only took one year to become an ad-hoc Pidgin and about 5-6 years to become a fixed Pidgin. Judging from this development, it probably would only have taken a decade or two to reach the expanded Pidgin state. Therefore, a rough estimate of 20 years could represent a plausible assumption. This, unfortunately, does not prove that the Lingua Franca originated much earlier than when it was first reported. However, as the Lingua Franca did develop through political, economic and linguistic factors, at least the plausibility is granted for the origin of a Pidgin due to these beneficial circumstances accumulating since the 14 th century. Additionally, several other historical, political and economic factors would plausibly suggest an (much) earlier date for the origin of the Lingua Franca.

9.2.1.3 Historical (politico-economic) prove

Historically relevant evidence strongly points to the 14 th century. From 7 th to the 13 th century Mediterranean commerce was on a low, although recovering, level of commerce and had yet to be established again from almost nothing, as after the collapse of the Roman Empire commerce “was characterized by the slow reconstruction of maritime trade networks on the ruins of the Roman Empire” (Valérian, 2014: 83). It took almost a century until it even reached a low volume and was then only slowly increased because commerce at that time was still a very risky venture (Pollard, 2014: 467 and Backman, 2014: 170) and remained as such for quite some time. Not many merchants could afford such a high investment or risk, as failure would definitely have

158 ruined many merchants of that time. Commerce, and with it language contact, only reached a noticeably higher intensity in the late 13 th century with its (Catalan) peak in the 14 th century. This definitely was an economically and politically important (historic) event, as well as a point of culmination which facilitated the (possible) origin of the Lingua Franca. By the 14 th century Catalan commerce flourished because (politically) new and steadily developing laws in favour of commerce and thus strengthening the economy were passed during the course of the 12 th to 14 th century, i.e. the earliest forms of “The Consulate of the Sea” (Chaytor, 1933 and The Consulate of the Sea, 1494). In fact, the 14 th century marked the (furthest) development in many different yet interconnected areas such as political development, i.e. with the Crown of Aragon reaching its furthest territorial expansion as a thalassocracy; economically, i.e. the peak of the Catalan dominated Mediterranean economy including many beneficial political decisions, which allowed not only major developments in, but also a great increase of, commerce; and linguistically, through language contact with the (whole) Mediterranean. It is important to note that political and economic developments can create the necessary conditions for language contact; for instance, political peace treaties allowing economic contact, supply and demand of a free market, a functioning economy creating various (even linguistic) exchanges among economico-political systems (and languages), political laws such as The Consulate of the Sea (1494) stabilizing the economy, the (politically motivated) funding of infrastructure and the (economically motivated) set-up of trade routes, technological developments based on experiences drawn through commercial needs such as increasing the range of ships, the employment of political relations with other nations through chancery and, most importantly, the bureaucratic establishment of (continuous) economic contact through (political) contracts. These are the intertwined political and economic basic requirements which allow (peaceful) exchange among different people and thus provide the essential conditions of language contact such as cultural contact and interaction, language identity and development etc., which (can) further allow and facilitate Pidgin development. Basically, one development led to the next as all these developments were connected. Political power stabilized but also increased (through) commerce, and through that influence commerce also became politically dominant and was further facilitated. At the same time, the Catalan language also gained importance in commerce, political importance in the Mediterranean and thus gained an enormous spread, which allowed close and intensive (language) contact between

159 Catalan and Arabic speakers, and with it the Lingua Franca gained the possibility to originate. The 14 th century represented the peak of Catalan commerce, which in itself would almost inevitably have led to a major development in language development and competence, but also linguistically, because the Catalan language was fully developed as well as spoken in the whole Mediterranean basin (and thus consistent language contact to Arabic speakers was guaranteed). This date would thus qualify from an economic point of view, which would not only suggest that the Lingua Franca originated around the (middle) 14 th century, but also that the further development of the Lingua Franca would have occurred immediately. The expanded Pidgin state would have been reached quite soon due to the accumulation of beneficial circumstances. This assumption, in turn, would fit the political development of increased language contact among Europeans and Arabic speakers, and the fast development and spread of the Lingua Franca (a century later). It also seems much more plausible, given the ability of Pidgins to develop at an immense speed if the necessary conditions are present, that this development would occur as soon as the circumstances would allow it, i.e. once commerce had reached the necessary degree to facilitate intensive interaction which would have allowed the Lingua Franca to become an extended Pidgin within a very short amount of time, a few decades even would suffice. Therefore, the 14 th century peak of Catalan commerce and language could be interpreted as both the point of origin of the Lingua Franca as well as the opportunity for it to develop into an expanded Pidgin. The origin of the Lingua Franca could possibly also have been in the 15 th century because both the Catalan language was still spoken and Arab merchants would still use Catalan as a means of communication. Furthermore, the decline of the Catalan (language) dominance would have allowed the pidginization of the same language, as there existed less pressure to use the full (Catalan) language. The reason why the Lingua Franca definitely could not have originated before the 14 th century is because language contact was not on a large enough scale which would have allowed the origin of the Lingua Franca as it was documented, and large-scale commerce was not fully established yet.

9.3 Conclusion

The reason why the Lingua Franca was created at the earliest 250 years after the actual opening up of the North African shore and initiating of Catalan-Arabic contact is due to several political,

160 socio-linguistic, but seemingly primarily economic factors. Although language contact itself might have started earlier and Catalan language competence previously existed, it took some time for the need of a common means of communication to develop and thus for a Pidgin to gain the possibility to originate, develop and spread. Looking at the political and economic development of the Catalan Empire, this can only be confirmed, as through the interdependency of political, economic and linguistic factors a solid basis for the Lingua Franca origin and its development could have been provided. The Lingua Franca presumably originated depending on what date exactly commerce had created language contact suitable for both large scale trade as well as constant interaction suitable for the creation of a Pidgin. As a rough estimate, 1352, the greatest expansion of political, social and economic dominance of the Crown of Aragon (Chaytor, 1933), seems the most plausible date of origin as linguistic developments are often linked with historic events, and this event really opened up commerce to a broader public. Once commerce had reached this large scale level, the increased language contact (for quite a variety of language speakers) created the necessity for a Pidgin and allowed its creation (and development). Concluding, the middle 14 th century could have been the starting point of the Lingua Franca origin, spread and development.

161 Chapter 10: The geographical spread and development of the Lingua Franca

Having elaborated an (assumption about the) approximate or plausible date for the Lingua Franca origin, the question about the spread and development, including the place of origin, can roughly be answered as well. The political and economic consequence of the historic event of opening up the Southern Mediterranean for commerce was that a number of European merchants were venturing to the Arab Empire to trade with Arab merchants and established language contact with Arabic speakers (as well as an economy). The location in which the Lingua Franca originated therefore must also be where commerce started and the opening up of the Arab Empire for European merchants was, for the most part, commenced in Arabic-speaking locations, i.e. the southern Mediterranean shore. Presumably it must have been the western part, as the beginnings of commerce would be bound to the available (nautical) technology which did not allow long distance travel. Furthermore, commerce definitely was conducted more closely to the sovereign territory of the Crown of Aragon as the eastern locations would be reached only later in time after the continuous (eastwards) expansion, which lasted until the most eastern point and full expansion of the Crown of Aragon was reached in 1352.

10.1 14 th century development of the Lingua Franca

In the 14 th century much changed, as the broader public of Catalan merchants was venturing out into the Mediterranean. Thus large numbers of Catalan speakers were in constant contact with Arabic speakers. This led not only to the acquisition of words and phrases, but made interaction dynamic and thus the final step necessary for the origin of a (fixed) Pidgin was set in motion. Large-scale commerce provided the social environment and numbers, but the development of the newly created Pidgin occurred (automatically) through constant interaction between speakers of Catalan and Arabic. It may be far-fetched to assume that the acquired grammar in itself was complex, far from it, but through newly assembling and applying grammatical indicators such as morphology and prepositions, words acquired new meanings and vocabulary was put together in new and unseen ways, influenced by two different languages. At this point, one specific curiosity should be pointed out; that not the whole vocabulary and the whole grammar of separate languages are combined in Pidgin languages, but rather a high

162 statistical value of one language and a tiny percentage of the other of both vocabulary and grammar. What is adapted from which language is, from a subjective point of view, seemingly coincidental, although objectively this distribution seems to be determined by political and especially economic dominance as it depends on which language forms a majority (often referred to as being dominant), which, in turn, is determined by political factors. While contact was constant but irregular before the 12 th century, once large-scale commerce was established, which happened around the very beginning of the 14 th century, contact between merchants (and possibly also the local population) changed immensely (Chaunu, 1979: 57 and Wansbrough, 1996: 139). By then it had become a daily event, (as well) as contact had become even more close than before involving a high degree of interaction (and participation) on both ends (i.e. by merchants). Consequently, language situations arose with repeated consistence, which led to people trading and conversing with each other extensively and, more importantly, to the origin, development and spread of the Lingua Franca. Without previous language contact and establishment of words and phrases, which created a certain (Catalan) language competence as well as a familiarity with the other language (Catalan) among Arabic speakers (merchants), a Pidgin could not have originated. Presumably this only occurred when considerable numbers of (Catalan) merchants came to the North African ports and cities to trade, interacting with the local merchants on a daily basis. Thus communication was established to such a high and mutually involving degree that the necessary setting and conditions for the origin of the Lingua Franca were created. It would appear that large-scale commerce facilitates language contact and therefore the (possible) creation of Pidgins in general, or at least the Lingua Franca. In any case, language contact is important, otherwise a Pidgin can not originate or develop. However, contact between speakers of different mother tongues does not lead to the creation of a fixed Pidgin, as regular language contact may lead to an approximation between speakers, which is expressed by the acquisition of words and phrases, but does not lead to an amalgamation of vocabulary and grammar (i.e. a fixed Pidgin), even though this by definition is an ad-hoc Pidgin. However, if contact sustains and increases, then a development occurs, leading to the origin and immediate development of a Pidgin language. Therefore, large numbers of interactions on a daily basis create a (fixed) Pidgin and its consequent development which, in the case of the Lingua Franca, was provided through large-scale commerce.

163 10.2 Analysis of language input

In the beginnings of the re-establishment of commerce there were not many other travelling merchants besides Italian and Catalan merchants venturing to the Southern Mediterranean, Occitan merchants were involved in trade (Valérian, 2014: 84) and Jewish merchants too (Astren, 2014: 399). However, it would seem appropriate to assume that Catalan was the main language influence and that Arab merchants would use Catalan words for communication, probably also with Occitan and Italian merchants. Due to the ongoing language contact that Catalan and Arabic speakers had through commerce, the Catalan language competence of Arabic speakers would slowly increase with the acquisition of words and (later) phrases. Although Arabic speakers would adapt the acquired Catalan words, it remains unclear if the Catalan speakers ever adapted to the Arabic language itself. After all, since the establishment of the Crown of Aragon in the Mediterranean, Catalan was establishing itself as the predominant commercial language. Thus it is doubtful that Catalan speakers, even merchants that had close contact to Arab merchants, would use any other language than Catalan, simply because Catalan was the official language of the Crown of Aragon and its colonies. Especially later in the 13-15 th century when Catalan became the official language (of commerce) used at sea and for maritime law ( The Consulate of the Sea , 1494), it was known and spoken in the whole Mediterranean. As a consequence there must have been some language influence of Catalan on Arabic speakers in both smaller ports as well as commercial cities such as Algiers, Tripoli and Tunis, creating a Catalan language competence in the southern Mediterranean. Presumably, larger cities would have had higher language proficiency due to a higher frequency of communication and more consistent Catalan input (compared to smaller ports). Once commerce reached a large scale it presumably caused Catalan to become almost inevitable in the sense that almost all Arab merchants had some level of Catalan language competence and used this knowledge to communicate with every ´franc´, i.e. European (merchant). Given that many Arabic speakers denominated all Europeans as ´francs´, assumed that all francs spoke the same language and that they themselves spoke the ´language of the francs´, meaning the Lingua Franca, it seems quite likely that they used the Lingua Franca with all Europeans (merchants, travellers, slaves etc). In a sense, Arab merchants provided Lingua Franca input to European merchants with the effect that non-Catalan speaking travelling merchants adapted the Lingua Franca. Catalan traders may have been the group who provided consistent Catalan language input, but it was Arab merchants that created, used and developed the Lingua Franca. 164 The result would have been that at the height of (Catalan) commerce Arabic speakers pidginized the Catalan language using it to communicate with both Catalan speakers as well as speakers of other Romance languages. These other Romance speakers would then adapt the Lingua Franca (or pidginized Catalan), thereby creating the dynamics of the Lingua Franca update process. Consequently, it must have been Arab merchants that also spread the Lingua Franca (although not (primarily) by them travelling). It can not be fully excluded that other travelling merchants might have been able to speak Catalan and thus not every travelling merchant would have adapted the Lingua Franca. However, those that communicated with Arab merchants and did not know Catalan most likely picked up this language (and spread it across the Mediterranean), because it allowed communication with both Arabic speaking merchants (all over the Mediterranean) as well as Catalan speakers. This was useful if one did not speak the Catalan language itself, i.e. it was useful for sailors, merchants and people in close contact with these groups. Furthermore, the Lingua Franca could also be used to communicate with other Romance merchants due to the compatibility of the Lingua Franca with other Romance languages. Therefore, if this scenario were correct, the Lingua Franca development and spread occurred through the Catalan language. At first, spreading through Catalan travelling merchants in the initial time of commerce (the 11 th to 12 th century) with the acquisition of words and phrases by Arabic speakers as a result. Once commerce started to increase during the 13 th century Catalan language competence developed constantly, eventually reaching a quite high proficiency level among Arabic-speaking merchants. Meanwhile, Catalan was spreading within the whole Mediterranean providing a solid basis for Pidgin formation, which eventually led to the origin of the Lingua Franca, presumably in the middle of the 14 th century. Furthermore, as many travelling merchants became involved in, and were conducting large-scale commerce, these merchants would adapt the most developed form of the Lingua Franca, i.e. the extended Pidgin. Due to the compatibility with Catalan and the Lingua Franca being omnipresent in the (already opened up) ports and cities (thus allowing communication with other merchants in the Mediterranean), the assumption that Catalan language competence, which then created a Pidgin that became adapted by both Arab as well as European travelling merchants does not seem unrealistic.

165 10.3 The dynamic cycle of language input

The language that primarily spread in the western and central Mediterranean until the 15 th century was Catalan and the result was that many Arabic speakers all over the various ports and cities in the (western) Mediterranean developed a certain Catalan language competence. Eventually, a Pidgin would develop which was adapted by (Romance) travelling merchants. The most probable scenario of travelling merchants adapting a Pidgin would be that it was a Pidgin which was in the centre of commercial relations. The focus of commerce lay within the commercial cities, consequently travelling merchants would adapt a Pidgin which had developed in (one of) these commercial centres, presumably Algiers, but Tripoli and Tunis would also be a possibility. Judging from economic developments but also from the linguistic documents, primarily Haedo (1612), Algiers was the biggest and most influential commercial city in the western part of the Southern Mediterranean. It could be assumed that Algiers could have had the most developed Lingua Franca which, once established, could also have been the Lingua Franca variety adapted by other travellers. Consequently, a dynamics of exchange developed between Arabic speakers and travelling merchants, which in turn had the effect that Catalan was the main source of input for Arab merchants but also pidginized Catalan input was provided. The effect must have been that the Lingua Franca remained very close to its original superstratum source language (i.e. Catalan) but was also susceptible to changes or rather variations in vocabulary and, to a lesser extent, grammar. The dynamics of language exchange between Lingua Franca speakers causing the Lingua Franca to constantly update itself truly started when the numbers of (other) travelling merchants had increased considerably (and with it the number of Lingua Franca speakers). This must have been the point where the Lingua Franca became adapted and carried around by a large number of speakers, and occurring after the 14 th century, it can also be assumed that this was the extended Pidgin variety of the Lingua Franca. Due to the increasing number of travelling merchants (in the 16 th century) speaking different languages, and who had different pronunciations, vocabulary and phraseology, a slight influence occurred altering some aspects of the Lingua Franca, e.g. in the field of (preference of) vocabulary, expressions but also grammar (e.g. prepositions). The Lingua Franca became intelligible beyond the necessity of vocabulary shift and developed into a more internationally intelligible (expanded) Pidgin. Once this dynamism of double, instead of single, input started (on a large scale) the whole dynamics of Lingua Franca spread changed, or rather adapted to this specific process as now not 166 only Catalan input prevailed but an additional factor emerged which allowed the acquisition of an already established (pidginzed) language. This changed or rather avoided the language contact situation of being unable to communicate with a different language speaker. Due to an already established Pidgin language, the whole slow acquisition process of Catalan-Arabic contact gradually establishing words, phrases and, at some point, a grammar (although limited to a very specific degree) as well as slowly rooting itself in a location, could be met with an already established Catalan-Arabic based language. The acquisition of the expanded Pidgin almost immediately provided, if not guaranteed, intelligibility among all involved speakers the instant other travelling merchants were using the Lingua Franca. One major aspect or result of people carrying around the Lingua Franca was that the dynamics of language exchange started to take effect and (with it) Lingua Franca speakers would constantly carry new vocabulary, phrases, prepositions and maybe even other or new grammatical traits to locations all over the Mediterranean causing a constant update of the Lingua Franca and, a century later, the Lingua Franca varieties. However, as long as Catalan was the official language, the Lingua Franca would only expand, not develop into several varieties, which would only happen later once the Catalan influence and previous Catalan input ceased. Even in the 15 th century, although losing its dominance, Catalan was still strongly present as a language (until the 16th century) and could therefore stabilize the Lingua Franca by providing a language basis for the (Catalan based) Lingua Franca, with the result of the Catalan Lingua Franca being a basis for both the Spanish and Italian Lingua Franca varieties. Consequently, the Lingua Franca was updated and stabilized by the Catalan language itself. This makes the development of the Lingua Franca (until the 15 th /16 th century) much more linear, with only words and phrases being added (but no varieties developing), presumably due to Catalan merchants being the primary language input and other travelling merchants being a minority. However, it became quite complex later on once multiple language input, i.e. Catalan and pidginized Catalan input from different European languages, influenced the pidginized Catalan of Arabic speakers.

10.4 Linguistic complexity of language input

Analysing the whole situation linguistically, a very complex contact situation emerged once the Lingua Franca developed and was taken over by a majority of travelling merchants, which is the constant input of pidginized Catalan being equal to the Catalan input itself. This situation created

167 both a constant Catalan input for grammatical stability but at the same time immense intelligibility due to various different choices in vocabulary as well as pronunciation (which probably would have included spelling if the Lingua Franca had ever been a written language), due to the constant pidginized Catalan input provided by travelling merchants. Meanwhile Catalan merchants would simply use Catalan as a language and therefore stabilize the Lingua Franca without being influenced by it. The (non-Catalan) travelling merchants were in a completely different linguistic position as they were not only taking over a new language but were also influenced by Catalan at the same time. This created a dynamic cycle of reciprocal influence which not only allowed the development of the Lingua Franca into an expanded Pidgin but also its spread within the Mediterranean. Later, once Catalan input ceased and other languages became dominant, the multiple language input facilitated the emergence of different varieties of the Lingua Franca, but at the same time ensured their (mutual) intelligibility. Therefore, this whole dynamic cycle allowed what would not have been possible through single Catalan input alone. Due to travelling merchants reaching many locations the number of people adapting the Pidgin constantly grew as well, and as more people spread the Lingua Franca across the Mediterranean even more adapted it, thus expanding this dynamic cycle. Being closely tied to (the limits of) commerce, this process continued until the furthest expansion of Mediterranean commerce and with it the expanded Pidgin state (in which the Lingua Franca remained for centuries) was reached. Basically, (some) travelling merchants took over the Lingua Franca from Arab merchants, but the Arab merchants themselves learned more and more through this process and passed the improved knowledge to other merchants who would carry this knowledge to other locations. With new people constantly acquiring and using the Lingua Franca, the Pidgin would develop, which is a process that occurs on its own (as the field study shows). Furthermore, the more people acquired the Lingua Franca, the more easily they could handle the linguistic problem of a language situation in which speakers of two different languages needed to communicate. Most likely at least one of the parties had a language available that could be understood by the other. Following this development in later years, Catalan merchants would communicate with local merchants, but additionally the Arab merchant would also have contact to other travelling merchants. Therefore, the local merchants would have spoken the Pidgin due to the remaining Catalan input, but also passed on the Pidgin to other travelling merchants and further developed the Pidgin through being exposed to the Lingua Franca (i.e. pidginized Catalan) input of these

168 travelling merchants. Consequently, every travelling merchant that would adapt the Pidgin (or new parts of it) would also adapt a more developed form of the Pidgin and spread it to the next location in which the same process would recur. The travelling merchants were not only going to other locations, they were also returning to these locations bringing with them a further (though only slightly) developed form of the Lingua Franca to share their knowledge with the local merchants. This established a dynamic cycle of both picking up and spreading new or slightly further developed forms of the Lingua Franca between different locations in which these developed (travelling) forms were adapted by Arab merchants replacing or rather updating other local forms of the Lingua Franca. Once in a new location (especially those that pursued large-scale commerce), the Pidgin would settle and be further developed by local merchants but still receive enough input from other locations or rather other Lingua Franca varieties. Intelligibility remained consistent, although (minor) divergences must have occurred due to each city’s unique traits, both linguistic and cultural, including different degrees of European dominance (or occupation).

10.5 Facilitation of intelligibility

Another important aspect of the Lingua Franca is its intelligibility. Commerce and maritime law definitely were responsible for the intelligibility of the different Lingua Franca varieties. The interconnectedness of Mediterranean commerce allowed a very complex process of exchange and input between the Lingua Franca varieties and thus a (kind of) reciprocal effect, i.e. an updating of the Lingua Franca varieties was established, a (so-called) dynamic cycle. Catalan being the official language and offering constant input provided one major archetype or standard to aspire to (i.e. similar to an acrolect) which the Lingua Franca followed. This cycle not only allowed the Lingua Franca to be more easily established in a location, especially as this was a continuous process that occurred consistently, but it was also a process that continued until the Lingua Franca had reached the extended Pidgin state which seems to be the limit of its development through commerce. In the beginning of the 19 th century the Lingua Franca almost developed into a Creole, but not through commerce, instead via the local population. What is most interesting is that the dynamism of commerce itself kept the Lingua Franca alive and shaped or even defined its development. In small ports and cities, in which commerce was at a lower level (of intensity), the Pidgin was not known by many people, the exception being Arab

169 merchants due to a repetitious but limited language contact with travelling merchants. However, in the major commercial cities the Lingua Franca was both known and also spoken by a large percentage of the population due to commerce focusing on, or amalgamating in, such large cities and resulting in increased language contact and exchange among Romance and Arabic speakers. The further effect was that the Lingua Franca even became a cornerstone of stability and intelligibility in these major commercial cities, presumably due to the fact that large-scale commerce was only to be found in the major commercial cities, not the smaller ports. This dynamic cycle functioned perfectly and the Lingua Franca had become an established entity in the three major commercial cities, i.e. Algiers, Tunis and Tripoli, and remained as such until the 19th century. The Lingua Franca having fixed locations in which it developed separately, after the demise of the Catalan language and dominance of Spanish and Italian, further suggests that while these Pidgin varieties had the most influence, at the same time these locations served as a stabiliser for the Lingua Franca. Within these locations the Lingua Franca was constantly spoken (and received input) and without cities such as Algiers, Tunis and Tripoli the Lingua Franca would have been replaced quite quickly, especially after the collapse of the Crown of Aragon. The Lingua Franca as a whole would have vanished or never developed beyond, or even into, an expanded Pidgin and been replaced through a Spanish and Italian Pidgin, if these established entities did not exist. However, due to the considerable number of Lingua Franca speakers within these cities other inhabitants adapted one of these varieties and also brought it to other locations, consequently updating the Lingua Franca varieties via themselves, thus further providing intelligibility among the different varieties. Therefore, the Lingua Franca intelligibility is most likely linked to large- scale commerce and large commercial cities such as Algiers, Tunis and Tripoli, with Algiers being the most probable assumption for being the first major city in which the Lingua Franca developed into an extended Pidgin due to its closeness to Barcelona, and therefore being the first major city to become integrated into large-scale commerce with the Crown of Aragon. If this were true, it would suggest that this occurred around the 14/15 th century, when commerce was established and was thus interactive and dynamic to reach an intensity which allowed the Lingua Franca to develop into the expanded Pidgin state. Presumably this happened in one of these specific locations, presumably Algiers, within a very short period of time as Catalan input ensured the intelligibility and even the relevance of the Pidgin. Meanwhile the already available pidginized form of Catalan, the Lingua Franca, proved to be easy and fast to acquire, and even

170 provided an ideal compromise for both communication and commerce. By learning the Pidgin, communication could be guaranteed and thus commerce was reinforced as well. One other aspect responsible for the intelligibility of the Lingua Franca was its (relative) closeness to (Old) Catalan. Catalan was a consistent influence on the development of the Lingua Franca until the 16 th century and Arabic speakers were assimilating Catalan language traits into their own (pidginized Catalan) Lingua Franca variety, which would then be passed onto other Lingua Franca varieties. The dynamic cycle of travelling merchants constantly spreading both Catalan and a further developed Lingua Franca to the next city (and combining it with or replacing a local form) was a (constant) flow of Lingua Franca knowledge that made the Lingua Franca such a well known and complex but intelligible language which was available for every merchant in the Mediterranean.

10.6 Lingua Franca varieties

The initial point for adapting the Lingua Franca in all Mediterranean locations must have occurred at least in its fixed, but more realistically its extended Pidgin state. Otherwise, such a high degree of intelligibility could not have been achieved and continued for almost half a millennium. Presumably the Pidgin of Algiers was carried to eastern locations where it became adapted but also shaped according to the different linguistic and cultural (i.e. geographic) settings which are apparent in the Lingua Franca varieties. These varieties developed individually and were shaped by the political entity having economic dominance and thus also having linguistic influence in a specific area, i.e. language influence depending on culture and political-economic dominance, with the result of these varieties going into different directions linguistically. What indicates that the Lingua Franca was carried to other (eastern) locations (i.e. Tunis and Tripoli) once it originated and then adapted to the prevailing languages of these locations instead of three completely separate Pidgin languages developing from different languages (Spanish, Italian and French) is that despite the Italianization, Frenchification, Hispanicization and closeness to Arabic of all three Lingua Franca varieties they are far too similar to each other and intelligible among themselves to truly have originated and developed as separate Pidgins. The reason why there cannot have been more than one (original) Lingua Franca simply resides in the fact that although three different Lingua Franca varieties existed they do stem from one and the same original Lingua Franca. Considering the century of Catalan influence before the Lingua

171 Franca varieties developed, this seems logical. It also serves as an explanation why the Lingua Franca remained intelligible in its varieties and hardly varied grammatically despite the local influence and the cultural differences prevailing in each major commercial city in which these varieties seem to have developed. Even if three separate Pidgins were subject to constant Catalan influence and constant language exchange between these different Pidgins would consistently update them, the cultural and linguistic difference of different superstratum languages, e.g. Italian, Spanish or French with an Arabic substratum, of separately developed Pidgins, would have been far too severe to show such strong similarities that exist among the Lingua Franca varieties. Even though these varieties were exposed to separate cultural, economic and linguistic influences (Cifoletti, 2004: 22 and Rossetti, 2005: 17), due to originating from one Pidgin and being in contact with the other varieties they remained intelligible. Cifoletti (2004) convincingly proves that different varieties of the Lingua Franca existed (2004: 31-32) and that these varieties developed in separate locations which were under (different) Romance dominance (2004: 22). Therefore, the Lingua Franca varieties must all stem from one Pidgin which spread across the whole Mediterranean but was influenced by separate (economically dominant Romance) languages.

10.6.1 The spread of one single Lingua Franca

The Lingua Franca adapted and changed over the centuries and arguably the earliest (14 th /15th ), 16th and 19 th century Lingua Franca is not exactly the same language. However, as change is a natural language phenomenon as well as an innate Pidgin feature and there further exist many similarities to the earliest Lingua Franca in the 19 th century variety, including the intelligibility among all other varieties, it can safely be assumed that it is one language. Therefore, the original Catalan Lingua Franca must have been the variety to spread across the Mediterranean, as Catalan traits were present even in the 19 th century Lingua Franca of all varieties. The Lingua Franca and its varieties were one and the same language, not separate Pidgins which developed in different locations and then developed and diversified over the centuries through different political and economic developments. However, developing in separate locations affected the Lingua Franca development in the sense that it shaped both vocabulary and grammar in a geographically restricted geo-political area. The main connection and the one specific feature that all Lingua Franca varieties mutually share

172 is that they are all based upon the same language situation of Romance-Arabic language contact, i.e. they were all subject to Arabic and Romance language input, although different Romance languages prevailed over the course of time. All of these Lingua Franca varieties developed individually and existed alongside each other independently but were still intelligible, in contact with each other and shared very many features as they stemmed from and have been influenced by one and the same Pidgin. Considering the influenceability of Pidgin languages, these varieties share much more than such a linguistic scenario would immediately suggest (which becomes visible when these varieties are directly compared to each other). The Lingua Franca in itself did not change dramatically in its core grammatical traits even after being Hispanicized, Italianized and Frenchified, presumably because the Pidgin state that sustained the Lingua Franca did not change. Once the Lingua Franca had reached a certain state in its development, i.e. the extended Pidgin state, it remained in this state and did not (or only slightly) change its core grammatical traits. The exception was the very common grammatical features such as prepositions, infinitives e.g. adding /e/ to the /ar/ infinitive in the Italianized variety, etc.

10.7 Conclusion

The Lingua Franca derived from Catalan-Arabic language contact but had always been constantly developed through language exchange (and interaction) caused by commerce, which went as far as when the original input (source) language, Catalan, was no longer available and became replaced by other Romance languages, the Lingua Franca remained stable in its core (grammatical) traits due to (unchanged) persisting language contact (although one Romance language was replaced with another). The main source for the stability and intelligibility of the Lingua Franca, i.e. commerce, also did not change, as contact itself still existed in an analogous manner and intensity among Spanish, Italian but especially Arabic speakers as before with Catalan speakers. Therefore the whole (economic) system continued to work as effectively as before with different varieties of the Lingua forming instead of being replaced by other Pidgins. The Catalan influence on the Lingua Franca vanished due to Catalan itself declining from the 15 th to 16 th century. However, once the Crown of Aragon had declined, Spain and Italy would take its political but especially economic place, and thus Italian and Spanish were dominant Romance languages in the Southern Mediterranean. The Lingua Franca adapted (locally) to this shift and

173 was (only) relexified by these Romance languages but remained, in its core (grammar), the same Lingua Franca of the Arab merchants throughout its course of existence. The dynamics of commerce became an integral part of the Lingua Franca development, as through many travelling merchants using Catalan or the Lingua Franca, the Pidgin would not only be updated by Catalan but also by the Lingua Franca itself. Consequently, one could concur that the Lingua Franca has always been intelligible and compatible, even after different varieties developed, since it was in constant language contact, at first with the Catalan language, and later via other Romance languages. The main reason for this intelligibility was that the Lingua Franca was always receiving Romance language input through commerce (and travelling merchants, pirates and slavers) which had been the main connecting factor between the different varieties of the Lingua Franca. This is the true connection between the Lingua Franca varieties, i.e. the consistent connection created through commerce to provide source language input.

Chapter 11: The spread of the Lingua Franca

11.1 Spread among Catalan speakers

It can be assumed that Catalan merchants would simply continue to use Catalan, as it was the official language of commerce, and other travelling merchants, Occitan, Spanish or Italian for instance, would adapt the (fixed Pidgin) Lingua Franca and apply it. Catalan speakers themselves may have gotten used to hearing the Pidgin and communicating with Pidgin speakers. However, as the Lingua Franca was rather close to Catalan it would make less sense for Catalan speakers to use pidginized Catalan instead of Catalan itself. This assumption alone does not qualify as proof for Catalan speakers to not have used this Pidgin, but there are some indications against it. However, Catalan speakers using the Lingua Franca is a possibility due to the potential presence of the expanded Pidgin form of the Lingua Franca in 14/15 th century ports and among specific groups and also because a majority of merchants (other than Catalan merchants) were using the expanded Pidgin. This would allow the assumption of at least some Catalan merchants becoming deeply involved in Catalan-Arab interaction, with the result of adapting the Lingua Franca itself. This, however, would seem more of an individual development rather than a common feature of Catalan-Arabic language contact. The reason is that as long as Catalan was the main language at

174 sea and in commerce, the Lingua Franca remained as close to Catalan as possible. Therefore, it would rather be the Arab merchants that really used the Lingua Franca and Catalan merchants that used Catalan (due to practical reasons), as there existed no pressure for the Catalan speakers to learn the Pidgin, since intelligibility between Catalan and the Lingua Franca, i.e. pidginized Catalan, was given and communication could be established through it. A curious point to note here is that with the Austrian-Rumanian Pidgin the Austrian speakers really made an effort to come as close to Rumanian as possible, which was successful in regard to the limited grammatical input. There were, of course, developments noticeable that led away from similarity to Rumanian (e.g. /p/ became /b/), which were definitely supported or rather caused by Austrian grammar as early as the fixed Pidgin state (i.e. with the integration of (fixed) grammatical traits). This would have probably continued even further had this Pidgin ever reached the expanded Pidgin state, i.e. making the Austrian-Rumanian more of an independent language. However, intelligibility to Rumanian speakers was still present, also due to the constant Rumanian input (including grammatical correction), which occured only in personal interactions. The development away from (standard) Rumanian was not severe, as it was counterbalanced through Rumanian input. Had there been either a more frequent or personal interaction basis, the Austrian-Rumanian Pidgin would have been a lot closer to actual Rumanian due to a constant stream of speakers´ correcting, a multitude of interactions and occurring auto-corrections because of Rumanians speaking Rumanian and not the pidginized form. Therefore, the Lingua Franca remaining as close as possible to Catalan itself (presumably without grammar correction) could be explained through either a rather personal interaction basis of Catalan-Arabic contact or a very frequent and numerous quantity of Catalan input. As Catalan was the official language used by a majority it seems rather plausible that the Arabic speakers were adapting to considerable Catalan input instead of developing personal interactions with a small number of Catalan merchants. Compared to the development of the Austrian-Rumanian Pidgin, the local Arab merchants presumably would have reacted highly responsively to the Catalan input as soon as the necessity to communicate was present and most definitely immediately adapted some Catalan words. The Catalan travelling merchants on the other hand would not adapt (to neither Arabic nor) the Lingua Franca. In the Austrian-Rumanian Pidgin development it was noticeable that the Rumanian speakers would not adapt to the Pidgin as much as the Austrian speakers, because the Pidgin always oriented itself to the superstratum language and once the fixed Pidgin state was reached it was apparently similar to Rumanian. Consequently,

175 it was less likely for people to adapt the Pidgin, and instead they were using standard Rumanian, although they would simplify it when speaking to the Austrian-Rumanian Pidgin users. As the Austrian-Rumanian Pidgin never reached the expanded Pidgin state, it cannot be stated for certain, but this would presumably not have changed in the expanded Pidgin stage, i.e. Rumanian speakers would not adapt the Pidgin. However, (I am convinced) it would have happened in the Creole stage that people would have taken up the Austrian-Rumanian Pidgin instead of simply using Rumanian or a simplified form because this language would have become distinctive from standard Rumanian. Applying this to the Lingua Franca, it would have been possible for Catalan speakers to adapt the Pidgin but this would have required a more evolved and quite distinctive Pidgin language (e.g. a Creole). Furthermore, as long as there existed no deeper involvement or interaction of users, no Rumanian speaker would adopt the Austrian-Rumanian Pidgin while understanding and accepting the pidginized form (i.e. conversing normally). The only exceptions in adopting the Austrian-Rumanian Pidgin were the Rumanian bus and truck drivers who were not only visiting much more frequently but were also involved in interactions more often. This seems to be a very specific condition for the adoption of a Pidgin by a superstratum speaker (than a general rule). The assumption, or even conclusion, could be drawn that superstratum speakers only adopt the Pidgin if they are personally involved with the substratum speaker. Another possible reason, I believe, to adopt the Pidgin as a superstratum speaker would be due to the necessity to learn the Pidgin since it was distinctively different or even unintelligible to speakers of the superstratum language itself. Although this definitely did not apply to Catalan and, presumably, Occitan merchants, but was most likely the case with other merchants such as Spanish, Italian, French, English, Dutch, Greek etc., who probably had to learn the Lingua Franca to be able to trade with Arabic speaking merchants speaking a pidginized form of Catalan with Europeans. It can, therefore, be assumed that the majority of Catalan merchants were only using their own language (and thus spreading Catalan not the Lingua Franca) and only some exceptions were adapting the Lingua Franca itself. The degree of involvement among merchants is only speculative but would suggest that merchants most likely were generally less involved in personal contact with each other 27 . The idea behind this assumption of non-Catalan speakers adapting the Lingua Franca is based on the linguistic development of Romance languages. Although the Romance languages were quite

27 This is a personal assumption based on the experience with the Austrian-Rumanian Pidgin development 176 close in the 10 th century, they developed and started to differentiate themselves between the beginning of the 10 th to 14 th century into the Romance varieties that are also known today. These still do share many similarities but are not fully intelligible. Even though they were more intelligible in the 16 th century, there already existed linguistic differences. Therefore, the assumption that travelling merchants had to learn the pidginized form of Catalan as they would not be able to understand it through speaking Italian or French alone seems plausible. However, it would also have been relatively easy for Romance speakers to acquire the Lingua Franca due to the (relative) closeness of Romance languages at that time and the so-called (grammatical) simplification of the Pidgin itself which, contrary to this belief, consists of basic, not simplified, grammar. Therefore, a necessity to learn a new language might have been created for non-Catalan speakers, while this was at the same time a simple and readily achieved task (i.e. not too time- consuming an effort). The amount of involvement in personal interaction would give more evidence about the Catalan merchants´ relation to the Lingua Franca. Especially if personal communication was indeed a fact, the assumption could be drawn that the Lingua Franca was definitely adapted by Catalan travelling merchants. However, there are a few indicators which would suggest the opposite. The first is the considerable similarity of the Lingua Franca to Catalan and with it the intelligibility of both the Lingua Franca for Catalan speakers but also of Catalan for Lingua Franca speakers. This is an argument against this conclusion because for as long as communication was kept simple, people who spoke Catalan could continue to use simplified (or basic) Catalan and Lingua Franca speakers would most likely be able to comprehend what Catalan speakers were saying. In a language situation in which Catalan speakers would use simple words and basic grammar for communication instead of learning the Pidgin itself, although they definitely had the ability to easily acquire the Lingua Franca, and still successfully achieved limited communication, it seems more logical to simplify one’s own language than to use a pidginized language. Another indication for Catalan speakers to not have adapted the Lingua Franca is its very development, or rather the lack thereof, as until the later centuries there has never been even an indication for the development of a Creole. The Lingua Franca had the potential, as do all Pidgins, to quickly evolve and reach another state of development within a short amount of time, as seen with the rapid development of an expanded Pidgin state or into a Creole(-esque) state in the 19 th century. However, as the Lingua Franca remained in one specific Pidgin state it would seem that if Catalan speakers had adapted the Pidgin too, there should have been a Creole

177 development. As the development of the Lingua Franca is solely based on commerce and hardly any personal involvement with the language is either part of or facilitated by commerce (or piracy or slavery) itself, and personal communication via the language cannot even be noticed before the 19 th century in the source texts, it seems unlikely that Catalan speakers acquiring the Lingua Franca would seem a realistic scenario. Even the possibility that personal interactions occurred among Catalan and Arab merchants themselves seems unlikely as there should have been a much greater degree of development within a shorter amount of time to at least a Creole(-esque) state if interaction had been the case. Instead, there does not seem to be a development beyond the expanded Pidgin state, at least according to the sources gathered in Schuchardt (1909), Foltys (1984), and Cifoletti (2004). The expanded Pidgin only occurs in the 16 th century sources, which suggests that either the expanded Pidgin state was a very late development or it could have occurred much earlier but was not noticeable to people outside the group of merchants. This lack of development does indeed suggest that commerce would only allow the development to an expanded Pidgin state. Concluding, the (majority of the) Catalan travelling merchants did not use the Lingua Franca and consequently were not the group spreading the Lingua Franca in the Mediterranean, but they spread Catalan, which became adapted by the Arab merchants and the dynamics of commerce ensured that Catalan spread across the Mediterranean. Later on, once the Lingua Franca had developed, other (non-Catalan) travelling merchants adapted the Lingua Franca due to the Arab merchants using the Lingua Franca to speak to all francs i.e. Europeans. The Lingua Franca, therefore, spread through (non-Catalan) travelling merchants and the whole dynamism previously referred to took place. A similar (language) situation also occurred at the gas station referred to previously, where one would use the Pidgin to talk to all Rumanian speakers as well as Polish, other Slavic and even Italian speakers (with surprisingly good results, presumably due to the inclusion of Italian vocabulary). The consequence of Catalan merchants not adapting the Pidgin and spreading Catalan (instead of pidginized Catalan) would have been that, after the establishment of trade, i.e. once commerce had reached a higher level in the late 13 th century, a Catalan language competence was created in locations all over the Mediterranean with the difference that not all of them remained on the same level of proficiency. Smaller locations, i.e. ports and cities, had a lower Catalan language competence and major commercial locations had a high level of Catalan language proficiency,

178 this being due to the smaller ports having less language contact. Catalan merchants only using Catalan to communicate brought about a constant input of Catalan to Arabic speakers and thus Catalan language competence definitely existed. However, as many Arab merchants remained stationary they could not spread the Lingua Franca or be updated in it themselves, but continued to receive Catalan input only, resulting in varying Catalan language competence developing separately in many Mediterranean ports, until (one of the) Catalan language competence(s) developed into a Pidgin (presumably in the 14 th century). By remaining, i.e. not vanishing, the Lingua Franca became known and adopted due to its popularity among the Arab merchants as well as presumably all travelling merchants that could not speak Catalan itself and therefore could not initially understand the Lingua Franca and speakers of other (Romance) languages that would also not be able to understand this Catalan-Arabic Pidgin. The Lingua Franca then gradually spread into other locations, i.e. major commercial centres, but also smaller ports and cities, replacing the previously acquired Catalan language competence by words and phrases, with pidginized Catalan (i.e. grammatical structures) becoming the Lingua Franca itself.

11.2 Spread among the local population

One very important aspect about the Austrian-Rumanian Pidgin referred to is that it was very useful to converse with Rumanian speakers but not useful at all to communicate with Austrian speakers. In fact, communication almost only worked one way and this aspect is very important when compared to the Lingua Franca. The Lingua Franca was ostensibly created under similar circumstances and thus the question arises who created and used it, and what it was created for. The only logical answer is that the Lingua Franca was created to be able to communicate with Romance speakers. Although it is said to be spoken by both Europeans and Arabic speakers, this seems to be a later, not an immediate, development that can only have been created by Arabic speakers, i.e. for the Arabic speakers to be able to converse with Romance speakers, as suggested by the Lingua Franca structure itself: Romance vocabulary combined with Arabic grammar. This situation can be compared to the Austrian-Rumanian Pidgin, which was definitely created by Austrian speakers to enable communication with Rumanian speakers, and therefore manifests Rumanian vocabulary and Austrian German grammar. Similarly, as has been put forward above, the Lingua Franca must have been created by Arabic speakers to be able to communicate with large numbers of Catalan speakers - hence (originally) Catalan vocabulary and Arabic grammar.

179 Because the Lingua Franca was created by Arabic speakers this might also be the reason why the Lingua Franca did indeed spread in larger cities in which large numbers of Catalan or (at least Romance) merchants were present as these two factors seemingly are directly proportional to each other, i.e. large-scale commerce and Lingua Franca competence. Communication with Romance speakers was definitely established and since commerce had reached a large scale there were many Romance speakers coming to these large commercial cities and thus the need to communicate was given. As Arabic speakers created the Lingua Franca and would also use and develop it, this Pidgin language was then, at some point, adopted by the local population of such commercial cities and could thus spread within the city. The Lingua Franca also became to be adopted by travelling merchants visiting the city, causing a further spread of the Lingua Franca into other locations. Variations occurred within different cities due to the cities´ individuality, both culturally as well as linguistically and because some cities were (commercially) dominated by Romance powers. The majority of European merchants would be Catalan and either Italian or Spanish, or French later on, all of which had a remarkable influence and outcome concerning the language used. Algiers is one of the few locations in which the local population is reported to have spoken the Lingua Franca, and as there are very few reports about the local population of other cities, this would make Algiers the cradle of the Lingua Franca. It is, unfortunately, not stated when this process originally was initiated. According to the sources, especially Haedo (1612), it would have been the 16 th century in which the adoption of the Lingua Franca by the local population was already a completed process, coinciding with the high peak of Spanish commerce. However, the question remains if this could also have happened earlier on, about two centuries before, which would coincide with the high peak of Catalan commerce. This assumption would suggest that the Lingua Franca was used with, and known among, the local population during the 14 th century. However, as this remains speculation and hardly any evidence that could prove this or be interpreted as such exists, it is only later that the first source states its existence. On due consideration, it would, however, seem unlikely to have happened in the 14 th century because if the local population actually had contact with the Lingua Franca through commerce then simply through the spread of the Catalan language and its importance the Lingua Franca would have been spoken in the whole Mediterranean including Europe and the local population of every city, as well as undergoing the development of becoming a Creole early on. However, instead, only three major cities are mentioned in the sources.

180 According to the documents and reports, the whole local population of cities such as Algiers (Haedo, 1612 quoted in Cifoletti, 2004: 197-202), Tunis (de Breves, 1628 quoted in Cifoletti, 2004: 202-203) and Tripoli (Histoire Chronologique Du Royaume de Tripoly, 1675 quoted in Cifoletti, 2004: 213) had acquired the Lingua Franca by the 16/17 th century. Contact, however, had existed previously, leaving open the possibility that knowledge of the Lingua Franca could have existed among smaller sections of the population before, presumably those that were more involved in trade with merchants. The possibility that this could have happened in the 14 th century lies in the increase of commerce, in which more and more people became involved in European-Arab contact, i.e. communication between individuals increased. This is (the only) one possibility as to how the Lingua Franca could have come in contact with the local population of these cities at this early time. In any case, this is unlikely and also highly speculative. European merchants, most likely Catalan, but it could also have been any other Europeans that spoke the Lingua Franca, could have established contact with the local population (commercially) which, however, only occurred in major commercial cities, not in small cities. The Lingua Franca came to be spoken only within these cities - not outside the city limits or in the countryside - by the local population. In small ports and cities where only merchants spoke the Lingua Franca it did not come to be adopted by the whole population, although by a smaller section of the local population to some degree. As seen above, the Lingua Franca was spoken in the major commercial cities (documented only in three, i.e. Algiers, Tunis and Tripoli), but is still absent or hardly present in reports about other cities. Unfortunately, these cities were never really specified, but these must have been smaller. Cities as Algiers, Tripoli and Tunis are always mentioned by name and at the time of the reports these were the most important cities in Mediterranean commerce in any case. The documents 28 about the Lingua Franca being spoken by the whole population in the (three) major cities intensively but hardly in other smaller cities could be interpreted to indicate that the Lingua Franca was not only created by, but also stayed among merchants and was used and spoken by them, and consequently sailors (Foltys, 1984: 7; Rossetti, 2005: 7 and Selbach, 2008: 34). This was primarily until large-scale commerce reconnected the Mediterranean, thus meaning that the Lingua Franca became common, known and spoken in the whole Mediterranean and was then subsequently adapted by (smaller) cities and their population. However, it is the lack of

28 Algiers is mentioned by Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Rehbinder, 1798 quoted in Cifoletti, 2004: 224-228; Tunis is mentioned by de Breves, 1628 quoted in Cifoletti, 2004: 202-203; Tripoli is mentioned in the Histoire Chronologique Du Royaume de Tripoly, 1675 quoted in Cifoletti, 2004: 213 181 documentation that serves as an indicator of another scenario, because if the local population of any port or city had spoken the Lingua Franca then there would be reports and sources stating this. Instead, only the local populations of the three North African cities (and only since the 16 th century) are ever explicitly mentioned as knowing the Lingua Franca. Otherwise it is only implied that a general knowledge of it was present in other cities of the Mediterranean, while some other reports about cities of that time do not mention the Lingua Franca at all. Curiously enough, one report unambiguously states that the Lingua Franca was known in every port of the Mediterranean especially on the coast of Tripoli and from the North African coast to the Levant (La Condamine, 1731 quoted in Cifoletti, 2004: 216), which leads to the conclusion that commerce alone did not lead to the Lingua Franca becoming adopted by, and integrated by, the local population. The Lingua Franca just never spread among the whole local population via the trade relations in (smaller) ports and commercial cities, although sections of the local population would have definitely known the Lingua Franca and can only have known it from merchants themselves. However, it could have happened that the local population of any city adapted the Lingua Franca and, arguably, the possibility of the Lingua Franca having been adopted by the local population would have existed during the 14/15 th century (where linguistically Catalan and the Lingua Franca matched); this being at the high point of commerce which made language contact between merchants inevitable and could have also favoured the Lingua Franca becoming spoken in the whole city. However, the question remains if this linguistic situation truly affected the local population to such a degree that it would have caused the adoption of the Lingua Franca. Realistically, it would seem unlikely that this actually happened in the 14 th century, because despite the dense frequency of commerce the local population itself would not have had contact to the Pidgin when trading, and definitely not the majority. It can be assumed that most people would have spoken some form of Arabic, which provides enough common understanding to conduct small deals. Furthermore, even if the local population had known or even used a Pidgin with merchants, this in itself would not be a guarantee for the adaption of such a Pidgin, i.e. a situation similar to Venice of the 18 th century in which the Lingua Franca was definitely known but presumably not spoken by the local population (Cifoletti, 2000: 14). Merchants, for instance, presumably used the expanded Pidgin state of the language while the local population only used its fixed Pidgin state (visible in the source texts). Therefore, commerce was not the reason for the adoption and spread among the local population. This

182 would also exclude the 14 th century in which Catalan commerce was at its peak, because a correlation between commerce and Lingua Franca adoption would have inevitably caused Lingua Franca competence in the local population (in any commercial city). Concluding, it is difficult to estimate at what point in time the local population became involved with the language. However, it seems as if the 16 th century (not earlier than the middle or end) was the time when the local populations became involved.

11.2.1 Spread via merchants

The local population adopted the Lingua Franca at some point in time and to further clarify if and how this happened, commerce and contact to merchants in regard to the local population needs to be analysed, i.e. if Catalan merchants had language contact to Arabic-speaking merchants or to the local Arab population itself. The possibility that Catalan merchants during any moment of commerce, even at the highest peak, would sell their wares directly to the population seems unlikely. Furthermore, even if Christians were allowed to trade, it would rather be unlikely for them to be allowed to directly trade within the cities, or only through specific licences or permits. Presumably not even in the cities such as Algiers, Tunis or Tripoli would Catalan merchants sell their merchandise directly to the local population. If the so-called merchant quarters included a sales stall as well is not clear and even if this was the case, rather trade partners and locals would have been hired in the shops. Presumably, local merchants probably bought the wares from European merchants and resold them to the local population. Merchants probably stayed and traded among themselves and did not have close and intensive contact with the local population. It would seem more plausible for merchants to establish contact among each others as trade could more easily and more effectively be conducted among merchants themselves. Trade with the local population would be cumbersome, negotiating with every single person for small parts of wares in an environment in which both participants would speak different mother tongues but possess no common language competence. It just seems more logical to sell wares all at once or in large parts, especially because Romance merchants probably did, and not spend a lot of time in the city itself or with the local population. That merchants only traded among each other and presumably Arab merchants selling their products to the local population also fits the development of the Lingua Franca. For the Lingua

183 Franca to have developed as a Creole language (of the local population), language contact between Catalan speakers and the Arabic speaking population must have existed. In fact, it must have predominated, as only close and interactive language contact could have facilitated the creation of a Creole. However, this contact did not seem to exist between these two groups before the 19 th century, and definitely not through commerce. The assumption that there was no intense (enough) contact between travelling or local merchants and the local population but instead among merchants themselves seems adequate as documents, i.e. the Geniza documents, about trade contracts in the 11 th and 12 th century show that commerce had been conducted among (Jewish) merchants over great distances, not in a market situation but among associates (Astren, 2014: 399-400). Therefore, the assumption that merchants would go to the market and trade for themselves and thus come into close contact with the local population seems to be unlikely. Concluding from the assumption that merchants remained among themselves, two further assumptions or even conclusions could be drawn. The first is that the Lingua Franca could not have become adopted by the local population through trade with travelling merchants, but must have been integrated into the consciousness of the local population through Arabic speaking merchants themselves. This also seems less likely. The Lingua Franca did not settle among the Romance population even though many merchants did speak the Lingua Franca and lived in locations beneficial for the spread of a Pidgin, as seen in the Cornwall report (Noall, 1809 quoted in Cifoletti, 2004: 220-221 and Coates, 1971 quoted in Collier, 1977: 284), in which both language contact and commerce were quite intense. Furthermore, contact between local (Arab) merchants and the local (Arab) population is very unlikely to have used the Lingua Franca, as communication among Arabic speakers would have been committed in Arabic. Even in unintelligible Arabic dialects simple communication for trade could be established. Concluding, ´Catalan merchant - Arabic population´ contact probably was not established, at least not to such an extent that could have facilitated an extended Pidgin. The second assumption would be that Judeo-Arabic would obtain its Lingua Franca vocabulary not through contact of the local merchants and their own people but through close and intensive contact of the Jewish population itself with the local population and consequently with the Lingua Franca itself. The Judeo-Arabic population, presumably after its expulsion and settlement in locations with a great density of Lingua Franca speakers, could have acquired the Lingua Franca through close contact with the local population. Even if the Judeo-Arabic population preferred to stay on their own, there would definitely have been occasions when contact with the

184 local population was necessary and this would have been the point when the integration of the Lingua Franca started. Furthermore, assuming that many Judeo-Arabic speakers would have also spoken Spanish, it would have been possible for them to adopt the Lingua Franca due to both, their knowledge of Spanish and Arabic. Fascinating as it is, this would also mean that Judeo- Arabic is one of the few (non-Pidgin) languages which integrated the Lingua Franca to such a great extent. Lingua Franca vocabulary might have occasionally become adopted by some languages, but nothing resembles the considerable presence of Lingua Franca in Judeo-Arabic (Corré, 2005). An alternative suggestion is offered by Guido Cifoletti (personal communication). He proclaims that the reason for so many similarities between the Lingua Franca and Judeo-Arabic, or to be more precise, between (all) languages of Jews having been expelled from the Iberian Peninsula, is because they themselves created the Lingua Franca. Consequently, not the local population would have shared the Lingua Franca with the Jewish population but the other way around. This is a truly valid point, as it remains difficult to explain how such a considerable number of Lingua Franca traits could have entered a fully fledged language. However, it seems (much more) logical for linguistic traits of Jewish languages to have survived in a Pidgin, and thus the origin through Jewish people seems most appropriate. Another perspective in which to view contact to the local population is that if contact between Catalan merchants and the Arab population existed since the beginning, then Arabic would have become the superstratum language of the Lingua Franca, as the majority of language speakers present in a Pidgin-forming situation defines the superstratum language. This is not necessarily linked to dominance, but is rather dependent on which language is approached and of which language more speakers are present. For instance, in a language situation where more Arabic speakers than Catalan speakers are involved, Arabic will become the superstratum language. The opposite will happen in a language situation where more Catalan than Arabic speakers are present, which then leads to Catalan as the superstratum language. In the specific situation where Catalan merchants would have had direct contact to the local population (of any North African port or city), Arabic would have been used by the majority as there would have been far more Arabic speakers (present in such a language situation) than Catalan speaking merchants, which would have led to the result of Arabic being the superstratum language of the Lingua Franca. The language situation would have been different insofar as Catalan merchants would have wanted to acquire Arabic words to be able to communicate with the Arabic population. However, as this

185 was not the case, contact presumably existed only among merchants themselves, which is also the reason why Catalan had become the superstratum language. In this kind of language situation a larger number of Catalan merchants would trade with a comparatively smaller number of Arab merchants. The lack of contact with the local population could also be interpreted as the reason why the Lingua Franca originated only later in time, centuries after the re-establishment of trade. After all, if contact to the local population existed, there would also have been an increase in exchange in communication which would have facilitated the creation of a Pidgin much earlier in time than the contact among a comparatively smaller number of merchants. It is unlikely that an adoption of the Lingua Franca by the local population would also not have led to an expansion of the Lingua Franca. Merchants’ use of the Lingua Franca might have been bound or restricted to commerce, but the local population’s use of the Lingua Franca would most definitely not be. There are many language situations and concepts that need to be expressed and require vocabulary to establish effective communication within the local population which are not found in or necessary for trade. After all, commerce as such does not require complex vocabulary, what is needed are numerals and basic vocabulary. However, this is not very effective for communication outside the field of commerce, as it is bound to fail simply because one cannot express every concept and the probability for ambiguity and different interpretations becomes exponentially too vast to express any intended statement clearly. If there had been such a development, the Lingua Franca would presumably have reached a Creole state very early in time. However, nothing of the sort occurred before the 19 th century. In fact, rather the opposite occurred (in the 16 th century), i.e. a downgrade of Lingua Franca competence.

11.2.2 Spread via slavery

Contact between Lingua Franca speakers and the local Arab population must have occurred in some way, as, curiously, there are also cities (e.g. Algiers) in which the Lingua Franca is constantly reported to be in use by the local population (Haedo, 1612). Other reports and documents about the Lingua Franca being spoken in other cities and especially the one report about it being used in every port but only by certain people, e.g. slavers, merchants or diplomats, would clearly indicate that contact among merchants existed aplenty, but hardly with or among the local population. Concerning the local population, contact situations between Romance and

186 Arabic speakers definitely did happen in all ports and cities of the Mediterranean including European cities, as seen with the example of Venice (Cifoletti, 2000), but explicitly so in North African port-cities. Thus it is fair to assume that the Lingua Franca not only originated in North Africa, most likely Algiers, with it being the most important commercial city, but was created through the interaction between merchants and stayed close to this class at all times. Therefore, the merchant class presumably created the Lingua Franca. The remaining question is how the Lingua Franca was adapted by the local population. It seems unlikely that the local population acquired the Lingua Franca from European merchants, as it is not clear if contact between these two demographic groups even existed or existed to the necessary point of Pidgin formation or adoptation. However, the local population must have had contact with the Lingua Franca and as merchants were the primary group of Lingua Franca speakers only the possibility of acquiring it from local Arab merchants remains. However, unlike the merchant class in which the Lingua Franca represented a defining aspect (of their work and lives), the local population did not have the same approach to the Lingua Franca and as a result, local merchants must have had greater Lingua Franca competence. The local population must have adopted the Lingua Franca once the need to learn it as a means of communication arose. Even though commerce had expanded to such a great extent in the commercial cities, it might not be the reason why the Lingua Franca became adopted by the local population, because for a whole demographic to adopt a language, even just a Pidgin, more than commerce is necessary to create the amount of (language) contact to initiate such a language adoption process. What remains unclear, however, is how in particular this adoption could have occurred if close language contact between local merchants and the local population did not exist. Presumably Lingua Franca language competence was easily accessible for everyone. Thus, to be specific, in the commercial cities communication between merchants and the local population existed only to the extent of occasional trade, which definitely served as a catalyst to acquire bits and pieces of the Lingua Franca but also did not allow the complete adaptation of the Lingua Franca by a separate demographic class. This would also explain the difference in Lingua Franca (state) competence. The adoption by the local population itself seems rather appropriate because if contact between the local population and merchants existed to a greater extent everywhere, then there would have been reports clearly stating this circumstance. The only defining force which could have accomplished the adoption of the Lingua Franca, by and integration into, the local Arab population from a historic, economic and political viewpoint

187 is slavery. The mentioned cities were not only the major commercial centres but also slave strongholds (Kahane & Tietze, 1958: 19), as slavery occupied a major part of commerce itself (Selbach, 2008: 33), especially during the 16-18 th century and became integrated even into mainstream society. Captives were forced into slavery where they could be kept as slaves up to five years and had to work in the city to be able to pay their ransom fee. Furthermore, slaves were not only the major trading good, at least in these major commercial cities, but they were to be found everywhere in the city as (forced) labour source for city building, working as water boys, being simple beggars in the streets, and slaves could even be acquired as house slaves where they were kept by locals. This made it possible for the Lingua Franca to spread within these cities (Selbach, 2008: 33). Presumably then, the Lingua Franca was integrated into the local population through slaves, which is the only other possible source which could have facilitated the spread among the local population. This was the only situation in which (close) contact between Romance speakers and the Arabic speaking population truly existed. It was because of slavery that quite a number of Romance speakers were present in the cities, even in the homes of Arabic speakers, and were forced to interact repeatedly with the local population. More importantly, slavery and the consequent presence of (European) slaves presented a reason for the local population to acquire and speak the Lingua Franca, as slaves were to be found everywhere in these commercial cities making communication inevitable. One of the most important points to note is that the local population did not learn the Lingua Franca from slaves but by contrast forced it upon them. The slaves themselves did not know the Lingua Franca in advance but acquired it through pirates, slavers, but most of all from the local population who would only speak Lingua Franca with them. With slaves being numerous and strongly present, it provided a good reason for the Lingua Franca to spread within these cities and among the local population. Therefore, the local population did not acquire the Lingua Franca from merchants, because if they did it would have been the expanded Pidgin state. However, instead through slavery and presumably the people involved in it (i.e. slavers and pirates), who did speak the Lingua Franca only spoke it in its fixed Pidgin state, as is documented in the source texts (de Bourdeilles, 1848 quoted in Cifoletti, 2004: 197 and Haedo, 1612 quoted in Cifoletti, 2004: 197-202 and Dan, 1637 quoted in Cifoletti, 2004: 203-206). One curious correlation is that both the Lingua Franca as well as the slaves themselves did not leave the city boundaries. Slaves were not allowed to leave the city, which could be a coincidence

188 or could be one reason why the Lingua Franca never left the perimeters of the city. Therefore, one could add slavery to the (natural) borders of the Lingua Franca. If one considers slavery (and its facets) in the spread of the Lingua Franca, the areas of slavery in which the Lingua Franca has spread do match up perfectly with the primary areas of (sea-based) commerce, i.e. large cities but never the countryside. One could further argue that even with slavery the Lingua Franca never left the ports and cities, as it truly was a language bound, defined and limited to commerce. Slavery itself only served as the infrastructure or the catalyst which made this spread permanent, but the acquisition of the Lingua Franca itself could only have happened because the Lingua Franca had already originated, developed and was broadly available. Furthermore, concluding from the fact that merchants, sailors, slavers and pirates (and diplomats) already were all Lingua Franca speakers and had their main seat (of business) in these major commercial cities, it seems obvious that it would be easy for the local population of these cities to quickly pick up this language and use it themselves once the necessity arose. After all, in the commercial cities, Lingua Franca competence must have been very common and readily available to everyone who wished to acquire it. Presumably, the spread into the local population would be a lengthy process. However, as slavery vastly increased at some point in the 16 th century, 1534 (Selbach, 2008: 33), it seems much more likely that this happened on a very rapid basis (and it is implausible that this should have happened in the 14 th century). After all, Lingua Franca competence, at least to a certain degree, existed in the commercial cities in the first place and the amount of slavery resulted in a very fast adaptation to this new type of language situation including the integration of this language into society (and social life). If slavery was not a variable in this equation, it would be doubtful that commerce alone would have led to an integration of the Lingua Franca into the daily life of the city. However, arguing that commerce consisted of slavery (at that time), it could be stated that it took a special kind of commerce to accomplish such an adoption, or to be more precise, the integration of slavery into commerce with the interrelation and mutual influencing of both commerce and slavery. This allowed the integration of a Pidgin language into the local population, as commerce provided the necessary basis, i.e. knowledge and popularity of the Lingua Franca, while slavery brought the Lingua Franca into the city itself and even into people´s homes.

189 11.2.3 19 th century spread and development

The development of the Lingua Franca in later years could be considered as an exception because through commerce alone the Lingua Franca would have remained in its extended Pidgin state. However, due to the Lingua Franca becoming spoken in commercial cities and thus coming into close contact with the local population, the possibility for the Lingua Franca to develop arose, although it took centuries until the Lingua Franca had reached this point. In the 19 th century the Lingua Franca presumably was in the midst of this very development, from an extended Pidgin into a Creole, but due to external circumstances it vanished during this process. As has been noted above, the reason for the (sudden) disappearance was the strong Frenchification and relexification process causing the Creole(-esque) Lingua Franca to become heavily Frenchified. At first, the Lingua Franca developed into a French variety, but then became replaced by a French Pidgin, which was itself influenced by the Lingua Franca and would later develop into a Creole. It is difficult to determine how extensive the connection truly was between the 19 th century Creole(-esque) Lingua Franca spoken by diplomats and state officials and the French Pidgin, often referred to as Sabir or later Petite and Grand Mauresque (Foltys, 1984: 30 and Schuchardt, 1979: 43) , spoken by the local population of Algiers. There were a few typical Lingua Franca traits present in the French Pidgin (Cifoletti, 2004: 258 and Castellanos, 2010: 4), although the French traits predominate, but none in the French Creole of the 20 th century. Possibly, this could be due to the process of depidginisation induced by French. The development from the extended Pidgin to a further Creole(-esque) state could only have happened because the Lingua Franca spread onto the local population. The reason why the Lingua Franca remained in its development state for centuries was because neither commerce nor slavery and not even the local population ever saw the necessity to develop the Lingua Franca beyond the expanded Pidgin state. However, the Lingua Franca was present in diplomacy of the 19th century, which was the initiation point for the Creole(-esque) state that is presented in the sources of that time. Furthermore, the Lingua Franca finally reached such a level of awareness and importance among the local population of Algiers in the middle of the 19 th century, due to the opening up towards European (i.e. French) culture, that a creolization process finally occurred, with the result of the creolization and (supposedly) depidginisation of the Lingua Franca. A further conclusion that could be drawn from this assumption is that the Lingua Franca also entered the Judeo-Arabic language in later decades in this very manner. Judeo-Arabic language contact to the local population could be the reason for this strong integration of Lingua Franca

190 vocabulary into Judeo-Arabic. Viewed from this angle this could only have happened in the late 18/19 th century when the Lingua Franca became omnipresent and spoken by more than just one group within the local population. Further judging from this assumption, Judeo-Arabic would also have gotten its Lingua Franca input due to the Lingua Franca gaining popularity among the local population in the major commercial cities. However, the thesis formulated by Guido Cifoletti (personal communication) of Iberian Jewish people creating the Lingua Franca after their expulsion is much more substantial.

11.3 Geo-political analysis of the Lingua Franca spread

One important aspect about the spread of the Lingua Franca, and Pidgins in general, is that Pidgins tend to remain, especially during the process of creation, in the territory, or to be more precise in the geographical and social environment in which the source languages are present and in contact with each other. Linguistically, such an area tends not to be strictly limited to geographical areas, but instead to contact situations that facilitate the same conditions of language contact, e.g. Catalan-Arabic (language) contact. This is not restricted to one specific location but to an unspecific larger area in which language contact occurs including several locations, as seen with the (rare case of the) Lingua Franca. Therefore, Pidgin formation can also be attributed to the social environment. Pidgins can expand in the sense that any (even an expanding) geographical area is suitable as long as the same language contact takes place under the same conditions. In other words, if the geographical area, in which input from the source languages is constant, expands then the area of language contact does as well. For instance, Catalan and Arab merchants coming into close language contact in different locations, with Catalan being the vocabulary superstratum source language and Arabic being the grammatical input substratum source language, should produce the same result: a language most similar (some words may differ) to any Pidgin from another city in a different geographical area formed under the same conditions. Curiously, almost the same Pidgin could arise through the same conditions without ever establishing contact between these two Pidgins and they could be joined together if contact between them was established, which is possible through the very basic grammatical structure of Pidgins. This can not have been the case in the Mediterranean because the network within it was by much integrated and connected, which established a language exchange between the cities.

191 It is important to note that a ´language border´ rather than a ´political border´ is being described here, as a language border registers even separate language development that would not exist outside of political borders. With the Lingua Franca being created outside of the Catalan Empire, a political border would not register the language development of the Lingua Franca in North Africa, only if the Lingua Franca would sustain itself within the Catalan Empire. The advantage of language borders is that for instance commercial centres can be registered in which Catalan was used although they were not part of the Catalan Empire. This creates a completely different and broader picture of the spread of the Lingua Franca, as it allows an impression of the spread of Catalan correlating with the spread of the Lingua Franca within the Mediterranean basin (i.e. a so-called geo-politcal analysis). The (geographical) expansion of the Lingua Franca itself can be explained geo-politically as well as linguistically. While linguistically seen, it seems very unusual for a Pidgin to be spoken in several distinct places, it still recognises that the Lingua Franca was spoken within the whole Mediterranean basin, although it may not be able to explain why exactly the lingua Franca was limited to ports, cities and ships. Geo-politically, however, this coincides with the political and geographical territory of both the Arab Empire as well as the Catalan kingdom, due to being the result of language contact between these two political entities. More specifically, due to commercial relations between these two political systems, language contact occurred in a location in which the political area of one political system overlapped with the commercial network of the other political system. As a geo-political analysis can also include different languages and the basis of the Lingua Franca origin, spread and development is language contact between Arabic and Catalan speakers, certain insights should be gained. Arabic was spoken along the whole North African coast and Catalan was spoken in the Crown of Aragon and its colonies, which was a (comparatively) small geographical area with its main seat on the north-western Mediterranean shore. However, language contact itself took place on the North African shore through Catalan merchants travelling to North Africa to trade with Arabic speakers, although not along every part of the Arab Empire, but limited to ports and cities. This not only coincides with the reports about where the Lingua Franca was spotted and spoken, i.e. in ports and on ships, by merchants and sailors, but also explains its diffused and seemingly random expansion of never occupying any area outside a city or any larger land mass in general. The basis of Catalan-Arabic language contact that created the Lingua Franca was commerce-

192 based and occurred only at ports and cities along the Catalan merchants’ sea routes. In this respect, a geo-political analysis does explain the unusual spread or indirect expansion of the Lingua Franca which is small in regard to the geographically vast area of the Mediterranean, but also small in contrast to the vast territorial expansion of empires, kingdoms or colonies of other nations, i.e. the Crown of Aragon, Spain, France etc. Concluding, the geographical area of language contact does tend to be limited. The potential of expanding this area (although usually not onto other areas) exists if both source languages are spoken and contact between the source languages is either constant or even increases. Once contact exists to a certain (i.e. greater) degree, a Pidgin is likely to originate. If the same language contact is maintained, the created Pidgin can sustain and expand both linguistically and geographically, i.e. in the sense of acquiring more speakers in the same location or, if this option is available, in other geographically distinct locations providing the same language contact exists (as well as a link between these locations).

11.3.1 Language border expansion

After some time, if the Pidgin exists long enough to become either a fixed or an extended Pidgin, it is easier for a Pidgin to expand beyond its original (language, social, political) borders. This is one explanation why the Lingua Franca could not only expand on the North African coast among Arabic speakers but (later) also with other Romance speakers. A further beneficial factor for the spread outside its original area of origin, especially in Romance speaking areas, is that of the similarities between the Romance languages and the Lingua Franca superstratum language being a Romance language itself. Another important aspect that further created the possibility for the Lingua Franca to expand beyond its original place of origin is the considerable flexibility of Pidgin languages. The lexis of a Pidgin can shift at ease allowing other language speakers to adapt the Lingua Franca and shape it towards their own language system without losing communication quality, i.e. intelligibility. Even shifting Lingua Franca vocabulary with Spanish or French words is irrelevant for Lingua Franca intelligibility (especially when the shifted words of the Romance languages are similar to each other). One last aspect which catalysed the (geographic) expansion of the Lingua Franca was created through a circumstance of commerce, i.e. connecting different ports and cities across the

193 Mediterranean through travelling merchants and thus facilitating language contact. The opening up of many commercial cities within the Mediterranean allowed the Lingua Franca to expand outside of its place of origin, western North Africa (most likely Algiers), as commerce created language situations that were similar, if not identical to the language situation in which the Lingua Franca originated.

11.3.2 Expansion into cities

Curiously, there is a correlation between the Lingua Franca being used within city limits, but not within the countryside of North Africa, and Catalan. Catalan was definitely known and probably even spoken by some in the cities, but there is no evidence that Catalan or Romance languages in general were even remotely known in the countryside. The reason for this circumstance might be that Catalan merchants only travelled by sea (and not by land like many Arab merchants). As Catalan (or any Romance language) was not spoken outside the city borders, consequently the Lingua Franca could not establish itself in this specific area. This could only have happened if the Lingua Franca became a separate (Creole) language that was spoken exclusively by Arabic speakers. For instance, the 20 th century (French) Creole might have been used even outside the city borders, presumably because French became spoken in the interior due to the political interest of France in the Algerian inland as “the prime interest of France increasingly became the vast hinterland behind the coast of the Maghreb” (Abulafia, 2014: 147). However, this would only have occurred in the mid 19 th century with the French Pidgin and in the 20 th century with the French Creole. The Lingua Franca itself only settled in commercial centres in the Southern Mediterranean, all of which had a port, and only three of them are reported to be centres of Lingua Franca communication i.e. established entities. Cities in the Northern Mediterranean are only indicated to have Lingua Franca speakers, but it seems these were not the local population but Lingua Franca speakers (merchants) from the Southern Mediterranean (Cifoletti, 2000: 14). ´Settled´ in this case actually means that the Lingua Franca was used in small and large ports (in the same way) but with different intensity due to large cities having much more commercial frequency. Merchants conducted business in both small and large cities, but language contact definitively was greater in large cities. However, commerce, despite the considerable amount of language contact it established, never caused the Lingua Franca to settle anywhere, only facilitated its

194 constant use. Instead, slavery or the constant presence of slaves supported the Lingua Franca in being continuously spoken by merchants, diplomats, pirates, slavers, even slaves themselves and the local population.

11.3.3 Expansion into ships

The Lingua Franca was spoken and used on ships by merchants, pirates and slavers, was then adapted by sailors, and thus the Lingua Franca became not only a trade language but also a sailors’ jargon. This language expansion or spread was possible due to Catalan being the official language of commerce and the official language on Catalan ships, although Catalan was presumably also used on other European ships and even on ships by the Ottomans. Most curiously, the Lingua Franca was spoken by the (Ottoman) Minister of Marine (Broughton, 1839 quoted in Cifoletti, 2004: 239). The Lingua Franca, however, was primarily a trade language not a sailors’ jargon, although Foltys (1984) mentions the “starke Verbreitung der LF [Lingua Franca] unter Seeleuten und in Hafenstädten des gesammten Mittelmeeres” (1984: 7), which is why there are reports about the Lingua Franca being spoken by sailors and, consequently pirates, slavers and merchants. Although sailors are not mentioned in the earlier documents (and centuries), this most likely changed ´during the 16 th century with piracy becoming the main business in the Mediterranean´ (Selbach, 2008: 34). What can easily be overlooked is that while Catalan merchants may have used Catalan to communicate with their crew (i.e. sailors), other travelling merchants could have used the Lingua Franca. Thus the groups that carried the Lingua Franca across the Mediterranean Sea were travelling merchants including sailors. An interesting fact is that travelling merchants were also mariners (Chaunu, 1979: 261 and Wansbrough, 1996: 6). Therefore, the assumption seems obvious for the Lingua Franca to become spoken among sailors. However, this development must have occurred when commerce was at a high peak and conducted not by single persons but by large groups with many people involved. Consequently, this caused the Lingua Franca to become spoken by a majority of sailors (although presumably in a lower lectal state). Gertwagen (2014) for instance mentions that the average size of a ship’s crew was 150 in the 14 th century, which only grew in the following centuries with the increasing carrying capacity and size of ships (2014: 161-167). However, the average size of crews did not drastically increase as the Cornwall report of 1769 states ´ 220 men were on board the pirate ship that arrived at Cornwall´ (Coates,

195 1971 quoted in Cifoletti, 2004: 220). As Catalan had become the official language on ships, i.e. a lingua franca, it seems plausible that this occurred not only on ships of the Catalan kingdom and Catalan merchants but also on other ships, similar to English as a trade language today. Presumably, sailors would have to speak Catalan or a pidginized form of it to be hired on ships. However, it would seem implausible that Catalan itself would be acquired, which is probably even true for most European sailors, due to the intelligibility of Catalan and the Lingua Franca and the considerably higher effort necessary to acquire Catalan (than the Lingua Franca). Fortunately, the Lingua Franca was quite close to Catalan but easier and thus quicker to learn for both Romance and non-Romance speakers, and hence was quite likely accepted on a Catalan speaking ship. Sailors, therefore, would most likely have spoken the Lingua Franca.

11.3.4 Conclusion

The spread of the Lingua Franca happened on two separate but parallel occasions: the expansion in fixed geographical locations such as commercial centres and ports, and at the same time the expansion on ships (i.e. at sea). The date since which sailors were using the Lingua Franca could be any century, although from what is known from documents it can only be deduced that the Lingua Franca was already established among sailors by the 16 th century through piracy and the internationality of pirate crews. Even though all of this could have happened earlier, the fact that the extended Lingua Franca is not known to be spoken by sailors makes it more likely that this would have been a rapid process that occurred in the 16 th century, as the acquisition of a fixed Pidgin (which was spoken by sailors) takes less time. Unfortunately, all that is definitely known is that the (fixed Pidgin) Lingua Franca had been established as a sailors’ language by the 16 th century.

11.4 Specifics of Lingua Franca spread

The historical event that led to the creation and (the chain of events of) the development and spread of the Lingua Franca presumably was the peak of commercial relations between the Crown of Aragon and North Africa, which occurred during the 14 th century and constituted a constant process until the 15 th century. The continuous commercial relations led to the spread of

196 the Lingua Franca onto all locations connected by Catalan commerce. What made it possible for the Lingua Franca to spread across the whole Mediterranean was that Catalan was a widely spread language and a Catalan Pidgin could be understood all across the Mediterranean. The Lingua Franca could have remained where it originated but instead, through the interconnectedness of Mediterranean commerce, it spread across the whole Mediterranean. This could have been either a consistent process and be directly related and proportional to the spread of both commerce and the Catalan language (as these were both intertwined) or a direct consequence of the establishment of Catalan commerce within the Mediterranean and therefore occuring after the peak of the Crown of Aragon. Although it is possible that the Lingua Franca spread parallel, and therefore slowly, with both commerce and Catalan to the rest of the Mediterranean, it would seem more logical that the Lingua Franca would originate and spread only once Catalan and commerce were established in the Mediterranean. However, even though the Lingua Franca presumably originated much later when commerce was already established, it must have spread almost immediately to all locations that were heavily involved in commercial activity in the Mediterranean commercial network. After all, these established locations with abundant language contact and human interaction would provide the perfect conditions for a Pidgin to (quickly) originate, develop and spread.

11.4.1 Specific place of origin

From a linguistic point of view, the Lingua Franca could have originated, developed and spread at, from and to every port and every city in the Mediterranean to which a trading route was accessible. However, if one were to define one specific location in which the Lingua Franca developed first, it must have been a location in which both commerce accumulated and language contact intensified to the point where a Pidgin could originate geographically, economically as well as linguistically (and thus historically). The most plausible location for the origin of the Lingua Franca would be a very western location on the Southern Mediterranean (judging from what has been established up to this point). If the Lingua Franca originated in the western part of the North African shore, then it could not have been any small port but instead a major commercial city to which Catalan merchants would venture regularly, in which plenty of commerce and language contact occurred and presumably during a time when Catalan dominated the Mediterranean basin. The whole North African coast

197 may have been incorporated into the Catalan trading network but presumably the western part had the most intensive commercial activity concerning Catalan merchants, which diminished the further east one went. Consequently, the most plausible assumption would be that the Lingua Franca originated in the western part of the North African coast and then spread eastwards from there (a port or city at a time). This whole process of spread would presumably have been accomplished within decades if the Lingua Franca originated during the peak of commerce when these cities had already been integrated into the trade routes of merchants. The specific location of origin itself is impossible to determine for certain. However, judging from linguistic specifics, information and indications from the source texts and documents, as well as economic and political reasons (that provide additional evidence), the origin of the Lingua Franca can only have occurred in one of the major commercial cities of that time, with Algiers being the most plausible location (as already indicated above).

11.4.2 The importance of Algiers (as a specific location of origin)

As to why Algiers had become such a central point for the Lingua Franca is most likely due to the eastward (commercial) expansion of the Catalan kingdom, with Algiers being the first commercial cornerstone that became opened up for European merchants. Merchants ventured eastwards and reached Algiers as the first major commercial city (sooner than Tunis or Tripoli). This seems plausible for several reasons, e.g. trade routes had yet to be established and thus the ports in the vicinity of the Crown of Aragon were the main destination. Long term sailing was difficult if not impossible at the time, as many inventions that allowed long term travel had yet to be invented or were not a standard, i.e. compass, portolan etc. (Chaunu, 1979: 84). Thus only a slow progression into the eastern Mediterranean was possible and merchants just started to venture out, which means that this was unknown territory and knowledge about currents, winds and ports was missing, and thus slowed down the eastward progress. With Algiers being the first commercial city in the eastward expansion, there clearly is a temporal advantage present. The reason why Algiers must have been the first major city in which this applies is not because it was the very first city that merchants ventured to, which is rather unlikely as there are many other (smaller) ports between Barcelona and Algiers on the North African shore, but because it was the city in which large-scale commerce definitely occurred. Algiers could have been the first place to create a Catalan-Arabic Pidgin, and the documents

198 provided by Haedo (1612) strongly suggest this impression. Algiers has always been strongly connected to the Lingua Franca and also seems to be the city the Lingua Franca was most accepted in. This might due to the strong ties to Catalan, Spanish and later French, i.e. European languages, or perhaps due to the (political-economic) importance of this location itself. The descriptions of Haedo (1612) suggest that the Lingua Franca was spoken everywhere in Algiers and by everyone, but also as having been established to an extent which would allow the assumption that this could have been an earlier development. According to Cifoletti (2004: 197), Haedo was writing during the 16 th century. This could easily mean that the Lingua Franca could have been in Algiers since the beginning of the 16 th century and would thus correspond to the earliest report on the Lingua Franca (Haedo, 1612 quoted in Cifoletti, 2004: 197-202). Presumably, this also was due to piracy and slavery being established by the Ottomans in the 15 th and 16 th century respectively; corsairdom started around the mid to late 15 th century to gain prominence in the Ottoman Empire after the latter were building up their naval forces since the mid 15 th century, and slavery was also increasing early on in the 16 th century with the Ottomans establishing themselves in Algeria (Kahane & Tietze, 1958: 12, 18-19).

11.5 Spread into other locations

The establishment of Catalan-Arabic contact and the increase of Catalan language competence in the location of origin is no exception to any other commercial city, although one major difference occurred, the development of a stable Pidgin. The increase of contact facilitated by commerce, which in turn led to increasing language links occurring in massive numbers consistently, was how the Lingua Franca presumably originated. Simultaneously to the Lingua Franca being established as a Pidgin, commerce itself had reached a large scale in the middle of the 14 th (to 15th ) century which could have caused the Pidgin to quickly develop into an extended Pidgin. Furthermore, an additionally large number of (non-Catalan) merchants became involved in trade, and the majority of these travelling merchants adapted the expanded Pidgin and carried it into the Mediterranean. Furthermore, these merchants presumably did not face the same linguistic problems of being unable to communicate in a mutual language like travelling merchants in the early centuries. By the 14 th century the Mediterranean had been opened up by the Crown of Aragon (but also by the Italian city states and Spain) and thus the Catalan language, its pidginized form and also Catalan

199 language competence were definitely present when the Lingua Franca became used (en masse) by non-Catalan travelling merchants. Once the Lingua Franca was established and available for a broader spectrum of merchants another major development in contrast to earlier times appeared. Both Catalan and Lingua Franca input became present in several locations which contributed to the much quicker development of the Lingua Franca (itself) and its establishment in other locations than in its initial place of origin. Not only would there have been Catalan merchants involved ensuring input for the Arabic speakers but other travelling merchants were soon following as well that were presumably speaking the Lingua Franca. This provided further input for the Arabic speakers meaning that both the development of the Lingua Franca and its spread into other locations in the Mediterranean (for the main part eastwards of Algiers) were catalysed (presumably also in the major commercial cities at first). Concluding, it can be stated that the Lingua Franca did not only remain in its location of origin but its spread immediately continued through commerce being carried to other locations by travelling merchants and thus spreading also in many more smaller ports and cities. Catalan commerce had reached the Mediterranean as a whole on a larger scale and within these cities an already developed Pidgin was able to develop far more quickly because the antecedent conditions necessary to form Catalan language competence and large-scale commerce were already established. Thus the constant and intensive Catalan as well as the Lingua Franca input provided by merchants sustained further.

11.5.1 Parallel spread

At this point it could be argued that even though the assumption of a simultaneous spread would seem logical linguistically, due to prevailing Catalan input and already established Catalan language competence this most likely was not the case, as due to the spatial distance alone a temporal shift must have occurred. Assuming that the Lingua Franca had one place of origin and spread from there to the next location, there must have been a delay by the distance between these locations alone, unless this really was a parallel process where the same language contact situation and therefore the same pidginization process occurred in all locations at the same time, i.e. Algiers, Tripoli and Tunis. This seems unlikely as such a process always depends on several individual factors bound to every city, thus eliminating a parallel aspect to this development process.

200 Travel at those times did take time and for a merchant to go to the next major commercial city could take weeks or, if the weather was not suitable for travel, months to reach the next (major) destination by ship. Wansbrough (1996) mentions how long it took to travel by ship by stating that travel ´from Venice to Alexandria took about 89 days, to Istanbul about 81 days and to Damascus the journey was about 102 days´ (1996: 9). The temporal difference in spread can be analyzed not only because of the geographical distance and the time it took to reach another location, but because the Lingua Franca could only reach these other destinations through travelling merchants that were themselves bound to (already established) commercial routes. Therefore the Lingua Franca could only establish itself in the ports and cities that were connected to each other. Furthermore, even if a location had been made accessible through a trade route, it took time before commerce and contact language between Catalan and Arab merchants was fully established for the Lingua Franca to establish itself in a location. The Lingua Franca served as a means of communication, no doubt, but would not be adopted in the smaller ports and cities as there was no need for it. Local merchants might have heard of the Lingua Franca, but without large-scale commerce there was no consistent or close contact with the Lingua Franca and thus even a Pidgin would not be adopted (in a new environment), possibly only partially integrated. After all, it was not the first merchants that came to a location who immediately created or established the Lingua Franca but it took hundreds and thousands of interactions to establish even the necessity for the creation of a Pidgin. For instance, in the Austrian-Rumanian field study, only after 50 people per day were constantly arriving and interacting (in Rumanian) did one start to even consider using Rumanian, but once this was established the Pidgin process started and the adoptation process happened quickly. With large- scale commerce, a constant supply of Catalan traders ventured to the cities, interaction became consistent and regular, repetition followed automatically, and with it the Lingua Franca could establish itself in a new location. At this point it is necessary to clarify (the speculation) if the spread and especially the establishment of the Lingua Franca in other commercial cities (Tunis and Tripoli) could be reached in far less time than in the original location (Algiers). Presumably it takes longer to create a Pidgin language with only (source) language input than to establish an existing Pidgin through (source) language as well as pidginized (source) language input. While the Lingua Franca became more established in one place, travelling merchants would

201 acquire and carry it to the next city. Once in another location, the development of the Lingua Franca started from a different point, i.e. on the basis of an (already) existing Catalan language competence (and thus not anew but) with both Catalan and Lingua Franca input. This definitely decreased the time necessary to establish a Pidgin in a location different from the original. However, the Lingua Franca still took time to settle and develop itself in a city, meaning that the travelling merchants had to communicate with people that yet had to learn the Lingua Franca. Concluding from this assumption, there would have been a short time shift of a few years between the origin and the spread of the Lingua Franca as well as a difference of about one decade for an extended Pidgin to establish itself in another major commercial city.

11.6 Linguistic analysis and interpretation of spread

A very dynamic system of (gaining new words and) stabilizing words was created through the constant (language) input and exchange on both sides, i.e. travelling merchants for vocabulary and local merchants for grammar and stability. Furthermore, without this constant language input the Pidgin could not have developed beyond even the fixed Pidgin stage (or at all), which was definitely the case in smaller cities but even in major commercial cities it only reached the expanded Pidgin state. The explanation for this phenomenon is that it does require a certain degree of (established) language contact for a Pidgin to develop further, which could only be found in Algiers, Tunis and Tripoli. The economic development of the three major commercial cities, which were massively involved in the Mediterranean economy due to commerce concentrating in these cities, is one major indicator for not only the economic relevance of these cities, but also for the linguistic importance of the Lingua Franca varieties which were developing in these locations. These were the main centres of Lingua Franca development and stability, in the sense that as long as the Lingua Franca was spoken in these locations it would remain (linguistically) stable and not vanish from the rest of the Mediterranean. The linguistic development of the Lingua Franca varieties crystallizing in the Mediterranean could have had a tendency to develop with more prestige due to their geographical location and commercial importance. Therefore, it is possible that these three Lingua Franca varieties had a much greater linguistic influence and importance due to their economic relevance. Furthermore, merchants probably adapted a variety from Algiers, Tunis or Tripoli, which might also be the

202 reason why these different varieties survived. However, if there were other important commercial cities in which Lingua Franca varieties developed separately is up for speculation. While it is conceivable that these varieties existed, e.g. in Amalfia or Fez, they could have vanished once their economic importance or political relevance disappeared and trade with these locations declined. Therefore, if such varieties existed, they presumably lost all their influence and either completely disappeared or were preserved in the other Lingua Franca varieties.

11.6.1 First language contact

Linguistically, ever since the first language contact between Catalan and Arabic speakers, there existed the problem of being confronted with a different language, i.e. (the reality of) language contact and depending on the intensity of contact, from sporadic to small scale to large scale, and different outcomes were present. While sporadic (but consistent) language contact created the learning of (a limited) vocabulary and phrases, small scale contact created the further extension of Catalan language competence through learning more vocabulary and phrases, i.e. extending or further developing Catalan language competence. Large scale contact, however, created a Pidgin language, the Lingua Franca, and continuous language contact (on a large scale) created the further development of the Lingua Franca into an extended Pidgin. The linguistic situation, was always the same, finding a common means of communication suitable for the intensity of contact with less language contact rather leading to the adoptation of words but more intensive language contact leading to the development of a Pidgin. The state of a Pidgin could thus be described as an adaption process depending on the intensity of language contact. A Pidgin language fulfills a pensum (or quota) of (required) expressions for different linguistic demands. While close contact between speakers of different languages creates difficulties in communication, it also enables solutions as it supports language contact which facilitates finding a common means of communication, as long as the intention of (successfully) communicating is mutual. In terms of historical linguistics, Catalan speakers came to Arabic speakers and communication was established (through the mutual interest in communicating and trading) with Catalan speakers using their language and Arabic speakers learning a few words from it at first, then acquiring more words and phrases and thus increasing their Catalan language competence. Finally, they rearranged the Catalan vocabulary (i.e. different words) via their own grammatical system, which is the origin of the Lingua Franca. This is clearly visible in the Lingua Franca

203 itself as the vocabulary is predominantly Catalan, at least Romance, and the syntax (e.g. inflection) is Arabic, i.e. /ir/, especially the /i/ replacing the /e/ is said to be specifically Arabic (Schuchardt, 1979: 34 and Collier, 1977: 291), although the third conjugation of Old Catalan is also based in /ir/ (Monk, 2013: 215). However, the second conjugation is /er/ (Monk, 2013: 215), which presumably is where the Arabic influence becomes identifiable, i.e. Catalan saber becoming Lingua Franca sabir .29 Furthermore, through the field study of the Austrian-Rumanian Pidgin, it can be stated that people who travel to a new location and want to communicate use their own language if they do not know the local language, while the local people rather adapt to the new vocabulary by learning new words and applying or embedding them within their own grammar; that is, of course, if a mutual intention to communicate exists as there are cases of people reacting negatively or not reacting at all to people of a different language trying to communicate. This may be a curious case, but not illogical when considering that one will interact and respond according to their own grammar when being confronted (but also when being motivated to interact) with another language. Curiously, grammar is often assumed to be similar to the mother tongue, or rather not considered at all, as vocabulary and its meaning is of primary concern. As communication continues people will still learn words of the other language but use them as if learning new vocabulary for their own first language. Applied to the situation of the Lingua Franca, this linguistic phenomenon, or problem depending on viewpoint, occurred with every new situation of language contact in every new location when speakers of two different languages concurred and communicated. This pattern of communication became consistent: i.e. travelling merchants were venturing to several spatially (and temporally) distinct locations and establishing Catalan-Arabic language contact with different Arabic speakers. Once (Catalan-Arabic) language contact had been established, the linguistic process of developing (Catalan) language competence started, and was followed by the process of creating (and establishing) a Pidgin, which presumably happened later once language contact was at a level that required a higher level of communication. The process of developing Catalan language competence was the result of Catalan merchants using Catalan with Arabic speakers in locations that already had established Catalan-Arabic contact. The outcome of Arab merchants receiving more Catalan input was closeness to Catalan

29 Evidence for the grammatical influence of Arabic can be found in Schuchardt, 1979: 27 and Cifoletti, 2004: 36. Both Schuchardt, 1979 and Cifoletti, 2004 extensively show the considerable influence of Arabic on the Lingua Franca in their works 204 itself, which influenced the effectiveness of communication and probably even dictated the course of development of the Lingua Franca into a Pidgin strongly similar to Catalan. After several such language situations of Catalan speakers communicating and thus providing linguistic input, with different outcomes depending on the intensity of contact and commerce, Arabic speakers most definitely developed a Catalan repertoire as they adopted (parts of) the Catalan language. At some point a Catalan language competence had become established in the cities (and to a lesser degree in the small ports). Eventually, this process led to the creation of a Pidgin.

11.6.2 Further linguistic development

Once a Pidgin originates, it evolves to a certain state and then stops if (or because) its communication needs are satisfied. A Pidgin can then evolve beyond that particular state, but only if the need for a higher level of communication arises. Pidgins themselves are by definition limited languages, limited in the sense that they are contact languages which are restricted to whatever facilitates (this particular) language contact (i.e. the entire reason this Pidgin exists). The most important aspect is that language contact itself does not immediately create a Pidgin. Such a process only starts once a certain degree of language contact is reached and constant repetition of this recurring pattern over a (certain and) constant period of time is ensured, meaning that as long as contact remains sporadic and irregular there will not occur a development into a (fixed) Pidgin. The same applies to the development into further Pidgin states. Any fixed or expanded Pidgin will remain in this state as long as the necessary requirements to initiate further development are not met, for instance through an increase in language contact or a demand to converse on a mesolect level. Assuming that the Lingua Franca was created through commerce, depending on which Pidgin state commerce allowed to (create and) develop is also the state which the Lingua Franca reached and then remained in (until the need for a further state arose). Curiously, judging from the earliest source texts this would be the expanded Pidgin state. The communication need of small-scale commerce does not require a Pidgin to be fulfilled, because communication would not be crucial and could be satisfied with the acquisition of vocabulary. However, as soon as commerce reaches a large scale, with growing demand and commercial relations, a great number of people will become involved and a means of communication will become unavoidable and inevitably be created. Therefore, contact and communication that were intense enough to provide the requirements for

205 the Lingua Franca to reach an extended Pidgin state would have been found in large-scale commerce. However, outside of commerce there was neither the need for the existence of this language nor the intention of further developing the Pidgin. Continuing this assumption, it would also seem rather logical that the Lingua Franca would originate once large-scale commercial relations were established, develop with the quickly growing quantity of language contact and thus reach its final state within a short amount of time. The most probable assumption is that large-scale commerce created an extended Pidgin state because (even large-scale) commerce itself only requires an extended repertoire of fixed expressions for predetermined language situations to successfully establish communication. However, at the same time variation should not occur to provide stability for the expressed statements. In other words, only a selected amount of vocabulary is truly required to gain the desired effect of communication in the context of commerce. Grammar, on the other hand, needs to be expressive enough to allow meaning to be comprehensible (e.g. for commerce to function on a certain competence level), but at the same time needs to be flexible enough (i.e. not bound to complex source language grammar) to allow for a certain quota of (grammatical) mistakes without rendering meaning unintelligible. Flexibility of grammar must be seen in a more general sense: instead of fixed grammatical rules there are only general indications as to how a particular meaning can be expressed. However, variations or mistakes (from a fully fledged language perspective) can occur without breaking the flow of understanding. As long as there are some (fixed) structural reference points, e.g. present tense, infinitive, prepositions, in which this language operates, meaning can be established.

11.6.3 Linguistic re-definition of Lingua Franca origin

The origin of the Lingua Franca may also be re-evaluated and re-defined. Origin, in this case, is not the immediate emergence of one new language but many interconnected steps with the common denominator of constant (Catalan) language input resulting in the visible emergence of (Catalan) language competence, which was the basis for the origin of one common Pidgin. In this perhaps exceptional case, the origin was a process that occurred over time, i.e. establishing language contact, expanding language competence, and continuation of language contact until an extended Pidgin, i.e. the Lingua Franca, was finally visible to a broader public. The origin of the Lingua Franca may lie in the late 16 th century, but technically it is (the final result of) a process

206 which only became noticeable in the form of a (developed) means of communication. This whole process of individually establishing and developing language contact and competence in several locations through continuous Romance input, which would lead to a commonly spoken Romance-Arabic Pidgin (including its varieties), could theoretically be called the true origin. One might define origin in this case not as the (sudden) appearance of one (expanded) Pidgin but as a complex process including the interplay between separate factors until an intelligible result became visible. Lingua Franca origin is one major aspect in Lingua Franca studies. Common definitions may not be accurate in this, admittingly, very special case of Pidgin and Creole language origin. The Lingua Franca origin itself may have been in one specific location, but its development and spread can not have been restricted to one geographical location, i.e. one major commercial city. Instead, as the development of commerce suggests, it must have been bound to a (trading) network. Locations such as ports and cities which were connected by the (then) established trading routes of commerce laid the foundation for the origin of the Lingua Franca. Therefore, one could argue that the Lingua Franca originated because language contact of local Arab merchants with travelling merchants in multiple and separate locations in a very wide area of the Mediterranean provided the basis for its origin.

207 Chapter 12: Lingua Franca stability

Quite unusual for a Pidgin language was the stability present in Lingua Franca development. Rossetti (2005) comments the “continuum” of the Lingua Franca stating that it “shows a remarkable level of consistency” (2005: 17) throughout its existence. There were shifts in vocabulary and grammar but astonishingly many words and a considerable amount of grammatical structure from the original variety still exist and are visible in the relexified forms 30 . Due to the innate flexibility of Pidgin languages with regard to vocabulary and only minor changes occurring through the shift to other (Romance) language systems, intelligibility could easily be achieved without a loss of previously established grammatical traits (although different pronunciations developed). The Lingua Franca remained astonishingly stable in its grammar (for a Pidgin language) and traces from its original variety are visible to this day and age. The stability of the Lingua Franca, presumably the only Pidgin language that remained close to its original form through all these centuries, is unlikely to have occurred often in the history of Pidgin languages and might even be the exception to the rule. It can be assumed that the stability must, to a major part, have come from the Arabic speaking half as the Lingua Franca was also an Arabic Pidgin. The Lingua Franca not only originated in North Africa but continued to be established in its society (maybe even culture) in its original form quite strongly. Consequently, vocabulary or grammatical shifts, even after centuries of changing language dominance, would not occur, at least not to the degree that would be expected in a Pidgin language. Presumably, the Lingua Franca in its original (Catalan) variety had become strongly established and continued to be spoken by Arabic speakers and thus it was almost impossible for it to grammatically change to more than a slight degree. For the Lingua Franca to be recognizable after centuries in both vocabulary and grammar is, theoretically speaking, most unusual, but serves as both an explanation and indication for the enormous stability of the Lingua Franca. In any case, it seems incredible for any language to remain so similar that grammatical traits of the original variety are still strongly recognisable, with only minor changes in grammar after several centuries of existence. Considering that Pidgins have an inherent flexibility and adaptability (which may be regarded as instability), can

30 Most detailed studies on the Lingua Franca and its (linguistic) development have been conducted by Schuchardt (1909), Cifoletti (2004) and Foltys (1984) to which I can only add a more general perception about Lingua Franca stability 208 quickly adapt and completely shift in vocabulary and even grammar (if the requirements are met i.e. through changing language input), it is truly remarkable that the Lingua Franca remained so close to its original variety.

12.1 Arabic enforcing stability

One explanation for the stability of the Lingua Franca lies in its source languages, most prominently in the substratum language. Considering that the Arabic influence on the Lingua Franca never completely ceased, despite the growing prestige of the Turkish language and vocabulary dominance of other Romance languages, it could be one major reason why the grammatical traits of the Lingua Franca stayed so similar to its original variety. Arabic grammar characterised but also stabilized the Lingua Franca since its creation and as Arabic continued to be an integral part of the Lingua Franca for all of its existence, this seems to be a plausible explanation for the (only) minor grammatical shifts caused by Romance languages. Only a few existing grammatical traits of the Lingua Franca were replaced with other most basic forms which consequently only slightly affected the linguistic appearance of the Lingua Franca (but also made the distinction of the Romance source language more difficult). Curiously, even after the Ottoman Empire annexed the Arab Empire, the Lingua Franca only acquired minimum Turkish vocabulary. This might be due to the Ottomans being very liberal towards other languages, but also not representing a (linguistic) majority.

12.2 Catalan enforcing stability

Another considerable reinforcement of Lingua Franca stability would have been Catalan. Through it being spoken in the whole Mediterranean for centuries, the Lingua Franca did have linguistic backup support, which made it almost impossible for the Lingua Franca to vanish. It would (only) have been the end of the Lingua Franca if commerce, i.e. the form that created the Lingua Franca, had vanished. However, as commerce in its re-established form stayed the same until the end of (politically motivated) piracy of the 19 th century, the Lingua Franca stayed close to its original form. Linguistically, Catalan greatly affected the Lingua Franca in the sense that it stabilized both vocabulary and grammar to such a high degree that the basic grammatical forms and the most common (Catalan) vocabulary remained, even after several relexification processes

209 (visible in the feature of doublets). Although Catalan was eventually replaced by Spanish (later Italian and French), it took almost a century for this replacement process to be completed, thus allowing the Lingua Franca to remain almost unchanged after the Spanish annexation. Catalan stabilized the Lingua Franca until Spanish became dominant and, presumably, the relexification process thus started rather late in the 15 th or even only in the 16 th century. Furthermore, as has been pointed out above there were (grammatical) similarities between the Romance languages of Catalan, Occitan, Italian and Spanish of the 16 th century and thus the Lingua Franca became relexified, i.e. Hispanicized (and Italianized). The Spanish (and Italian) relexification process only slightly changed the appearance of the Lingua Franca, i.e. pronunciation, intonation, lexis etc. adapted to the new superstratum language, while other basic (Catalan) features remained almost unchanged. Presumably, this was not primarily because Spanish speakers used the Lingua Franca but rather because Arabic speakers and speakers of non- dominant languages used its original form to communicate with Spanish speakers (thus establishing a group behaviour and dynamics). The Lingua Franca probably would have vanished the moment the Spanish (crown) became commercially powerful in the Mediterranean. The number of Spanish speakers (i.e. travelling merchants) was steadily increasing, while the Catalan influence steadily decreased until it was finally replaced by Spanish. However, because (Arab) merchants (and the local population) continued to use the Lingua Franca, i.e. the original Catalan variety, the Lingua Franca did not vanish or became replaced by another Pidgin. Through a circumstance quite unique to Pidgins, the vocabulary shift, the Lingua Franca could continue to exist with the same grammatical traits but a different vocabulary. This was quite a fortunate coincidence as otherwise the Lingua Franca would have been replaced by a Spanish or Italian Pidgin and presumably would have vanished after some time (as it was the case with French in the 19 th century). Instead, the Lingua Franca was relexified in Algiers with Spanish, in Tunis and presumably Tripoli with Italian and later again in Algiers with French. This clearly shows the enormous (grammatical) stability the Lingua Franca maintained throughout these centuries. However, this also shows that the Lingua Franca, just like any other language, is influenced by, and changed through, political and economic factors. Political influence itself, I believe, did not necessarily lead to the extinction of the Lingua Franca. For instance, diplomacy, which can be regarded as a political influence, caused the stability, development and even creolisation process (i.e. the Creole-esque state) of the Lingua Franca. In most cases political influence only caused a vocabulary shift e.g. towards Spanish and Italian.

210 However, colonization, which definitely entails political influence, caused the most intensive shift towards French. Primarily, it was the language influence of colonization which caused the further development of the Lingua Franca resulting in its replacement by a French Pidgin and presented a further (de)creolisation process changing the French Pidgin into a Creole. It can be argued whether the French Pidgin of the 20 th century had any connection to the Lingua Franca or if it immediately fully replaced the Lingua Franca and caused its extinction in the late 19 th century. In any case, the Lingua Franca finally lost its influence and became extinct due to its loss of relevance in Mediterranean commerce.

12.3 Relexification as a sign of stability

With regard to Pidgin languages, a vocabulary shift can be interpreted as a sign of stability. Due to the innate instability of Pidgins and their tendency to disappear quite quickly, instability would manifest itself either in the replacement (by another Pidgin) or complete vanishing of the Lingua Franca through the rise of Spanish or the existing Italian influence. Instead, only a relexificatiom process occurred causing a vocabulary shift. This shift of the Lingua Franca vocabulary occurred in the same way as the stabilization process, i.e. through language input. The Lingua Franca was continuously updated by, and receiving Spanish or Italian input (through Spanish or Italian speakers) while being provided with Spanish or Italian Lingua Franca input provided by other Romance or Arabic speakers. One further change occurred once the commercial system became replaced (or rather dominated) by the piracy-slavery (commercial) system. Language influence shifted as the Lingua Franca received multi-language input due to the many different languages present in this sytem, resulting in a great expansion of Lingua Franca vocabulary. What happened was that when Catalan became replaced by Spanish (or Italian) an already established Lingua Franca would be confronted with a rising number of speakers of a different Romance language, i.e. Spanish or Italian. The established Catalan Lingua Franca would still strongly be represented among Arabic speakers and thus would not change considerably. Presumably, Arabic speakers continued to use the Lingua Franca, thus maintaining its grammatical basis, while Romance speakers were only influencing and thus (ex)changing vocabulary. However, through individual contact of Arabic speakers with Europeans (dominated by Spanish and, spatially separated, also by Italian) the Lingua Franca vocabulary became slowly replaced with new Spanish or Italian words without changing its structural grammar (or its way

211 of operating in commerce). A process that happened not only twice, but thrice as the Spanish Lingua Franca was later replaced by French. The Lingua Franca had (backup) support through the Arab people as both merchants and the local population created the necessity for its use and with it the inevitability of its survival. The Lingua Franca could have died out with the overtaking of the Crown of Aragon by Spain, the displacement of Catalan by Spanish and Italian in the Mediterranean or another Pidgin could have taken its place. However, through the survival of Catalan for another century due to Catalan being spoken in its former colonies until the 16 th century as well as the Catalan Lingua Franca still being spoken by the North African population, the Lingua Franca could outlast the demise of the Crown of Aragon. Through trading with mainly Spanish or Italian speakers (depending on location), the slow but constant process of increasing Spanish influence and the already strong presence of Italian, only a vocabulary shift occurred instead of the replacement with another Pidgin or the demise of the Lingua Franca itself. Without such a massive backup from more than one side, the Lingua Franca would have disappeared without a trace. This, however, occurred centuries later when French became the official language as well as the language adapted and spoken by the colonists, the local population and in commerce (Abulafia, 2014: 147). There definitely were more reasons to adopt French instead of the Lingua Franca. However, the replacement process did not occur immediately. At first, a Frenchification process was initiated, but the process of creating a French Creole was completed around 1890 (Foltys, 1984: 33) and with it the end of the Lingua Franca was determined, as hardly any typical Lingua Franca trait remained in the French Pidgin and none in its Creole. The extinction of the Lingua Franca (and its last traces) occurred before the 20th century when there truly was no linguistic (backup) support left for any Lingua Franca variety.

12.3.1 The process of French relexification and Lingua Franca replacement

As expected from a Pidgin, the Lingua Franca became relexified, i.e. Frenchified and developed into a French Lingua Franca variety called Sabir. However, through the local population supporting French (or, depending on viewpoint, lack of support towards the Lingua Franca), the Lingua Franca was then replaced with a (completely different) French Pidgin which was also called Sabir. The designation of Sabir for two hardly related languages (Waille, 1884 quoted in

212 Cifoletti, 2004: 274) is very diffuse and the distinction of Petit Sabir which supposedly developed into Grand Sabir is described by Schuchardt (1979) as “quite arbitrary and fantastic” (1979: 43). The only true connection between these languages seems to be the French language influence and Arabic basis. What is noticeable in the early Sabir documents is that Sabir appears to be quite close to the Lingua Franca. This proximity became completely absent in the French Creole in which (almost) no traces of the Lingua Franca are noticeable except the /e/ to /i/ sound change which, however, could also be due to Arabic influence. The transition from the Lingua Franca to a French Pidgin and a French Creole is most astonishing considering the great development in such a small time period of about 30-40 years (Foltys, 1984: 32-33). The reason presumably was due to language contact between French and Arabic speakers being very intense, although perhaps not in the positive sense as Algiers was colonized and the people were basically forced to learn French and thus the language support from the local population was more enforced than voluntary. Both the stability of the Lingua Franca in the 16 th century as well as the replacement of the Lingua Franca in the 19 th century was due to the population supporting either the Lingua Franca or French and integrating, or at least using, it in everyday life. For instance, in 16 th century Algiers the Lingua Franca was distributed, known and spoken among a majority of the population and thus stabilized, while in the 19 th century French was supported. Thus a French Pidgin, later a French Creole, developed and was stabilized. Algiers of the 19 th century, with French becoming the official language and means of communication (and thus replacing the Lingua Franca) was unlike Algiers of the 16 th century with Spanish becoming the official language because the local population was still supporting and using the Catalan based Lingua Franca and thus only relexifying the Lingua Franca in the process and allowing the continuation of its existence.

12.3.2 Colonization as reason for replacement

Colonization would serve as an explanation as to why the Lingua Franca became replaced in the 19th century. When for instance Spanish or Italian dominated commerce (economically) they did not dominate the Lingua Franca speaking areas politically (or linguistically in that regard), as these were separate states and empires. Neither Spanish nor Italian were the official languages of these states and thus there was no language forced upon the Arabic speaking population, only close and intensive contact between Spanish or Italian and Arabic speakers predominated, which

213 led to a relexification process. A similar case is the Crown of Aragon, which did not colonize the Southern Mediterranean where the Lingua Franca was spoken, although various islands inside the Mediterranean basin from Malta, Sicily even up to Crete were annexed. What was, however, truly colonial was the aspect of exploitation (Abulafia, 2014: 150). The true difference was that the Lingua Franca in previous times and locations was influenced language-wise but even if one language dominated commerce there was no political dominance as such and with it language was not forced upon the people. This would lead to the assumption that the Lingua Franca relexified in economic dominance but became entirely replaced in political dominance, e.g. when a language becomes politically enforced upon the population. This is the most plausible reason why the Lingua Franca did not become replaced during the long time of its existence, because it did not originate nor was it ever spoken in a true colonization setting. Language-wise, the Lingua Franca never was enforced to become adopted and, instead, adapted via influence through intensive and close language contact. This is the one common denominator that all these (different) Lingua Franca relexifications share: intensive (language) contact between speakers of different mother tongues. One important point to note as to why the French Lingua Franca relexified in the first place before it became replaced during colonization is that French as a language became predominant in the Mediterranean as its influence grew since the 16 th century; especially the 18 th century was dominated by the French (Greene, 2014: 94, 101). Therefore, a relexification process started due to close and intensive language contact via commerce between Lingua Franca and French speakers, which continued regardless of the annexation of Algiers and could even have started before the annexation itself due to the economic dominance of France. Second, the Lingua Franca only became replaced through a French Pidgin later on which can be explained through the time it took to truly colonize Algiers. After all, colonizing an area takes time and, depending on the efforts of the colonizers, the (pre)existing conditions can sustain themselves and thus remain for a certain amount of time until institutions, facilities and a colonial government are set up to enforce language, regulations and laws, i.e. gain political control, including the implementation of the official language. However, until this time the colonization language may be present enough to exert influence and cause a relexification process but is not yet dominant enough to initiate a pidginization process, i.e. create a Pidgin of its own and replace an already established Pidgin. In other words, French had yet to set up and enforce its influence and dominance to create a French Pidgin before it could replace the Lingua Franca, which took some time.

214 The reason why the Lingua Franca relexified towards French in the first place, but only became (fully) replaced after almost 50 years after the annexation of Algiers, is because relexification can be established within a short amount of time if contact is sufficient, intense and of great enough quantity. The creation of a Pidgin including the development, and especially the replacement of another Pidgin, on the other hand, is a (much) longer process. After all, an extended Pidgin takes years, possibly decades, to develop and a fixed Pidgin is unlikely to replace another already established extended Pidgin. This also shows one further aspect about the relexification process: the incredibly little amount of time it takes to relexify an already existing Pidgin (given that no tremendous differences exist among the original and relexifier languages). The rapidity of this process seems to have an enormous temporal advantage over the pidginization process itself. Therefore, a temporal difference between the relexification and the replacement of the Lingua Franca (by French) can be observed. Altogether, the same vocabulary shift also occurred when major changes within the Mediterranean power distribution took place, i.e. once the Catalan commercial empire vanished and Spain, Italy and the Ottomans, later France and England, took its place and its commercial dominance. Three independent systems dominated by each political power were created and caused the emergence of the Lingua Franca varieties, as defined by Cifoletti (Cifoletti, 2004: 32), where a major focus of each group lay on one city, which in all three cases happened to be the most important particular commercial city of this part of the Southern Mediterranean.

12.3.3 Arab population ensuring stability

The reason the Lingua Franca was only relexified and remained spoken until its final replacement was not because (of the common notion that) it was easy to learn. In fact, any Pidgin is easy to learn, especially in its early Pidgin states such as ad-hoc or fixed ones, but even the extended state does not take too long to acquire, especially for someone that knows the superstratum or vocabulary source language. It would have been even easier to acquire a newly created (fixed) Pidgin as it would have been less developed and less complex than the Lingua Franca. Therefore, the reason for the Lingua Franca to still be spoken was because its use was assured by the creators of the language itself, the Arabic speakers. North Africa was thus not only the place of origin but also the reason for the Lingua Franca’s further existence and stability, which in further consequence ensured the Lingua Franca existed long enough to (spread and) become spoken

215 throughout the Mediterranean. The reason itself for Arabic speakers to use the Lingua Franca was either due to its usefulness or because of the necessity to communicate with Romance speakers (merchants) in general, but it was neither the simple fact that the Lingua Franca was easy to learn, nor its so-called simplicity. The reason was the continued use of the Lingua Franca by Arabic speakers that caused its stability.

12.4 Economic explanation for the Lingua Franca stability

It could be argued at this point that it was also due to economic reasons that the Lingua Franca survived for such an extended period of time. The rise of piracy definitely played a major role in the development of the Lingua Franca and was encouraged by political reasons, i.e. due to the licensing of corsairing or marauding activity which were either nullified or legitimized through peace contracts or declarations of war, trading licenses, etc. Despite the rise of piracy the economy and consequently commerce were further established, which resulted in the economy becoming heavily dependent on slavery itself. In fact, piracy almost defined the whole economy (Selbach, 2008: 34) and with it also the need for (or probably even the forceful use of) the Lingua Franca. In fact, it was used by pirates and slavers to communicate with their captives and slaves. One further reason for the continuous stability and the survival of the Lingua Franca were the centres of piracy, which were basically the same commercial centres as before. However, the drastic shift towards slavery also changed the main Lingua Franca group of these commercial centres from merchants to pirates, slavers and the local population. Pirates were coming from (or to) locations in which the Lingua Franca was spoken (before) and the pirates adopted, if they did not already know, the Lingua Franca. Once they took it onboard their ships the Lingua Franca definitively became a sailors` language as well as (infamously) known within the Mediterranean. The main focus of the Lingua Franca, however, remained commerce (Pananti, 1817 quoted in Cifoletti, 2004: 245). The constancy or even intransigence of the stability of the Lingua Franca might therefore be due to the pirates and slavers themselves using the Lingua Franca, even if they only took over parts of the language or a lower state of the Lingua Franca as a means of communication. As long as piracy thrived, as long as Algiers as well as Tripoli and Tunis remained piracy strongholds, the Lingua Franca had a cornerstone in form of locations in which it was spoken regularly, constantly

216 and even interactively (due to the individual reactions of the slaves). In other cities than Algiers, Tripoli or Tunis this might have been different as there are no detailed reports or documents about other cities. Thus only (a historically based) speculation can be offered. What individually happened in detail in other locations is up to speculation, as other factors (or groups) might have determined how or why the Lingua Franca survived or was stabilized. However, from the historical sources and linguistic documents it can be deduced that there were not that many (other) factors or groups to keep the Lingua Franca alive, i.e. commerce including merchants, sailors and pirates, slaves and the local Arab population of the Southern Mediterranean. All in all, the stability and longevity of the Lingua Franca could have been caused by both factors determined via the groups involved and from the economy, which was being politically motivated. Furthermore, although the Lingua Franca was adopted by other groups, i.e. North Africans including pirates, slavers, renegades and locals who only spoke the Lingua Franca because it was the common means of communication, would not necessarily mean that the Lingua Franca abandoned the group originally associated with it, i.e. merchants. Arguably, it might have been different as to how or why the Lingua Franca could stay alive for centuries, but unfortunately hardly any proof remained except that all merchants, pirates, sailors and slavers used the Lingua Franca in the piracy strongholds of Algiers, Tunis and Tripoli in which the Lingua Franca flourished during the 16 th to 19 th century until piracy was discarded as illegal.

217 Chapter 13: Anthropological approach

To further study the Lingua Franca and reconstruct its origin, spread and development an anthropological approach is necessary. Human behaviour (in general), the behaviour of (specific) groups including background knowledge about their trade (i.e. what people were involved), who they interacted with, to what extend and what kind of (language) contact existed, and also which locations were involved are factors that offer background knowledge that is indispensible. It is absolutely necessary to consult the behaviour of specific groups that are (even remotely) involved with the Lingua Franca due to a lack of documents, reports and especially an absence of serious linguistic studies before 1880. Schuchardt himself was the first to study the Lingua Franca in 1880 (Schuchardt, 1979: 26) to provide a plausible explanation how, when and where the Lingua Franca originated, spread and developed.

13.1 Notion of purity

One (definitively political) reason that explains the lack of (serious) linguistic studies and documents was the notion about “purity” of languages which made many people either fully condemn the Lingua Franca as ´impure´ and ´bastardized´ or completely ignore its existence. As there were many captives and slaves that wrote about their experiences, the whole Mediterranean became aware of the Lingua Franca. The circumstance that the Lingua Franca obtained its fame due to piracy, especially since the 16 th century, consequently led to it being not always mentioned in a positive manner due to being identified with piracy and slavery. This bad reputation continued and found its way into linguistic circles which described this Pidgin as a bastardized or broken language, not a true language, and therefore unworthy of being studied. The field of research of linguistic studies itself is not very old (150 years), and therefore, a linguistic study before the 19 th century is quite rare. Even though there were individuals that did involve themselves in linguistic studies, such as Dante Aligheri in his De Vulgari Eloquio (Dante, 1304), these treatises, although offering quite professional and valid observations, are lacking linguistic essence and are thus not as useful as later studies in this context.

218 13.2 Primary and secondary spread

As the groups involved presumably are primarily travelling European merchants and local North African merchants it is indispensible to study and analyse their specific behaviour. This analysis should include the whole commercial network that allowed the whole situation to be created in which these two groups came into contact with each other to be able to draw considerable conclusions as to why the Lingua Franca was spoken throughout the Mediterranean but was hardly publically known or accepted. Therefore, one aspect that should be taken into consideration is the behaviour of merchants and how this relevant group behaved to fully reconstruct their involvement in the origin, development and spread of the Lingua Franca. While language contact between two groups, i.e. Catalan and Arab merchants, was the reason for the creation of the Lingua Franca, presumably, one further group (i.e. non-Catalan travelling merchants) was responsible for its spread. The first group of Catalan merchants was responsible for the intelligibility and the other group of travelling merchants was responsible for the spread of the Lingua Franca, while both groups were geographically active, i.e. in motion. The other group of Arab merchants was responsible for providing a cornerstone of stability while geographically being passive, i.e. remaining stationary. This results in an asynchronous spread across the Mediterranean Sea, in particular within commercial cities. It seems to be restricted to (only three) Southern Mediterranean cities, as the Northern Mediterranean cities, at best, seem to report the Lingua Franca to be noticeable (but less notable). This could lead to a classification of a primary and a secondary spread. Primary being the spread across the Mediterranean through the travelling merchants, while secondary being the spread in several commercial cities across the Mediterranean through the local merchants.

13.3 Group dynamics

This theory about human behaviour further involves two more assets or levels of interaction. One is the behaviour or role of the individual and the other is the behaviour or role of the group, the (so-called) group dynamics. The behaviour of a whole group or system (even a nation) consisting of thousands of people repeating a certain set of actions, interactions and behavioural patterns in both the linguistic and the geographic sense can also be analysed as a commercial system exchange. Wansbrough has referred to the repetition of such an economic behavioural pattern as “orbits” (Wansbrough, 1996: 6).

219 The behaviour of the individual may be context- and personality-dependent and thus a lot of variety in reaction and difference in language contact situations can arise. However, even single reactions are limited in the sense that they create a reoccurring pattern of similar or same language situations which, through mere repetition, are experienced by quite a number of individuals. This, in turn, creates a similar set of experiences for a group, i.e. a group dynamics. Even though single experiences may vary a lot, on a broader perspective people’s experiences show similar, if not the same, results. One such situation, and the one most is known about from the reports, is the situation of captivity, slavery and imprisonment, i.e. being captured by pirates, being held for ransom, being kept in prisons (or bagnos ) and being forced to work. This situation, created by a certain group, pirates that were also slavers, faced quite a number of individuals. Although there must have been (general) variation among the individual reactions and dealings with this situation, these were rather exceptions, i.e. responses that were limited to very few individuals (and thus not creating a group dynamics). For the most part, people reacted similarily, i.e. in a specific reoccurring behavioural pattern which then created recognizable speech patterns (or forms) of the Lingua Franca. It soon becomes clear through many reports that the way such situations were handled by the slavers was always the same. Not only was it business routine but also a repetitive (behavioural and speech) pattern that did not respond to individual behaviour(al input). In the reports, clear structural behavioural patterns of orders, insults, beatings and repetition of these (orders, insults, beatings) shouted in Lingua Franca are well documented. The individual was treated like any other slave no matter the individuality or personality. The situation that repeated itself for centuries was that travellers, but also merchants, were captured by pirates who spoke the Lingua Franca, were then brought to slave capitals such as Algiers, Tunis or Tripoli and thrown into prisons in which the slavers also spoke the Lingua Franca. They then were forced to work in the city or as house slaves in an environment in which the Lingua Franca was used by the local population to communicate with the same slaves. Probably, the Lingua Franca was spoken in homes due to the slaves being omnipresent and thus it might have become customary to use the Lingua Franca. As some captives were being forced to stay as slaves for years they inevitably heard, learned and, in the end, some spoke the Lingua Franca themselves.

220 13.3.1 Dynamics of political-economic systems

A functioning communication system was already established by the 16 th century. Presumably, it was created years or decades earlier through the rise of piracy itself which established a routine procedure of capturing and selling slaves and then settled in three major cities. This created a group dynamics (and behaviour) of both ´piracy and slavery´. Some individuals were using the Lingua Franca to communicate with their captives and with it created the group behaviour of piracy and slavery. Other individuals also inevitably acted as a group due to their (forced) role of being captives or slaves and being forced to interact linguistically with the Lingua Franca, thus forming the group behaviour of captives and slaves. In some regard, both sides acted according to the created (dynamic) group (behaviour) they belonged to, i.e as captives or slaves and pirates or slavers. This system was realized through the contact and exchange between different (nations´) commercial systems. It was not because the commercial system itself demanded slavery, this was rather due to political reasons of nations legalizing slavery and corsairing, but because the system of slavery could parasitically abuse the commercial system to its advantage. This slavery (based) system developed on the basis of the commercial system and involved several nations. Thus piracy and slavery became a part of Mediterranean commerce itself. At some point in the 16 th century, the Lingua Franca was an inherent part of a functioning slavery and piracy based system that forced all its participants, regardless if these were individuals or groups, voluntarily and involuntarily, to participate in communicating with the Lingua Franca. The result was a considerable stability for this Pidgin language and a constantly rising number of Lingua Franca speakers. This, however, was not because of the captives which did pick up (parts of) the Lingua Franca but probably never used it again, but because of the pirates and slavers who increased massively in numbers (i.e. a group development) including the so-called renegades, (Europeans) who decided to switch sides (Selbach, 2008: 34). Even the local population itself started using the Lingua Franca to communicate, which also increased its numbers. These groups were the main initiators of Lingua Franca use. Diplomacy itself was not directly involved in the piracy and slavery system as it only negotiated peace treaties and the relations between nations, not with the pirates or slavers themselves, although this would serve as a good explanation as to why diplomats and ambassadors spoke the Lingua Franca in the first place. Guido Cifoletti has a good explanation for this case, pointing out that the Turkish administration and ambassadors, at some point, consisted of soldiers and rather

221 low-educated people (Cifoletti, 2004: 48).

13.3.2 Group dynamics and origin

It is important to note that although everyone reacted and experienced captivity individually, there still is a dynamic group behaviour and (shared) group experience recognizable with both of the captors as well as the captives.. The important aspect about group behaviour is that even individual situations can be summarised as a group pattern or behaviour if they occur often enough. The same can be applied to the origin of the Lingua Franca in which an interrelation between group dynamics and individual variation (presumably) created the Lingua Franca. Through repeated language contact between two different language groups, which was caused by group dynamics and resulted in constant interaction between individuals, a common means of communication was created. However, for this process to reach a large scale, which probably took decades or even a century, a large amount of repetition of similar (or same) language situations was necessary. This could only be provided by commerce, i.e. a group behaviour repeatedly creating such a situation. Such an “orbit” (Wansbrough, 1996: 6) or consistently repeating pattern also produces certain patterns in both language and behaviour. The (group) behavioural pattern that can specifically be identified with this can be found in the merchant group. Specifically, there were travelling merchants sailing from port to port to buy or sell their wares and this group (dynamics) was met with another group (dynamics), the local merchants. Through trade and interaction between these two groups a language contact situation was created in which two different groups speaking two different languages were encouraged to interact with each other repeatedly. The group dynamics here was that Catalan merchants were allowed to enter North African ports and thus large-scale commerce was able to (re-)establish itself (although this took centuries). Within this functioning system of commerce, the group (dynamics) of European merchants interacting with Arabic-speaking merchants a language competence was created and stabilized through its constant repeated use and repetition, which then caused the creation of the Lingua Franca. The Lingua Franca was created due to an orbit. However, despite the constant repetition this group pattern enabled, this orbit only provided the necessary basis for the origin and persistence of the Pidgin. It was the individual that allowed for the Lingua Franca to develop further due to

222 the great amount of variation that occurred on the individual level. Within this group dynamics, individual variation was unavoidable as many merchants operated on an individual level (one-on- one basis), thus creating the necessity for expressing a greater variety of linguistic topics, vocabulary, etc. This reveals quite a complex overall picture as through both the sophisticated group behaviour and the varation through individuals (of different languages), communication evolved. There was considerable language contact between two groups of separate language systems with one group establishing language contact and the other developing language competence. This situation repeated itself in several locations of the Mediterranean involving different individuals that were confronted with the same language situation (until a Pidgin language developed). Once the origin of the Lingua Franca had occurred, the basic (group behaviour) setup as such remained. (However, with continued use of this Pidgin the language evolved and was spread further into the Mediterranean through individual language contact.)

13.4 The importance of the individual in group dynamics

The individual level definitely plays a role in the development of the Lingua Franca. It is present in the state of creation itself, but especially the further development of a Pidgin depends on variation which is provided by the individual level. What can be said is that the Lingua Franca originated through simple individual one-on-one (basic) language contact, but it was group dynamics that allowed such a situation to arise in the first place. Furthermore, once established, the continued existence of a Pidgin depends on the continuity of this group behaviour process. The Lingua Franca has always been following a strictly repetitive (group) pattern in creation, development and use, as it has been conducted in this repetitive fashion ever since its origin; a repetitive situation handled by a (simple) means of communication used for trade which becomes visible especially on a massive scale. Group behaviour further allowed the same situation to arise in different locations along the Southern Mediterranean in which the same pattern of individual language contact occurred. The Lingua Franca definitely started on the individual level. However, due to large-scale commerce an increase of language contact appeared involving more and more (Catalan and Arabic) language contact until it can not be defined as a single occurrence but group behaviour. Assumingly, the Lingua Franca was taken on by both groups involved which led to the group

223 dynamics of one group spreading the Pidgin further while the other group developed it. However, through the group dynamics of group A, the travelling merchants, the developed Lingua Franca could travel from its original location to another and thus a dynamics was created that assured a constant level of development all over the Mediterranean. Technically, group B, the local merchants, would expand the language through language contact with group A. However, once travelling merchants of different nationalities appeared (group X), group B was updated through groups A-X. Furthermore, as new locations of the Lingua Franca arose including almost every commercial city which involved other local merchants (group Y), group B was expanded to group B-Y. This caused further spread and development of the Lingua Franca until it could not be distinguished anymore which location was the original place of origin. By that time, all of these locations assured the constant development, spread as well as the stabilisation of the Lingua Franca.

13.5 Pidgin expansion through individual dynamic

The individual aspect of interaction becomes relevant for the further development of a Pidgin, as it creates variation in repeated situations and this aspect is what encourages a Pidgin to develop (and evolve) beyond the fixed and even the expanded Pidgin state. Individual (another descriptive term would be personal communication (and interaction)) is the complete opposite of repetition and stability. It opens the possibility to explore different expressions, i.e. a level of interaction beyond fixed patterns which explores other new topics or areas of communication. Individual interactions between speakers of different languages facilitate the further development of a Pidgin language and with repeated individual or personally involving interactions a Pidgin can evolve into a Creole. New language situations (or problems) arise that can and have to be solved by choosing (different) words and phrases, expanding grammar and grammatical expressions and mutually finding solutions to establish communication. Individual variation provides the necessity for new vocabulary, new meanings (sometimes derived from the old vocabulary), phrases, prepositions, etc. The need to express something new only arises in a new (kind of) language situation, definitely not through the repetition of the same (old) situation. Considering the development of the Lingua Franca, the only way this could be achieved was via individual variation. Individuals coming into a language situation that had not been experienced before and thus could not rely upon (previous) context or expressions needed to express

224 themselves differently. Certain kinds of situations require certain expressions, and if these do not exist new ones are created or old ones gain new meaning through a different context. Consequently, individuals may express themselves differently in the same language situation using different words or phrases and/or extending their meanings. Even the use of grammar can individually differ, i.e. endings, suffixes, plural, etc. Morphology is affected in general, while morphological derivation is possible as well, although inflection is less likely to occur in a Pidgin (rather in a Creole). However, these variations are still bound to the rules of Pidgins (and Creoles), thus not exceeding certain expressive limitations. Expressions, phrases or grammatical traits may be created and developed through individual variation, but the most commonly used expressions, phrases, endings etc. are adapted by the group (dynamics) to become fixed in form and meaning through this process. Furthermore, (individual) variation that can not prevail itself or does not lead to successful communication is abandoned (early on) through group dynamics. Therefore, the Lingua Franca may have spread through group dynamics but developed through the repeated process of individual one-on-one interaction which was brought about through the sheer number of individual encounters (i.e. group dynamics).

13.6 Interaction of individual and group dynamics

Language contact started via repeated one-on-one contact between Arabic and Catalan speakers, which was the beginning of the group dynamics or “orbit” (Wansbrough, 1996: 6). First, Catalan merchants were venturing to Arabic speaking locations and a language situation requiring communication among different language speakers arose. This linguistic problem reoccurred and was solved repeatedly on a one-on-one basis. Communication between a Catalan speaker and an Arabic speaker was established. A solution was found through the Arabic speaker adapting to the Catalan language and the Arabic speaker would use the previously acquired Catalan words with the next Catalan speaker. While the Catalan merchant would venture further to other locations and face similar language situations, the Arabic speaker would remain stationary (from a broader geographical perspective, i.e. not leave the city) and communicate with the next Catalan speaker. After several thousands of (individual) encounters the Arabic speaker would most likely have developed a certain degree of Catalan language competence and apply the acquired Catalan words with other Catalan speakers,

225 all from one one-on-one communication to the next. Single Catalan-Arabic communication exchange might be the basis of language contact but the repetition of this language contact situation was the crucial factor for the inclusion of Catalan as the source language of the Pidgin. Basically this entails that there existed a group dynamics that was responsible for the Lingua Franca origin, as it caused frequent language contact and exchange between speakers of different languages on a regular basis. Furthermore, origin and development may be based on a one-on-one language situation as market situations usually require a one-on-one communication in which many people are involved simultaneously, although the repeated aspect is what defines this process as a group dynamics.

13.7 The process of update

Once Catalan and Arabic speakers communicated, Arabic speakers adapted some traits of Catalan, thus expanding their Catalan language competence, and they applied their knowledge with the next Catalan speaker. The Catalan speakers (i.e. travelling merchants) brought their language to other locations in which they would apply Catalan in a similar language situation with other Arabic speakers (i.e. local merchants), who would adapt to this new language situation by acquiring some Catalan words as well. The Arabic speakers in this location would then also apply their acquired Catalan with other Catalan travelling merchants and the success in this kind of communication facilitated the further use of individually acquired Catalan. As long as only Catalan speakers were involved this facilitated the acquisition of Catalan words, although this Catalan language competence did not represent (or facilitate) a connection among different locations. Furthermore, Arabic speakers with Catalan language competence would not use Catalan among themselves and Catalan merchants did not adapt (to) Arabic. However, the linguistic situation became much more complex once commerce reached a large scale and other Romance speakers (i.e. non-Catalan travelling merchants) started to appear in large numbers. Through this development all of the Catalan language competence of Arabic speakers, who had previously received only Catalan input, was replaced by the Lingua Franca. While still receiving Catalan input, Lingua Franca input appeared as well due the large number of non-Catalan speakers acquiring the Lingua Franca. The interconnectivity of commerce allowed speakers of different languages to spread the Lingua Franca to locations in which Catalan language competence had persisted before. As a result a chain reaction of updating and

226 combining these separate competences and inputs began. Travelling merchants (of all nationalities) came into contact with Arab merchants and once some acquired the Lingua Franca, an exchange and update of separate language competences occurred. With this development Lingua Franca input started to appear, resulting in a constant Lingua Franca update process due to Lingua Franca speakers travelling all across the Mediterranean. This further resulted in the spread of the Lingua Franca in all important commercial locations in the Mediterranean (among merchants and sailors). Furthermore, this constant update process caused the Lingua Franca to be intelligible among all Lingua Franca speakers. Catalan input was serving as a main stabilizer of intelligibility and the probability for the Lingua Franca being updated by other Lingua Franca speakers presumably was low during the prevalence of the Catalan language. This changed dramatically after the 15 th century with the collapse of the Crown of Aragon (and the gradual disappearance of the Catalan language in commerce), allowing the Lingua Franca to become acquired by a variety of speakers. Commerce, however, remained on a large scale and travelling merchants of all nationalities started to adapt the Lingua Franca and spread it across the Mediterranean. Especially non-Catalan speakers would most definitely pick up the Lingua Franca as it would allow them to conduct business with a large number of merchants within the Mediterranean (and, in part, even outside of it). These merchants would ensure the spread of, and provide further input for, the Lingua Franca. The effect of updating was that many varieties would coexist alongside each other while all varieties and pronunciations of the Lingua Franca remained intelligible to each other. The process of updating continued onwards not only once the Pidgin had spread in the whole Mediterranean (presumably before but definitely) by the 16 th century, but continued once the Lingua Franca varieties had developed and even lasted until the final vanishing of the Lingua Franca itself (in the 19 th century).

227 Chapter 14: The underestimated involvement of the Jewish population and Judeo-Arabic

According to recent studies by Guido Cifoletti (2004) and Corré (2005), a deep level of involvement of Jewish people and the Lingua Franca existed which had a huge impact on their language (Corré, 2005). The Jewish population in the Mediterranean and presumably the Jewish merchants had such a close connection to the Lingua Franca that many words and expressions entered their own language, Judeo-Arabic, which would suggest that the Lingua Franca was not only known but also spoken by a large number of Jewish people. The problem concerning the Jewish people in regard to the Lingua Franca is truly the documentation or rather the lack of it, as there is no reference of Jewish merchants being in contact with the Lingua Franca. The assumption that Jewish merchants were using the Lingua Franca can not be confirmed by the known sources and articles about the Lingua Franca, as there is no mentioning of a connection between the Lingua Franca and Jewish merchants at all. Even if Jewish traders were using the Lingua Franca it is never explicitly mentioned or even indicated in the known sources, the only exception being Jewish pirates. The largest corpus of documents about Jewish merchants being involved in commerce, the Cairo Geniza documents, a large corpus of (Hebrew) texts from the 11 th and 12 th century, also gives no indication about the involvement of the Lingua Franca and Jewish merchants which, at least, would confirm that the Lingua Franca presumably did not exist in the 11 th or 12 th century (Astren, 2014: 399). The only document indicating a close connection to the Lingua Franca is the Shay Lamora , mentioned by Corré (2005) in his cybergraph, which shows that the Judeo-Arabic language of 19th century Oran is full of Lingua Franca words (Shay Lamora, 1838 quoted in Corré, 2005). This document would not only confirm the use of the Lingua Franca by Jewish merchants but also the deep involvement of the Lingua Franca in their lives and language. The insight of Corré would assume that Jewish people would use the Lingua Franca, presumably some words of it, in their own language, Judeo-Arabic, and consequently among themselves, thus integrating the Lingua Franca vocabulary into their own language. Corré (2005) does not specify if the Lingua Franca words found in the book are Italian, Spanish, French or Catalan based, but he mentions that they have been integrated into the Judeo-Arabic language (Corré, 2005). For all that is known, this could have happened in the 19 th century with the Spanish Lingua Franca as Corré mentions that the book is from 19 th century North Africa, Oran (Corré, 2005).

228 Unfortunately, there is no indication as to when this adaption of Lingua Franca words into Judeo- Arabic happened. It could have been as early as the 15/16 th century, but probably not before this date as this would have happened rather after the expulsion of Jewish people from the Iberian Peninsula in 1492 (i.e. Tagarins, Serafins and Moriscos), as they emigrated to locations such as North Africa, Italy and the Ottoman Empire (Astren, 2014: 402), i.e. into parts of the Mediterranean in which the Lingua Franca was spoken, which was not the case in the Iberian Peninsula. However, it could also have happened in the late 18 th or even only in the 19 th century once an opening up towards European culture as well as a creolization combined with a rising popularity of the Lingua Franca occurred.

Following the theory of integration, the process itself is very inaccurate with regard to specifying a particular date and could have thus taken place in any time during the possible existence of the Lingua Franca (14 th-19th century), and tracing it back to an exact date would prove to be difficult (for several reasons). The evidence that Lingua Franca words entered the Judeo-Arabic language was discovered (only) in the 21 st century by Corré (2005) in a book from the 19 th century i.e. 1838 (Corré, 2005). Unfortunately, this only allows a general conclusion that this process did happen before this book was written but does not allow the estimation of an exact date, as there is no indication as to when all these words could have entered the Judeo-Arabic language. Corré (2005) reports about this book, but unfortunately there are not enough examples which could further help to determine the time period since it could contain clues – as to what Lingua Franca words were used; if they were from the Spanish or Italian variety, if they were Romance words or from a Berber dialect, or words only found in the Algiers variety – which seems the most likely case as the book is reported to be from this area – or even the Catalan Lingua Franca, which would indicate that this process started in the beginnings of the Lingua Franca origin and spread, i.e. presumably before the late 15 th century. Unfortunately, there is no further documentation that gives a clear indication when or even that this could have happened, which could only lead to the conclusion that this was a rather late development. Furthermore, that the Lingua Franca was acquired early on, from a source text perspective, would seem unlikely, although the Lingua Franca could easily have reached Mediterranean locations with a Jewish population; but contact is never reported and a deep involvement of the Lingua Franca in Judeo-Arabic can only be attested in the Judeo-Arabic of the 19th century. It simply can not be determined for certain if this event occurred before the 19 th

229 century, as only a few hints, indications and suggestions, which might not be regarded as proof as they can be interpreted in quite a variety of ways, of which all seem plausible enough, are available. A lot remains unclear in what little evidence is left about the Jewish involvement in the Lingua Franca, it could have happened quite recently, it could also date further back or it could have been a slow but continuous process that started since the earliest contact of this group and the Lingua Franca. This would not seem that implausible considering the considerable interconnection of the Mediterranean and possible spread of the Lingua Franca at that time. As the Lingua Franca, or parts of it, has survived as substantially as Corré (2005) states, a close connection between Jewish people and/or merchants and the Lingua Franca definitely existed, and judging from the mere quantity of vocabulary in Judeo-Arabic it could be assumed that this could have been a longer development. However, even if some Lingua Franca words became adopted early on into Judeo-Arabic it can not have been as many as Corré (2005) has found because the Lingua Franca was most likely only an expanded Pidgin and did not have that much vocabulary to begin with, as is seen in the earlier reports which show that in previous centuries the Lingua Franca was not that heavy on words; and only in the 19 th century could whole paragraphs be filled with the Lingua Franca. The plausibility of the Lingua Franca entering the vocabulary of another language seems more likely once the Lingua Franca evolved out of the expanded Pidgin state and reached an almost Creole-like state, and also became massively integrated into people’s lives. This happened in the 18 th to 19 th century, and as the document is from that time too (19 th century) it would seem more likely that this was a later development. It can also be assumed that this would only have happened in the south (eastern) parts of the Mediterranean as this was a region in which the Lingua Franca was most integrated into local life and experience. Perhaps it was not the Judeo-Arabic Jewish merchants, who definitely did have close contact to the Lingua Franca, who adapted Lingua Franca words but, instead, rather the local Jewish population that integrated the Lingua Franca into their own language. Further considering the enormous popularity and spread of the Lingua Franca in the later decades, it seems very plausible that the Jewish community itself had contact with the Lingua Franca, and this scenario would suggest that the Jewish population could easily have adopted it. This, however, remains speculation as it is never clearly identifiable when or how exactly all of this happened.

230 14.1 Problems with the theory of Lingua Franca integration

The theory that Jewish merchants used the Lingua Franca and (partially) integrated it into their language can only be applied to Iberian Jewish merchants, as these merchants were, in fact, known for travelling far distances while remaining close to their own people (Astren, 2014: 399). Thus they must have been in close contact with both the Lingua Franca and their community. Jewish merchants would have known the Lingua Franca before their expulsion from Iberia, (that is) if the Lingua Franca existed at that time it would probably also have been known in the whole Mediterranean by the 15 th century. Hence it is possible that the Jewish merchants could have acquired the Lingua Franca before their expulsion, brought it with them and kept it as it was used among them already. While this is a (faint) possibility, however, it would seem more plausible that the Lingua Franca was acquired later on and integrated only once in a different location. After the expulsion of Jewish people from the Iberian Peninsula in 1492, the Jewish people migrated to the Islamic Maghreb and Ifriqiya with communities in Morocco and other cities (Catlos, 2014: 371): furthermore, these locations with a Jewish population are known as Lingua Franca speaking areas and thus the plausibility of people acquiring the Lingua Franca in such a location is quite present. Jewish people were known for remaining close to their own group, tradition and language (Astren, 2014) and this closeness applies not only to their own people but also to their own language, which would imply that there existed no communication problem (among them). This in turn would have prevented another language being either adapted by the community or being (heavily) integrated into it, as already established and perfectly functioning communication system prevents (especially) grammatical traits being integrated since these traits are already stabilised. The only possibility for one language becoming accepted by the whole community would be an increased borrowing of loanwords which would enter the lexicon but would not affect its grammatical traits. In fact, only words and expressions seem to have entered (a stabilised language such as) Judeo- Arabic, and therefore, this theory would suggest that not the whole Lingua Franca but parts of it, i.e. vocabulary, was integrated into their own language. Curiously, this would also imply that having no communication problems among a group prevents grammar becoming adapted, although single words and expressions can be heavily integrated into the native vocabulary if a close connection between the donor language and the community exists.

231 Linguistically, it would be possible to interpret the variety in spelling and pronouncing of Lingua Franca words in the source texts as due to the vast expansion of the Lingua Franca itself, seen for instance in the different spellings of bono (Selbach, 2008: 46) or mucho (Cifoletti, 2004: 32). The variety of spelling and pronounciation could also serve as an indication of how many different mother tongue speakers acquired (and spoke) the Lingua Franca and that speakers of the Lingua Franca came to a great number of different ports in and outside of the Mediterranean. This further implies or, at least, allows the assumption of how easily understandable the Lingua Franca was among speakers of the Romance and Arabic languages and how widely spread the Lingua Franca must have been, which further, and most importantly, gives the impression that Jewish merchants, who are said to have known a great variety of different languages (Astren, 2014: 399), must have known and used the Lingua Franca. However, this, as well, does not necessarily mean that they integrated the Lingua Franca. There are no further reports that Jewish people used the Lingua Franca instead of Judeo-Arabic among themselves, instead they preferred to stay among themselves, only trading with others when unavoidable (Catlos, 2014 and Astren, 2014: 398- 399). As already stated, Jewish merchants were known for speaking numerous languages and thus they must have spoken the Lingua Franca too. However, speaking numerous languages may not count as an argument for learning and, more importantly, integrating another language, because knowing many languages can create the opportunity of using these languages. A Jewish merchant who spoke Arabic presumably would rather speak Arabic than the Lingua Franca with an Arab merchant, as well as the other way around because an Arab merchant would rather use Arabic than the Lingua Franca for communication if the Jewish merchant knew Arabic. However, considering that once the Lingua Franca spread across the Mediterranean this could have created the necessity to acquire and use the Lingua Franca for the Jewish merchants, as many other merchants would have used the Lingua Franca to trade. From an economic perspective, it seems unavoidable for a Jewish, or basically any, merchant to acquire the Lingua Franca since the 16 th century. That Jewish merchants spoke the Lingua Franca seems most likely. Thus the possibility that they, just like many Arabic-speaking merchants, would have adapted parts of the Lingua Franca as well seems plausible and while in use remaining very close to their own language. Such words could easily have been integrated into their native vocabulary.

The real problem is that this very widespread adaption of Lingua Franca words would suggest a

232 close connection which allowed the integration of the Lingua Franca into Judeo-Arabic. However, at the same time the fact that there was no communication problem within their community because communication among the Jewish community was established serves as a counter-argument because there would have been no reason at all to adapt (even parts of) another language. The Jewish people seem to be the only group to have integrated the Lingua Franca into their language, which is unique as no other culture seems to have done this (to such an extent). It remains questionable if Jewish people only acquired the Lingua Franca once it had become an integral part of commerce and the Mediterranean itself. The only definite aspect is that at some point there must have been a deep involvement and a close connection between the Lingua Franca and the Jewish people (which does not seem to be the case specifically with Jewish merchants). The most plausible assumption would be that the Jewish population itself acquired the Lingua Franca with the Lingua Franca leaving quite an impact. The possibility that this was initiated in the 16 th century when the Lingua Franca was present but not favoured and used only in its fixed Pidgin state in the North African commercial cities by the local population would seem less likely than the 19 th century when the Lingua Franca was very widespread among the Arabic-Turkish speaking population, which would also include the Jewish minority. Despite the spread of the Lingua Franca, the inevitable contact of the Lingua Franca and the Jewish (minority) population, and the Jewish merchants speaking the Lingua Franca too, it still seems implausible that the integration of another language would occur to such a high degree, especially since there has been no other case of such language integration. All in all, this theory (of integration) leaves some inconsistencies. However, the Lingua Franca and Judeo-Arabic definitely were involved and judging from the apparently deep influence there must have been a very close connection between the Lingua Franca and the considerable Jewish population for a longer period of time. Everything considered, the question may not be if the Lingua Franca influenced Judeo-Arabic but, perhaps, if it was the other way around, i.e. that Jewish people themselves created the Lingua Franca.

14.2 An alternative theory

The Judeo-Arabic language itself is not identified in the sources. Schuchardt (1909) has not identified a specifically Judeo-Arabic background and neither the documents analysed by Foltys

233 (1984) nor linguists such as Collier (1977), Lang (1992) or Selbach (2008) have analysed a specifically Judeo-Arabic background in the reports about the Lingua Franca. Schuchardt (1909) also does not identify Judeo-Arabic words which would invoke the assumption that Judeo-Arabic was involved in the creation, spread or development of the Lingua Franca. One could argue that Schuchardt (1909) only examined the late 19 th century Italian variety of the Lingua Franca. However, this is not the case as Schuchardt not only analysed the Italian variety, which probably was still present, but also the French variety presented in the Dictionnaire , the Spanish variety of Haedo and he was in correspondence with other people who knew the Lingua Franca themselves e.g. Jellinek. Furthermore, as Schuchardt (1979) states himself, he did study the Lingua Franca before in his youth but was not completely satisfied with the result and thus took on the study once again in later years; but he never mentions any connection to Judeo-Arabic (cf. Schuchardt, 1979: 26). Furthermore, Judeo-Arabic it is not mentioned as being an integral part of the Lingua Franca by any reports or documents. Here the only exceptions are Corré and Cifoletti, and it does strike one as odd that no Judeo-Arabic roots or other references or even mentionings are to be found in the reports, although there seem to be a large number of Lingua Franca words in Judeo-Arabic (such a large number, in fact, that influence in the opposite direction is quite conceivable and thus it could be assumed that Judeo-Arabic words were present in the Lingua Franca). After all, if a close connection existed at some point at least some linguistic traces should be visible, regardless if the involvement was short-lived or existed since the beginning. If the influence of Jewish people was present in the beginning and declined at some point or if Judeo-Arabic was an integral part of the Catalan variety and got replaced by the Spanish variety of the Lingua Franca, at least some traces could still be found in the Lingua Franca. A suitable example for this scenario would be the Portuguese influence that lasted only for a short amount of time but created an immense increase in Portuguese vocabulary (Haedo 1612 quoted in Foltys, 1984: 13 and Cifoletti, 2004: 198), and which remained visible centuries later (as seen with doublets). Consequently, the abundance of Lingua Franca words found in Judeo-Arabic might be explained not through the Lingua Franca influencing Judeo-Arabic but instead through a more complex scenario. This much more, if not the most, accurate interpretation is offered by Guido Cifoletti, who theorizes that it was, in fact, Tagarins, Sefardis and Moriscos, i.e. Jewish people themselves, who created the Lingua Franca. This theory would perfectly explain why parallels to the Lingua Franca can be found in the

234 Jewish language, i.e. words had not been integrated but were already part of the language of Tagarins, Sefardis and Moriscos. While Iberian-Romance languages were used to create the Lingua Franca there were still Jewish, or to be more precise, Judeo-Arabic words which found their way into the language. The connection to Judeo-Arabic exists not because these words were integrated by the Lingua Franca but, instead, because Judeo-Arabic was used as a substratum language to create the Lingua Franca itself, making the Lingua Franca not a Romance-Arabic but a Romance Judeo-Arabic Pidgin language. It is also important to note that it was expelled Iberian Jewish people who created the Lingua Franca, as they had extensive knowledge of Iberian Romance languages and their influence was strong enough to assimilate their language(s) among the Jewish communities they emigrated to. Consequently, the question to gather further evidence would have to be which words of the Tagarins, Sefardis and Moriscos entered the Lingua Franca (not which Lingua Franca words entered Judeo-Arabic). Furthermore, this would explain the close connection of the Lingua Franca to Old Catalan as the Tagarins were speakers of the Old Catalan language and this could further imply that the Tagarins had a strong influence in forming the Lingua Franca. However, due to the intelligibility with other Romance languages such as Spanish, the involvement of other Jewish people such as the Sefardis, who spoke Spanish, and also the Moriscos, can be explained as well. If Tagarins, Sefardis and Moriscos created the Lingua Franca together, many different aspects of different Romance languages would have been combined in the Lingua Franca, with a strong Catalan and Spanish influence, but also including other Iberian Romance dialects which became visible in different local varieties of the Lingua Franca due to the (forced) emigration of different Iberian Jewish people to several different locations within the Mediterranean in 1492. It is a most fascinating thought, maybe even the most valid theory, that the origin and spread of the Lingua Franca can be attributed to Jewish people. This would, of course, as predicted by Cifoletti (personal communication), mean that the Lingua Franca originated after the expulsion of the Jewish people, presumably soon after the beginning of the 16 th century due to many Jewish people having a thorough knowledge of Romance languages such as Catalan (Tagarins) or Spanish (Sefardis, including the Moriscos). This quite plausibly is also the reason why the Lingua Franca is so close to Old Catalan and Old Spanish, i.e. to Iberian Romance languages, and at the same time compatible with all other Romance languages as seen through the intelligibility of the Lingua Franca varieties. Furthermore, Jewish

235 people were in close contact with each other, and consequently a language connection was easily formed among them. Consequently, due to Iberian Jews heavily influencing the Jewish communities they emigrated to, it seems quite plausible that a Romance lexicon (for the Lingua Franca) would have been chosen.

14.4 Jewish merchants and Lingua Franca origin

Following this theory, Jewish merchants could have played a very important role in the process of Lingua Franca origin, but also its spread. Given that a part of the Jewish Mediterranean population that was living on the Iberian Peninsula before the expulsion of Jewish people in 1492 were merchants, there is a possibility that some of these Jewish people who created and spread the Lingua Franca across the Mediterranean belonged to this group of merchants. Jewish merchants themselves probably would have used the Lingua Franca when trading with other European merchants (not Arabic ones), which would also have contributed to the spread of the Lingua Franca, as Jewish merchants were travelling far distances. In this way Jewish merchants might also be the group that created, used and spread the Lingua Franca. Whether Jewish merchants actually created the Lingua Franca can not easily be answered, because whether they created the Lingua Franca is a question of language usage, or, to be more precise, what language they used when trading and communicating with other merchants. Jewish merchants were described as speaking many languages (Astren, 2014: 399) which would include both Catalan and Arabic. Consequently, the crucial issue would be whether they used the Catalan language, (Judeo-)Arabic or the Lingua Franca to communicate with other merchants. After all, an important pre-condition for the creation of the Lingua Franca was that a Romance language was involved, and consequently Catalan (Aragonese and/or Spanish) must have been used, as the Lingua Franca would not have originated if only Arabic had been used as a means of communication. The possibility that Jewish merchants used Arabic (with Arab merchants) is given, because why use Catalan at all if communication with the (Arab) merchants could be established in a mutual language? Furthermore, Jewish merchants were allowed to trade with Arab merchants before the re-establishment of trade with Christians, and thus Jewish merchants would probably have used Arabic as a means of communication ever since. Another aspect supporting the use of Arabic is that “Jewish minority communities became profoundly enmeshed linguistically, intellectually,

236 and theologically with Arabo-Islamic culture” (Catlos, 2014: 367), as indeed is further implied by the nature of Judeo-Arabic, which is Hebrew in Arabic script, strongly suggesting that a connection to Arabic existed (Catlos, 2014: 367). To further establish if Jewish merchants used Catalan or Arabic might be answered by a compatibility of Judeo-Arabic with the Arabic of the North African coast, which presumably existed. However, even if the compatibility was not given, Jewish merchants could have learned this dialect of Arabic easily or at least established communication in it in the context of trade. Furthermore, since Jewish merchants were the only (European) merchants that were allowed to trade with the Arab population before the 11 th century, Jewish merchants had the opportunity to venture to the North African coast. This happened before the Lingua Franca developed and consequently the Jewish merchants would have been unlikely to have used Catalan or the Lingua Franca (as both languages were either not spoken in these parts or did not exist at this point in time). Additionally, it would seem unlikely for someone who knows the mother language of another speaker to not use it, although there are exceptions, e.g. political rejection or social prejudice etc. Furthermore, Jewish merchants could also have traded only with other Jewish merchants. Summarising, Jewish merchants presumably used Arabic to communicate with Arabic-speaking merchants and thus did not create the Lingua Franca through language contact with Arabic speakers. However, using Arabic to communicate with North African merchants does not necessarily mean that Jewish merchants would use Arabic to communicate with European merchants. The question arises why Jewish people, presumably merchants – as contact with Europeans would have rather appeared in a commercial context – did not use the Catalan language itself, especially when they had a perfect knowledge of Romance languages, and instead used a Pidgin variety of a language that they already spoke. An important point is noted by Cifoletti (personal communication) stating that Jewish people refused to speak the language of their enemies, i.e. Romance languages, which definitely is to be traced back to the expulsion of the Jewish population from the Iberian Peninsula. This is the point when using a Pidgin instead of a full language does make perfect sense, as viewing this situation in a Pidgin-language context allows a full explanation of the origin of the Lingua Franca. The Lingua Franca was created by Jewish people communicating with Europeans using not their full Romance language competence, which they definitely had, but instead reducing it to a Pidgin language based on their own Judeo-Arabic language, i.e. only used a

237 Romance vocabulary with their own Judeo-Arabic grammar to show their enmity towards the Romance speakers who in turn considered it an insult that their language was being spoken in such an inadequate or pidginised manner. This would definitely be an argument against an early development of the Lingua Franca, as this historic event took place later in time, i.e. when Jewish people were expelled but still conducted trade. Consequently, before the 16 th century there would have been no reason to use a Pidgin as Jewish people did speak both Catalan and Arabic and would also have used both languages, since such an enmity against Europeans would not have prevailed before 1492. It would seem more logical that before the expulsion of the Jewish population from the Iberian Peninsula, despite previous degradations of Jewish people, there existed no reason for the formation of a Pidgin, at least on the side of Jewish people. Consequently, the Lingua Franca was not created by Arabic speakers but by Judeo-Arabic speakers. However, either due to the influence or perhaps usefulness of the Lingua Franca other Arabic speakers (or Arabic-speaking merchants) started using the Lingua Franca too. Presumably this was a development in the major commercial cities before people everywhere started using the Lingua Franca. That the Lingua Franca was adapted by a considerable number of speakers of different languages seems highly probable especially for Jewish, Arabic but also Romance speakers, especially merchants and sailors. The spread of the Lingua Franca presumably was also favoured by that fact that basically any merchant would have started using the Lingua Franca for business communication. Considering that the Lingua Franca was known in the whole Mediterranean, the only possibility which would allow the Lingua Franca this popularity and spread is that it developed in a time when Romance languages had become widely accepted as trade languages or influential languages in general in the Mediterranean. During such a time, people, especially merchants, would have gained huge benefits from speaking a Romance language or Pidgin, while during the early re-establishment of commerce and certainly before this event it would have been much more advantageous to speak Arabic. It would thus also seem possible that the Lingua Franca also became established due to European commercial dominance. In any case, once originated and with growing success, the Lingua Franca spread across the Mediterranean, and due to it becoming such an important or even infamous language many people would have acquired the Lingua Franca to communicate with people whose language they did not speak. Astonishingly, the increase in different Romance language speakers rather

238 increased the use and usefulness of the Lingua Franca, which also became closely connected to slavery to such an extent that the Lingua Franca seemingly became ubiquitous in the Mediterranean.

Chapter 15: Conclusion

The Lingua Franca as a language itself has died out in its long lasting form of being spoken throughout the whole Mediterranean and only its traces may (hopefully) live on in other languages. If a language like the Lingua Franca will arise once again is rather a political than a linguistic decision as the linguistic possibility is always present when two speakers of different languages communicate and perpetuate communication. It has always been political, and with it economic, factors that have created the necessary conditions for a Pidgin to develop and survive over time, as was the case with the Lingua Franca. After all, it was not linguistic complications that extinguished the Lingua Franca after an assumed existence of (almost) half a millennium, as communication could always be established in any basic or more complex form. However, political changes both supported, but in the end ousted, the Lingua Franca from the general Mediterranean landscape and thus were the true cause of the eclipse of this Pidgin language. The Lingua Franca, a (Judeo-)Arabic-Romance Pidgin, finally became replaced and lost in the 19 th century. The story of the Lingua Franca is a story of a long trepidation about its survival. Although many political changes were not in its favour, its sudden disappearance at the end of the 19 th century was still quite unexpected, as it had become an established part of the Mediterranean (its history and culture). The disappearance of languages, however, especially with Pidgin languages, does happen often and usually entails the rise of another (similar) language to take its place (as was the case with Sabir). English, so far, is the most promising lingua franca or Lingua Franca successor on a global level and might be taking the same place as a means of communication that the Lingua Franca had, as seen for instance in commerce, science and transportation.

239 Bibliography

Abulafia, D. 2014. “Thalassocracies.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 139-153.

Alleyne, M. C. 1998. Roots of Jamaican Culture . Worcester: Pluto Press.

Ardener, E. 1971. “Social Anthropology and the Historicity of Historical Linguistics.” In: E. Ardener (ed.). A.S.A. Monographs 10 – Social Anthropology and Language . London: Tavistock Publications Limited, 209-242.

Arends, J. 2002. “The Historical Study of Creoles and the Future of Creole Studies.” In: G. Gilbert (ed.). Studies in Ethnolinguistics 9 - Pidgin and Creole Linguistics in the Twenty-First Century . New York: Peter Lang Publishing Inc., 49-68.

Arends, J. 2003. “Reduplication in Lingua Franca.” In: S. Kouwenberg (ed.). Twice as Meaningful: Reduplication in Pidgins, Creoles and other Contact Languages . London: Battlebridge Publishers, 227-231.

Astren, F. 2014. “Jews.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 392-408.

Backman, C.R. 2014. “Piracy.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 170-183.

Blake, J. W. 1942. Europeans in West Africa, 1450-1560. Documents to illustrate the nature and scope of Portuguese enterprise in West Africa, the abortive attempt of Castillians to create an empire there, and the early English voyages to Barbary and Ghana. Volume I . (Trans. John William Blake). London: The Hakluyt Society.

Catlos, B.A. 2014. “Ethno-Religuoua Minorities.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 361-377.

Chaunu, Pierre. 1979. [1969] European Expansion in the Later Middle Ages. (Trans. Katharine Bertram). In: R. Vaughan (ed.). Europe in the Middle Ages – Selected Studies 10. Amsterdam/NewYork/Oxford: North Holland Publishing Company.

Cifoletti, G. 2000. “La lingua franca a Venezia nel settecento.” In: V. Orioles (ed.). Documenti letterari del plurilinguismo . Roma: Il Calamo, 9-18.

Cifoletti, G. 2004. La Lingua Franca Barbaresca . Roma: Il Calamo.

Collier, B. 1977. “On the Origin of Lingua Franca.” Journal of Creole Studies 1/2, 281-298.

De Granda, G. 1976. “A Socio-Historical Approach to the Problem of Portuguese Creole in West Africa.” Linguistics 173, 11-22.

240 Dickins, J., Harvey, S., Higgins, I. 2002. Thinking Arabic Translation . London/New York: Routledge.

Doumanis, N. 2014. “The Mediterranean and Asia.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 441-456.

Foltys, C. 1984. “Die Belege der Lingua Franca.” Neue Romania 1, 1-37.

Gertwagen, R. 2014. “Nautical technology.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 154-169.

Greene, M. 2014. “The Early Modern Mediterranean.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 91-106.

Hall, R. A. 1966. Pidgin and Creole Languages . New York: Cornell University Press.

Harvey, L.P., Jones, R.O., Whinnom, K. 1967. “Lingua Franca in a Villancico by Encina.” Revue de littérature compare 41, no. 4, 572-579.

Hancock, I. F. 1977. “Recovering Pidgin Genesis: Approaches and Problems.” In: A. Valdman (ed.). Pidgin and Creole Linguistics . Bloomington & London: Indiana University Press, 277-294.

Herzfeld, M. 2014. “Po-Mo Med.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 122-136.

Holm, J. 1989. Pidgins and Creoles . Cambridge: Cambridge University Press.

Huber, M. 1999. Ghanian Pidgin English in its West African Context . Amsterdam/Philadelphia: John Benjamins Publishing Company.

Kahane, H., Kahane, R., Tietze, A. 1958. The Lingua Franca in the Levant – Turkish Nautical Terms of Italian and Greek Origin . Urbana: University of Illinois Press.

Kea, R.A. 2014. “The Mediterranean and Africa.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 425-440.

Lang, G. 1992. “The Literary Settings of Lingua Franca (1300-1830).” Neophilologus 76, 64-76.

Le Page, R. B. 1961. “Conference on Creole Language Studies.” In: R. B. Le Page (ed.). Creole Language Studies II – Proceedings of the Conference on Creole Language Studies . London: Macmillan &Co Ltd, 123-128.

Mallette, K. 2014. “Lingua Franca.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 330-344.

Pollard, E.A. 2014. “The Mediterranean and the Indian Ocean.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 457-474. Ruiz, T.F. 2014. “The Mediterranean and the Atlantic.” In: P. Horden, S. Kinoshita (eds.). A 241 Companion to Mediterranean History . West Sussex: Wiley Blackwell, 411-424.

Schuchardt, Hugo. 1909. “Die Lingua Franca.” Zeitschrift für Romanische Philologie 33, 441- 461.

Schuchardt, Hugo. 1979. [1909] On Lingua Franca. (Trans. Thomas L. Markey). In: Markey, T. L. (ed.). The ethnography of variation: Selected writings on pidgeons and creoles . Ann Arbor: Karoma, 26-47.

Selbach, R. 2008. “The Superstrate is not always the Lexifier: Lingua Franca in the Barbary Coast 1530-1830.“ In: S. Michaelis (ed.). Roots of Creole Structures - Weighing the Contribution of Substrates and Superstrates . Amsterdam/Philadelphia: John Benjamins Publishing Company, 29-58.

Thompson, R. W. 1961. “A Note on some Possible Affinities between the Creole Dialects of the Old World and those of the New.” In: R. B. Le Page (ed.). Creole Language Studies II Proceedings of the Conference on Creole Language Studies . London: Macmillan &Co Ltd, 107- 113.

Valérian, D. 2014. “The medieval Mediterranean.” In: P. Horden, S. Kinoshita (eds.). A Companion to Mediterranean History . West Sussex: Wiley Blackwell, 77-90.

Voorhoeve, J. 1961. “A Project for the Study of Creole Language History in Surinam.” In: R. B. Le Page (ed.). Creole Language Studies II – Proceedings of the Conference on Creole Language Studies . London: Macmillan &Co Ltd, 99-106.

Wansbrough, J. E. 1996. Lingua Franca in the Mediterranean . Surrey: Curzon Press.

Whinnom, K. 1965. “The Origin of the European-based Creoles and Pidgins.” Orbis 15, 509-527.

Whinnom, K. 1971 “Linguistic Hybridization and the ´Special Case´ of Pidgins and Creoles.” In: D. Hymes (ed.). Pidginization and Creolization of Languages – Proceedings of a Conference held at the University of the West Indies. London: Cambridge University Press, 91-116.

Whinnom, K. 1984. “Lingua Franca: Historical Problems.” In: A. Valdman (ed.). Pidgin and Creole Linguistics . Bloomington: Indiana University Press, 295-310.

Wiener, N. 1948. Cybernetics . Cambridge, Mass.: MIT Press. 1961 (2nd. ed.)

242 Webliography

Alfajarín, J. R. 2000. Ésser, estar i haver-hi en Català antic . https://books.google.at/books?id=3LRN34zpFa8C&pg=PA449&lpg=PA449&dq=%C3%A9sser, +estar+i+haverhi+en+catal%C3%A0+antic&source=bl&ots=gtcyevrRmL&sig=Jo_4D1oHj0lQ7 z69y_oshHhvMCo&hl=de&sa=X&ved=0CCEQ6AEwAGoVChMI88rcj_OWyAIVKABzCh3sv wR5#v=onepage&q=%C3%A9sser%2C%20estar%20i%20haverhi%20en%20catal%C3%A0%2 0antic&f=false (accessed 28 September 2015)

Castellanos, C. 2010. [2006]. La Lingua Franca. Consideracions critiques . (Trans. Alan D. Corré) https://pantherfile.uwm.edu/corre/www/franca/edition3/Castellanos.html (accessed 13 November 2010)

Carreras, C. & Yates. A. 2009. “The Catalan Language”. In: C. Carreras (ed.). The Architect of Modern Catalan: Selected Writings/Pompeu Fabra (1868-1948) . https://books.google.at/books?id=BKvQg1323YkC&printsec=frontcover&hl=de#v=onepage&q &f=false (accessed 28 September 2015)

Ceccarelli, G. 2007. “The Price for Risk-Taking: Marine Insurance and Probability Calculus in the Late Middle Ages”. Journ@l Electronique d`Histoire des Probabilities et de la Statistique Vol. 3 n o1. http://www.jehps.net/Juin2007/Ceccarelli_Risk.pdf (accessed 28 September 2015)

Chaytor, H. J. 1933. A History of Aragon and Catalonia . http://libro.uca.edu/chaytor/achistory.htm (accessed 25 September 2015)

Corré, A. D. 2005. A Glossary of Lingua Franca . https://pantherfile.uwm.edu/corre/www/franca/go.html (accessed 22 September 2015)

Dante, A. 1304. De Vulgari Eloquentia . https://de.wikisource.org/wiki/De_vulgari_eloquentia (accessed 22 September 2015)

Fischer, S. 2003. The Catalan Clitic System. A Diachronic Perspective on its Syntax and Phonology . https://books.google.at/books?id=8KTpDSzscNwC&pg=PA23&lpg=PA23&dq=catalan+clitic+sy stem&source=bl&ots=BI9xKqMag&sig=g7VsAZ19YeHCpnU2WSmGq125hdU&hl=de&sa=X &ved=0CCsQ6AEwAWoVChMInc7qqe6WyAIVRNwsCh3AjgUx#v=onepage&q=catalan%20cl itic%20system&f=false (accessed 28 September 2015)

Forster, J. 1883. The Chronicle of James I, King of Aragon Surnamed The Conqueror (T. L. John Forster) http://libro.uca.edu/chronicleofjames/chronicle.htm (accessed 28 September 2015)

Generalitat de Catalunya, Diversification of the language's uses . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/index619b.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=72f4edfc49ed7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

243 Generalitat de Catalunya. Territorial expansion and contact with other languages . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/index4dbc.html?vgnextoid=23885c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=23885c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=28d4edfc49ed7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. The birth of Catalonia. The Feudal and Counts' society . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexe277.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRCR D&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&co ntentid=045d3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. Old Catalan. The beginning of Catalan . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/index69e4.html?vgnextoid=77595c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=77595c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=9db4edfc49ed7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. The expansion of the Barcelonan dynasty throughout the Mediterranean . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexce40.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRCR D&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&co ntentid=298d3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. The conquest of the New Catalonia, the Kingdom of Valencia and Majorca . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexb767.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=9e6d3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. Muslims in Catalonia . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexd53b.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=8c1d3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. Influence of other languages: Germanic, Mozarabic, Hebrew . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexddf3.html?vgnextoid=23885c43da896210VgnVCM1000000b0c1e0aRCR D&vgnextchannel=23885c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&c ontentid=1174edfc49ed7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. The lower Middle Ages' crisis. The Compromise of Caspe. The Trastámara dynasty . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/index2c12.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRCR 244 D&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&co ntentid=b3ad3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. The Catalan State. The institutions. 'Cort General' and 'Diputació del General' (or Generalitat) . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/index7b79.html?vgnextoid=841c5c43da896210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=841c5c43da896210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2& contentid=70ed3c084ded7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Generalitat de Catalunya. Chronicles . http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.be2bc4cc4c5aec88f94a9710 b0c0e1a0/en_GB/indexeb3f.html?vgnextoid=b619d5e5d74d6210VgnVCM1000000b0c1e0aRCR D&vgnextchannel=b619d5e5d74d6210VgnVCM1000000b0c1e0aRCRD&vgnextfmt=detall2&c ontentid=80b7110e279d7210VgnVCM1000008d0c1e0aRCRD&newLang=en_GB

Igual, V. 2003 El Català de l'Alguer: un Model d'àmbit restringit . http://www.iec.cat/butlleti/pdf/103_butlleti_model.pdf (accessed 28 September 2015)

Kitchin, D. B. 1987. Old Occitan (Provençal)- English Glossary . http://ieoparis.free.fr/delo/Darcy%20BUTTERWORTH%20KITCHIN%20- %20Old%20Occitan%20(Proven%E7al)-English%20Glossary.pdf (accessed 28 September 2015)

Lindsay, R. 2009. A Reclassification of the Occitan Language . https://robertlindsay.wordpress.com/2009/05/24/a-reclassification-of-the-occitan-language/ (accessed 28 September 2015)

Loos, E. (ed.). 2003. Glossary of Linguistic terms http://www-01.sil.org/linguistics/GlossaryOfLinguisticTerms/Index.htm (accessed 28 September 2015)

Mateu. J. 2009. “Gradience and Auxiliary Selection in Old Catalan and Old Spanish”. In: Dins P. Crisma & G. Longobardi (eds.). Historical Syntax and Linguistic Theory . http://filcat.uab.es/clt/publicacions/reports/pdf/GGT-06-20.pdf (accessed 22 September 2015)

Monk, J. 2013. Catalan language. In: E. Michael Gerli (ed.). Medieval Ibera An Encyclopedia . https://books.google.at/books?id=euVJAgAAQBAJ&printsec=frontcover&dq=Medieval+Iberia+ an+encyclopedia&hl=de&sa=X&ved=0CCsQ6AEwAGoVChMIm- Kjw5nyxwIVx7QUCh2n1wyL#v=onepage&q=Medieval%20Iberia%20an%20encyclopedia&f=f alse (accessed 25 September 2015)

Pape, R. 1883. [1859] Die Wortstellung in der provençalischen Prosa-Literatur des XII. und XIII. Jahrhunderts . https://archive.org/details/WorstellungProvProsaLittPape (accessed 28 September 2015)

Rosembach. J. 1502. Vocabolari Molt Profitos per Apendre Lo Catalan Alamany y Lo Alamany Catalan . https://archive.org/details/vocabularicatal00barn (accessed 27 September 2015)

245 Rossetti, R. 2005. An Introduction to Lingua Franca . https://pantherfile.uwm.edu/corre/www/franca/edition3/lingua5.html (accessed 13 November 2010)

Scrivner O. and Kübler S. 2012. “Building an Old Occitan corpus via cross-Language Transfer”; Proceedings of KONVENS 2012 (LThist 2012 workshop) http://www.oegai.at/konvens2012/proceedings/58_scrivner12w/58_scrivner12w.pdf (accessed 28 September 2015)

Soler Llopart, A. 2003 [1999]. Literatura Catalana medieval: Un recorregut multimedia pels grans autors i els seus textos . http://www.editorialuoc.com/extra_content/84-7306-864- 5//web/index1.html

The Consulate of the Sea. 1494. http://www.lluisvives.com/servlet/SirveObras/jlv/09255007699869439647857/index.htm (accessed 28 September 2015)

Wikipedia. Catalan Verbs . https://en.wikipedia.org/wiki/Catalan_verbs (accessed 28 September 2015)

Wikipedia. Mediterranean Lingua Franca : https://en.wikipedia.org/wiki/Mediterranean_Lingua_Franca (accessed 28 September 2015)

Wikipedia. Occitan Conjugation . https://en.wikipedia.org/wiki/Occitan_conjugation (accessed 28 September 2015)

Zago, R. 2005. A Dissertation on Lingua Franca . https://pantherfile.uwm.edu/corre/www/franca/edition3/lingua6.html

(Lafont, R.) 2004. Lexic Occitan Medieval (òc francés) . http://lengadoc.chez.com/lexic_medieval.htm (accessed 28 September 2015)

246 Further Reading

Catalan Language http://www.orbilat.com/Languages/Catalan/catalan.html (accessed 28 September 2015)

Encyclopedia Catalana http://www.enciclopedia.cat/ (accessed 28 September 2015)

Generalitat de Catalunya http://www.gencat.cat/culturcat/portal/site/culturacatalana/menuitem.9d62b5d2eb879f88f94a971 0b0c0e1a0/en_GB/indexfee4.html?vgnextoid=ba29ebd463996210VgnVCM1000000b0c1e0aRC RD&vgnextchannel=ba29ebd463996210VgnVCM1000000b0c1e0aRCRD&newLang=en_GB (accessed 28 September 2015)

IEC - Institut d´Estudies Catalan http://www.iec.cat/activitats/entrada.asp (accessed 28 September 2015)

Italian Language www.orbilat.com/Languages/Italian/italian.html (accessed 28 September 2015)

Languedoc - The Troubadours www.languedoc-france.info/1904_troubadours.htm (accessed 28 September 2015)

LIBRO - The Library of Iberian Resources Online http://libro.uca.edu/title.htm (accessed 28 September 2015)

Occitan Language http://www.orbilat.com/Languages/Occitan/Grammar/index.html (accessed 28 September 2015)

Old Occitan http://www.lingweenie.org/occitan/ (accessed 28 September 2015)

Rialto - Repertoririo informatizzato dell´antica letteratura trobadorica e occitania. http://www.rialto.unina.it/ (accessed 28 September 2015)

Romance Languages orbilat.com/Languages/index.html (accessed 28 September 2015)

Trobar - Troubadours http://www.trobar.org/troubadours/ (accessed 28 September 2015)

247