A Non-Passive Markedness Process in Persian a Functional Sentence Perspective Approach Mohammad Jafar JABBARI

Full Length Research.

A Non-passive Markedness Process in Persian A Functional Sentence Perspective Approach Mohammad Jafar JABBARI

Department of English, Yasouj University, Yasouj, Iran

Accepted 20 th October 2012

In Persian, agentive passive sentences are not formed normally and frequently. The passive voice is used only when the doer of the action is not present for any reason. On the other hand, in Persian, unlike English and many other European languages, case is assigned to NPs overtly and by affixation. This article aims to find the correlation between case assignment and passive formation, to discover the reason(s) behind the absence of agentive passive structure in Persian, and to introduce a well-formed alternative for agentive passive sentences, within the framework of Functional Sentence Perspective (FSP).

Key Words : Aagentive passive voice, Markedness, FSP,Theme and Rheme.

1. Introduction produced in Persian and d) introduce an alternative markedness process for marking Persian sentences While agentive passive sentences are formed naturally and frequently in English, they, according to many 2. Review of the Related Literature Persian grammarians, are not naturally and frequently produced in Persian. The passive voice is used only 2.1 Case Theory when the doer of the action is not mentioned for any Case is an “inflectional category, basically of nouns, reason. On the other hand, while in English and some which typically marks their role in relation to other parts of European languages, case is determined syntactically, in the sentence" Matthews (2007). In other words, it is “a Persian, it is assigned morphologically. Analyzing the grammatical category that shows the function of the noun markedness process of passivization within the or noun phrase in a sentence" (Richards and Schmidt, Functional Sentence Perspective (FSP) approach, and 2002). elaborating the above-named two different ways of case Among the most frequently distinguished cases are assignment, this article intends: a) introduce such nominative, accusative and dative, respectively assigned concepts as case assignment, markedness, etc., b) find to the subject, the direct object and the indirect object or the correlation between these concept, c) discover the object of preposition. Case assignment in sentences with reason(s) why agentive passive sentences are not different numbers of NP arguments is illustrated in the following table: (1)

Subject Verb DO IO/OP NOMINATIVE ACCUSATIVE DATIVE a Maryam went - -

b Zhaleh invited Maryam -

c Jamshid gave the flowers to Maryam

024 Glo. J. Edu. Res.

The NP Maryam in (1a), (1b) and (1c) has nominative, 1981). From this viewpoint, an active sentence and its accusative and dative case respectively, while they, lacking corresponding passive one, though identical in the basic case endings, are equal in form. Case Theory goes far meaning, are uttered in two different situations and cannot beyond case endings of nouns. "It deals not just with case be interchangeably used. In other words, they are suitable forms visible in the surface sentence but with 'abstract' answers to two different questions. For example (2 repeated Case."(Cook and Newson, 1997). For some scholars case as 4b) is a suitable answer to (4a): assignment is a language universal. Chomsky (1986a) maintains that "every phonetically realized NP must be (4) a. What did Maryam eat? assigned (abstract) case." For Cook and Newson (1997), b. Maryam ate the cake. "abstract case is an important element in the syntax even when it does not appear in the surface. The fact that Maryam Sentence (3, repeated as 5b) is, however, a felicitous lacks any visible case ending in the above examples, does answer to the totally different question (5a): not mean that it lacks case. While case is a universal (5) a. By who(m) was the cake eaten? concept, case assignment processes is language-specific. b. The cake was eaten by Maryam. Case assignment differences in English and Persian will be elaborated in due course. Sentences (4b) and (5b), are grammatical, acceptable and semantically equal. They, nevertheless,are not felicitous 2.2. Voice answers to (4a) and (5a) respectively: Voice is a linguistic category which indicates whether the surface subject of a sentence is the performer of the action (6) a. What did Maryam eat? or is acted upon. Trask (2007) defines voice as “the b. ? The cake was eaten by Maryam. grammatical category governing the way the subject of a sentence is related to the action of the verb. In other words, (7) a. By who(m) was the cake eaten? voice is “the way in which a language expresses the b. ? Maryam ate the cake. relationship between a verb and the noun phrases which are associated with it” (Richards and Schmidt 2002). FSP describes how information is distributed in In many languages the most familiar voice contrast is that sentences: a sentence conveys two types of information, between active and passive constructions. “A verb is in the namely old (given) and new information. The old information, active voice when the subject of the verb actually performs referred to as theme , is that knowledge which the speaker the action indicated by the verb….. A verb is in the passive assumes to share with the addressee. Theme is “the starting voice when it expresses an action performed upon its point of utterance” (Mathesius, 1942), or it is “the point of subject” (Shaw, 1986). In other words, passive is "a departure for what the speaker is going to say" construction in which an intrinsically transitive verb is (Halliday,1985). The new information, referred to as rheme , construed in such a way that its underlying object appears is what the speaker states about, or in regard to, the theme. as its surface subject" (Trask, 1993). Two sentences can be In other words, rheme is "everything else that follows in the different in voice and yet convey the same basic meaning. sentence" (Brown and Yule, 1989). In still other words, Consider the two sentences (2) and (3): rheme is the information that the speaker assumes not to be (2) Maryam ate the cake. inferable by the addressee from the text. Theme is the (3) The cake was eaten by Maryam. information repeated in both the question and the answer, and what appears only in the answer is rheme. Sentence (2) is an active sentence. Sentence (3) is the corresponding passive of the active sentence (2). They are (8) equal in the basic meaning. If (2) and (3) are equal in the basic meaning, then in what they do differ and why should two semantically identical Theme Rheme sentences be formed? Different linguistic approaches may give different answers to these questions. The Functional a. What did Maryam eat ? b. Maryam ate the cake. Sentence Perspective (FSP) has a pragmatic answer to this question.

2.3. Functional Sentence Perspective (9) The Functional Sentence Perspective (FSP) is a type of analysis associated with the Prague School of linguistics Theme Rheme which studies the sentence in terms of the part its elements play in communication. According to FSP "the structure of a. By who(m) was the cake b. The cake was eaten utterances is determined by the use to which they are put eaten ? Maryam and the communicative context in which they occur" (Lyons,

Jabbari 025

In (8b), Maryam ate , repeated in both the question and the 2.5.1 Formal Difference answer, is the theme and the cake, which appears only in A translation of sentence (2) into Persian, can be of help in the answer, is the rheme. It can be concluded that, the short comparing passive voice in Persian and English : answer to a wh-question is the rheme but the complete (10) /mærjæm kejK-ra xord/ answer contains both theme and rheme. By the same token, Maryam the cake-ACCmarker ate in (9b), The cake was eaten is the theme and (by) Maryam is the rheme. Thus, according to FSP, an active sentence and Passive sentences in Persian, like English (and many other its corresponding passive sentence differ with regard to the languages), are formed with passive auxiliary and passive theme-rheme distribution: what is the theme in an active participle. Sentence (11) is an example of a Persian passive sentence can be (totally or partially) the rheme in the sentence : corresponding passive sentence and vice-versa. ( 11) / kejk xord-e ʃod / 2.4. Markedness the cake eaten became A linguistic form is said to be marked if it is " less central or less natural than a competing one (unmarked form) on any (11) is a translation of the following English passive of various grounds, such as lower frequency, more limited sentence (12): distribution, more overt morphological marking, greater semantic specificity or greater rarity in languages generally" (12) The cake was eaten. (Trask 1993). Thus, “a passive sentence like Janet was arrested by the police is marked with respect to the active Such passive structures as (12), in which the agent is not The police arrested Janet, since the passive contains more included, are called “agent-less” or “short” passive material, has a more complex structure, and is rarer than the sentences. Passive sentences, which do include the agent active”(Trask, 2007). or doer of the action, as (3), are referred to as “agentive” or The markedness of a sentence can be determined by its “long” passive sentences. While an agentive passive word order. Richards and Schmidt (2002) introduce the conveys the same amount of information as its Markednss Theory as: corresponding active sentence, in an agent-less passive in languages of the world certain linguistic sentence one constituent, i.e the agentive PP ( by phrase in elements are more basic, natural, and frequent English), and consequently one piece of information (i.e. the (unmarked) than others which are referred to as agent or performer of the action) is ignored. “marked”. For example, in English, sentences which have the order: Subject-Verb-Object: I dislike such people . Are considered to be unmarked, 2.5.2. Functional Differences whereas sentences which have the order: Object- Agent-less passive sentences are very common in many Subject-Verb: Such people I dislike. Are languages. They are formed when the agent or the doer of considered to be marked (P.320). the action is absent in the sentence, because it is unknown, unimportant or predictable, or for any other reason. On the By the same token, an English sentence like (2), which has other hand, long passive sentences, in which agent is also the basic Subject-Verb-Object word order, is regarded as included, are formed to serve another function. They are "unmarked" and any sentence which violates this basic word formed to chane the theme-rheme organization of the order, like (3) is marked. sentence. For FSP, any process which changes the theme-rheme There is a difference of paramount importance here: organization of the sentence is referred to as a “markedness unlike English, agentive passive sentences like (13) are not process” (Andersen 1989), passivization being the most produced in Persian. productive one. Some other markedness processes, the elaboration of which falls beyond the scope of this study, are (13) ?* /kejk be-væsile-je mærjæm xorde ʃod / Topicalization, Inversion, Clefting, to name but a few. Within the cake by- means-of Maryam eaten became one language, structurally different sentences may undergo different markedness processes. Moreover, different Sentences like (13) though not absolutely ruled out in languages may utilize different markedness processes in the Persian, are regarded by many scholars as unnatural and same situation. In other words, whereas markedness is a consequently ill-formed. Such sentences, if any, are universal concept, markedness processes are language- produced via incompetent word-for-word translation, mostly specific. from European languages. It is worth mentioning that, in Persian, inspired by Arabic grammar, the active voice is 2.5. Persian and English Passive Sentences referred to as /mæ?lum/ (literally meaning known) and Passive sentences in Persian and English differ in a number passive voice as /mæ hul/ (literally meaning unknown ). of ways. They are specially different formally and This naming is far from accidental. There is almost a functionally. consensus among Persian grammarians, both native and

026 Glo. J. Edu. Res.

non-native, that in Persian a passive or /mæ hul/ (literally and simple in nature, represents a standard level of meaning unknown) sentence is only formed when the language. Simple examples are deliberately chosen for the performer of the action is unknown, or not intended to be convenience of the discussions and argumentations. known for any reason. Persian grammarians have always argued that the passive in Persian is an agent-less 4- Discussions construction. Lazard (1957) believes that "the usage of the passive In the previous sections, case assignment was introduced in voice is specially limited to the occasions in which the doer brief. Now, consider (14): of the action is not mentioned 1. Bahar et al ., (1963) assert that "the passive verb is called (14) Zhaleh called Maryam. mæ hul/ (unknown) as its agent is unknown 2". Lambton (1966) maintains that ” the Passive Voice is not Zhaleh is the subject and Maryam is the direct object of the used in Persian, if the active voice can be used. sentence. Although the two NPs do not have any (overt) For Najafi (1987), there is “an important difference case markers, they have nominative and accusative case between Persian, on the one hand, and European respectively. This case assignment becomes overtly languages, on the other, namely a passive verb is used only manifest if the NPs are nominalized: when the agent is unknown or is not supposed to be known 3. (15) She called her. Shari’at (1988), divididing the transitive verb into active and passive, adds that “ if the agent of a verb is known, the In (15), unlike in (14), the two NPs, s he and her are in verb is active, otherwise it is passive 4” . contrast formally. Interchanging the two NPs results in an Natel (1991) believes that in Persian passive sentences, ungrammatical sentence: "verb is attributed to the receiver of the action; the doer of (16) * Her called she. the action is not mentioned, that is to say it is unknown. 5". For Vahidian (1992), “ in Persian, the passive verb does Sentence (16) is ungrammatical since in English, the not have much usage, whereas in such languages as canonically pre-verbal NP has the nominative case, whereas English, a sentence can be used in the passive voice, even if the cononically (immediately) post-verbal NP has the the doer of the action is known….. In Persian, the verb is accusative case. Thus, the pre-verbal slot must be occupied used in the passive voice, if and only if the doer of the action by a nominative NP, and the immediately post-verbal slot is unknown, or is not intended to be known (for any reason), must be occupied by an accusative NP, as in (15), not vice- or is obvious 6” versae. In other words, what determines the case of NPs, in If in Persian only agent-less passive sentences are English and similar languages, is the word order. Thus, formed, then how English sentences like (3) are translated changing the word order, results in either ungrammatical into Persian, while it is believed that "anything that can be (17b) or semantically different sentences (18b, 19b &): said in one language can be said in another, unless the form (17) a. Maryam ate the cake ⇒ b.* The cake ate Maryam is an essential element of the language" (Nida 1982). What is aimed here, regardless of translational issues (18) a. Zhaleh called Maryam ⇒ b. Maryam called Zhaleh. which are beyond the scope of this article, is to answer two (19) a. You can call Maryam. ⇒ b. Maryan can call you. questions: (i) while in Persian agent-less passive sentences are made frequently and naturally, why are agentive (long) Sentences (17 to 19) show that, in English, it is the word passives not naturally made, and (ii) what is the natural order that determines and assigns case, and nothing else. In well-formed alternative for a Persian agentive passive still other words, case in English and similar languages is sentence? abstract and is assigned syntactically. In Persian, unlike English, case markers do have 3. Data Analysis phonetic manifestation. In other words, Persian benefits from overt case markers. To illustrate this, sentence (10), which is The data for the study are simple sentences of the type of the Persian translation of the English sentence (2) will be re- those found in elementary Persian and English language examined as (20). Morpheme boundaries within the NPs, are textbooks. The data corpus, being very limited in number hyphenated, and case markers are underlined for more convenience: (20) / mærjæm kejk-ra xord/ 1 - My translation from French. Maryam the cake-ACCmarker ate 2 - My translation from Persian. 3 In (20), the postposition /ra/ attached to /kejk / (the cake), is - My translation from Persian. 4 the accusative marker. - My translation from Persian. 5 - My translation from Persian 6 - My translation from Persian.

Jabbari 027

What is the nominative marker in Persian? Nominative In (22) , the direct object /kejk-ra/ ( the cake) precedes the marker, in Persian does not have any phonetic subject /mærjæm-ø/ (Maryam). Thus, (22), having the OSV manifestation. This, however, does not mean that nominative word order, is marked. It is in contrast with (21) which has NPs in Persian are not marked. The nominative marker in the unmarked SOV word order. Sentences (21) and (22) Persian is a null element. A null element is "an abstract unit have the same basic meaning. However, they are felecitous with no physical realization in the stream of speech; The answers to two different questions: term is commonly used for the absence of a morpheme in the context where one would normally occur" (Crystal, 1992). ((23) a. / mærjæm-ø ʧi-ra xord /? It is said that "most approaches to grammatical description Maryam-NOM what-ACC ate recognize the existence of at least some null b. / mærjæm -ø kejk-ra xord / elements"(Trask, 1993). A null elements is symbolized as /- Maryam-NOM cake-ACC ate ø/. For instance "the plural morpheme is accordingly said to be realized by a zero morph: sheep-ø" (Matthews, 2007). (24) a. / kejk-ra ki-ø xord / ? This representation facilitates linguistic discussions and cake-ACC who-NOM ate argumenatations, and is frequently used, even, by rival b. / kejk-ra mærjæm-ø xord / schools of linguistics. In Persian, the nominative case cake-ACC Maryam-NOM ate marker /-ø/, though not phonetically realized, exists in the minds of native speakers, as part of their (sub-conscious) Sentence (24b), is a well-formed and naturally made linguistic knowledge, helping them differentiate nominative sentence in Persian. It is not passive, yet it is marked. It has NPs from accusative NPs, which are represented by the the same theme-rheme organization as the English passive post-position /ra/. In other words,in Persian, any NP followed sentence The cake was eaten by Maryam. by the post-position /ra/ is accusative, whereas any NP with no (overt) marker is nominative. Sentence (20) is (25) reconstructed as (21), using /-ø/ as the nominative case marker: (21) / mærjæm-ø kejk-ra xord/ Theme Rheme

Maryam -NOMmarker the cake-ACCmarker ate a.By who(m)was the cake eaten? b.The cake was eaten by Maryam It is worth mentioning that any absence of an affix is not necessarily a zero morpheme. A zero morpheme must always be in contrast with one or more non-zero morphemes. Thus, the lack of case markers in English is not comparable to nominative zero marker in Persian. The (26) difference is that, in English unlike Persian, both nominative 8 and accusative NPs lack markers, so they are not in Theme Rheme Theme contrast. In Persian, on the contrary, the nominative (/-ø/) a. /kejk -ra ki-ø xord /? b. kejk -ra mærjæm-ø xord / marker is in contrast 7 with the accusative case marker /ra/ . cake-ACC who-NOM ate cake-ACC Maryam-NOM ate Therefore, Persian NPs are not word order bound and can move flexibly within the sentence, as the case marker /ra/ is always attached to accusative NPs and differentiates them Sentence (26b) has the same amount of information and the from nominative NPs. same theme-rheme organization as (25b). Thus, (26b) can The above difference has resulted in the fact that the be a natural well-formed Persian counterpart (i.e. translation) Persian sentence word order is much more flexible than the of (25b). Thus, such sentences as (13, repeated in 27): English sentence. Sentence (21), having the word order 9 SOV, is a typical unmarked sentence in Persian. Sentence (27) * /kejk be-væsile-je mærjæm xord-e ʃod / (22), however, has a different word order: the cake by- means-of Maryam eaten became (22) / kejk-ra mærjæm-ø xord/ the cake-ACC Maryam-NOM ate are not formed in Persian, as it is not needed. The relationship between the existence of overt case markers and lack of agentive passive in Persian is far from 7 - In linguists and semiotics lack of a sign can be regarded as accidental. The linguistic principle of the 'least effort', a sign itself. For example ⇧is used to show a one way street. The absence of this sign is, as a psychological reality in the 8 - If the Theme is split by the Rheme, for any reason, it is driver’s mind, indicative of a two-way street. referred to as “Discrete Theme”. 9 - /xord-e/ comprises of the past stem/xord-/ and past participle suffix /-e/.

028 Glo. J. Edu. Res.

introduced by Zipf (1949), asserts that "speakers do not Chart (2) exert themselves more than is necessary for successful communication"(Matthews, 2007). This principle indicates that other things being equal, the simpler and shorter a 100.00% linguistic form, the more frequently human beings tend to 80.00% use it (Crystal, 2002). It is said that an active sentence could 60.00% be a better choice than its passive corresponding sentence, Active 40.00% because: 20.00% Passive Active voice sentences are often more concise than passive voice. Expressing the same idea in passive voice 0.00% takes 30% to 40% more words: Phonemes Morphemes Words The fighter punched Ali and dodged the uppercut . (Active voice-8 words) Ali was punched by the fighter and then an uppercut was dodged by him . Furthermore, the active verb/xord/(ate) in (28) is (Passive voice-14 words, about 40% longer) qualitatively (structurally) much simpler than the passive (Passive Voice : 1) 10 verb /xord-e ʃod/ (eaten became) in (29). Thus, marked active sentences like (27) could be better choices than The same calculations can be applied to the afore-said such unnatural passive sentences as (29), as are much pairs of active and passive sentences (26b and 27), simpler both quantitatively and qualitatively. Having the repeated as (28) and (29) respectively: same theme-rheme organization, as the English sentence (25b), sentence (28) is marked and can be used in the same context. So, the formation of (28) kejk-ra mærjæm xord / sentences like (29) is blocked. the cake Maryam ate 5. Concluding Remarks (29) * /kejk be-væsile-je mærjæm xord-e ʃod / the cake by- means-of Maryam eaten became It can be inferred from the foregoing that (i) the agentive and agent-less passive sentences serve two different functions. Sentence (28) includes 16 phonemes, 4 morphemes and 4 The former is made only when the agent is not mentioned for words 11 , while sentence (29) consists of 28 phonemes, 8 any reason, but the latter changes the theme-rheme morphemes and 6 words. Thus, the active sentence (28) is, organization. (ii) In English, case is assigned syntactically, phonogically approximately 42.80 %, morphologically i.e. the word order determines the case, whereas in Persian 50.00% and lexically almost 33.00%, shorter than (29). The case is assigned morphologically, i.e. via overt case above-mentioned quantities and percentages are illustrated markers. An unmarked SOV Persian sentence can be in Charts (1) and (2), respectively: converted to the marked OSV, without any need for passivization, as the ACC case marker /-ra/ is always attached to the object NP, and differentiate it from the Chart (1) subject NP, which lacks any overt marker. Consequently (iii), agentive passives are not naturally formed in Persian, as (iv) the theme-rheme organization can be achieved via a much 30 the simpler and more natural markedness process of 25 changing the word order. 20 15 Active 6. Suggestion for Further Studies

10 Passive 5 There are a good number of languages which do not form 0 agentive passive, as they benefit from overt case markers. Old Persian, Avestan, Arabic (see Jabbari, 2003, Jabbari Phonemes Morphemes Words and Kafipour, 2011) to name but a few. Other studies may be conducted on testing the hypothesis that in languages with overt case markers agentive passive sentences would

not be naturally formed. There are, usually, exception(s) to any rule. There are 10 -http://web.cn.edu/kwheeler/gram_passive_voice.html languages with overt case markers in which agentive 11 - The Persian object marker /ra/ is a post-position and it passives are formed. German is a good example. A special appears in the sentence as a separate word. research on German is strongly suggested.

Jabbari 029

References

Shari’at, J. (1988). Persian Grammar . Asatir Publication, Tehran, Andersen, H. (1989). Markedness Theory: The first 150 years, Iran.p.165 Markedness in Synchrony and Diachrony. I.O.M. Tomic, ed, 11-46, Berlin: Mouton. Shaw, H. (1986). Handbook of English. McGraw-Hill International Editions.p.12 Bahar, M.T., Rashid Yasemi, Gh.R., Gharib, A. , Homa'I, J. & Foruzanfar, B. (1963). Persian Grammar. Iranian Text-book Trask, R.L. (1993) A dictionary of Grammatical Terms in Publications. Linguistics. Routledge.191-319

Brown, G. and Yule, G. (1989). Discourse Analysis . Cambridge Trask, R.L. (2007). Language and Linguistics, The Key Concepts , Textbooks inLinguistics.p.126 Routledge .

Chomsky, N. (1986a). Knowledge of Language: Its Nature, Origin and Vahidian Kamyar, T. (1992). The Past Participle and its Use . New York: Praeger.p.74 Significance in Persian. Iranian Journal of Linguistics, Vol.9, No. 2. 60-72. Iranian University Press. Cook, V.J. and Newson, M. (1997). Chomsky's Universal Grammar , Blackwell Zipf, G.K. (1949). Human behavior and the Principle of Least Effort . Crystal, D. (1992). A Encylopedic Dictionary of Language and Cambridge, Massachusetts: Addisru-Wesely. Languages . Blackwell.p.87-426 http://web.cn.edu/kwheeler/gram_passive_voice.html. Crystal, D. (2002). The Cambridge Encyclopedia of Language . Cambridge University Press.

Halliday, M.A.K. (1985). An Introduction to Functional Grammar . Arnold.p.36

Jabbari, M.J. (2003) Passive Voice in Persian and English. Iranian Journal of Linguistics, Quarterly, Vol. 18, No. 1 , Iran University Press.

Jabbari, M.J. and Kafipour, R. (2011). Long Passive Formation in English and Arabic. European Journal of Social Sciences . Vol. 26 No. 2 (2011), pp.193-200

Lambton, A.K.S. (1966). Persian Grammar . Cambridge University Press.

Lazard, G. (1957). Grammaire Du Persan Contemporain . Paris. Kincksieck.p.161

Lyons, J. (1981) Language, Meaning and Context . London, Fontana.p.227

Mathesius, V. (1942). From comparative word order studies, Casopis pro Moderni Filologii.p. 28

Matthews, P. (2007). Concise Dictionary of Linguistics . Oxford University Press.p.48-442

Najafi, A. (1987). The vase was broken by the kid. Nashr-e Danesh . Iran University Press. P.3

Natel Khanlari, P. (1991). Persian Grammar .Tus Publication, Tehran.p.84

Nida, E.(1982). Establishing Translation Principles and Procedures- The BibleTranslation. 33:4-190.

Richards, J.C. and Schmidt, R. (2002). Dictionary of Language Teaching and Applied Linguistics , Longman, UK.p.63-580