Pausal Forms
1. Introduction 2. Pausal forms in Standard Arabic A pausal form is the form a word has at the end of a sentence or major phrase or before a Pausal forms are derived from the basic, full pause or stop in the speech flow (waqf ), if that forms by (i) deleting final short vowels (so the is different from the form it takes in the begin- pausal form of kataba ‘he wrote’ is katab); (ii) ning or middle of a phrase. In Classical and deleting a final suffix -n (so the pausal form of Modern Standard Arabic, most words have kitàbun ‘book’ is kitàb); (iii) replacing the suffix different pausal and medial forms. Phonetic -at- with -ah (kitàb-at-u-n ‘writing’ becomes pausal phenomena probably occur in all lan- kitàbah). This Standard Arabic morphological guages, although they may differ from language alternation between full and pausal forms is to language, but morphologically conditioned absent from the modern vernacular dialects and pausal changes are much rarer, and they are was lost in the medieval period, if not earlier; the ones most often referred to when discussing the vernacular form of a word is generally pausal forms in Arabic. The morphological derived from the Standard Arabic pausal form, pausal phenomena of Arabic prose are of a not the full form. single general type: the pausal form is produced The deletion of final short vowels is integrally by subtracting from the base form of the word related to the fact that, in both Old Arabic a final short vowel and the final consonant and the modern vernacular dialects, with few of certain suffixes. The pausal form is thus exceptions, the phonemic opposition of vowel shorter than the medial form, and grammatical and consonant quantity is neutralized in pausal features (case and mood) that in a full form are position. (In some modern dialects this applies marked by suffixes consisting of a short vowel only to unstressed vowels.) Thus, the final are absent from the pausal form. Therefore, the consonants of the words yad ‘hand’ and radd medial or context form of a word may be appro- ‘response’ are identical, when not followed by priately called its full form or basic form, in another word (both may be pronounced with that the pausal form can be deduced from it but a longer or a shorter d). The same is true it cannot be derived unambiguously from the with vowels. In the Damascus vernacular, for pausal form. When mentioning an Arabic word example, “If a final vowel is . . . unaccented, it in isolation, one usually cites the pausal form, varies between long and short depending on for two reasons: a word in isolation is, in effect, the phrasing and intonation. Thus . . . the i in in pausal position; and the Standard Arabic xëdi [‘take:imper:fem:sg’] is unaccented (i.e. pausal forms are, on the whole, more similar xódi), but is sometimes actually long” (Cowell than the full forms are to the way the words are 1964:19), so the difference in the vowels pronounced in the modern vernacular dialects of between xódi ‘take!’ and xëd\-ha ‘take her!’ Arabic. Thus, if one asks what the Holy Book of is that in xëd\-ha the stressed ì must be long, Islam is called, or how to say ‘city’, the answer but in xódi the unstressed final i may be long is given in the pausal form, al-qur±àn or madìna, or short. Something similar was likely the case not the full form al-qur±ànu or madìnatun. in Old Arabic (although stress is not known The most detailed description of the pausal to have been a factor), and the same is true forms in a Western language is by Fleisch (1990: of Modern Standard Arabic, where the final 172–197). Wright (1898:368–373) gives a con- vowels of ±anti ‘you [fem. sg.]’ and bintì ‘my cise but full statement of the facts, Retsö (1994) daughter’ are pronounced identically. presents a lucid synthesis of them, Birkeland A word that ends in a long vowel is unchanged (1940) focuses on the historical development in pausal position, but when a word that of the system, and Roman (1982:493–554) basically ends in a short vowel appears in pausal attempts to reconstruct the phonetics and position, it either loses that vowel, lengthens phonology behind it. the vowel, or adds h (Retsö 1994). Loss of
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 1 88/22/2007/22/2007 8:56:098:56:09 PMPM 2 pausal forms the vowel is the norm for final short vowels form qà∂ì, qà∂, or qà∂i (Carter 1990). Two of that are suffixes or part of suffixes, so the these are problematic: qà∂ is not acknowledged pausal form of al-bayt-u ‘the house [Art-house- by the foremost grammarian, Sìbawayhi, and Nom]’, mu≠allim-ùna ‘teachers [teacher-Nom. probably is not used in oral Modern Standard p]’, bayt-u-ka ‘your house [house-Nom-2ms]’, Arabic; and qà∂i ends in a short vowel, which is daras-a ‘he studied [studied-3ms]’, daras-nà- unexpected in pausal position (where in any case hu ‘we studied it [studied-1p-3ms]’, ≠an-hu it is not phonemically distinct from a long ì). ‘from it’ are al-bayt, mu≠allimùn, baytuk, daras, A special rule applies to the suffix -at-, darasnàh, ≠anh respectively. Lengthening is which marks several different morphosyntactic frequent at the ends of lines in poetry. Final features on nouns and adjectives, most often short vowels that are not part of suffixes are (in feminine gender but also some masculines normative Classical Arabic) followed in pause and plurals. Regardless of function, -at- has by -h (called by the grammarians hà± as-sakt, cf. the pausal form -ah, so for the full forms Fleisch 1990:185–186), so the pausal forms of mu≠allim-at-u-n [teacher-fs-Nom-Abs], xalìf- the jussive yaqi ‘he protects’, the imperative ra at-a-n [caliph-at-Acc-Abs], al-™aràmiyy-at-u ‘see!’, and kayfa ‘how’ are yaqih, rah, kayfah. [Art-thieves-at-Nom], the corresponding pausal The suffix -n, marking the absolute state of forms are mu≠allimah, xalìfah, al-™aràmiyyah. nouns and adjectives or the energetic mood of (This does not apply to the suffix -at which verbs, is also deleted in pause; thus, the pausal marks the 3rd person feminine singular on forms of bayt-u-n ‘a house [house-Nom-Abs]’ is verbs; katab-at ‘she wrote’ is unchanged in bayt. However, for words ending in a-n, the -n pause.) In many dialects, the suffix is -a rather is deleted, but the -a (which may be the marker than -ah, and for this reason the suffix is of accusative case or part of the stem) is not often represented in conventional transcriptions dropped but rather lengthened, so the pausal simply as -a. That the suffix was -ah in Old forms of bayt-a-n ‘a house [house-Acc-Abs]’, Arabic is clear from the facts that some modern fata-n ‘a boy [boy-Abs]’ are baytà, fatà. dialects preserve the h and that in classical Pausal forms are based on the corresponding poetry it rhymes with stems ending in ah, and medial, full forms, and not directly on the not with final à. One might suppose that the abstract underlying form. For example, fatan formation of this pausal -ah from -at- is a two- ‘boy’ is derived from an underlying form /fatay- step process, first deleting the -t specifically in u-n/ [boy-Nom-Abs], via an intermediate stage this suffix and then epenthesizing -h after the fatà-n; if the pausal form were derived directly final short vowel by the general process. This from /fatay-u-n/, deletion of the final -u-n might well have been the historical sequence would yield the incorrect form *fatay (which of events, it this does not account for ™ayàh, is, however, attested in certain ancient dialects, the pausal form of ™ayàtun (™aya-at-u-n) ‘life’, cf. Rabin 1951:116). Rather, the pausal form where deletion of the final t-u-n would leave must be derived from the full form fata-n (or ™ayà, which does not end in a short vowel and an intermediate form fatà-n), yielding fatà. so would not get an epenthetic h. In Modern Similarly, in both the indicative and jussive Standard Arabic, there is an alternative pausal forms of the verb ‘he stands’: yaqùmu and form ™ayàt, a back-formation from suffixed yaqum, the stem vowel is underlyingly long forms like ™ayàt-ì ‘my life’. (indicative /yaqùm-u/, jussive /yaqùm/), but in Arabic orthography does not normally the jussive the /ù/ becomes u by the general indicate the difference between pausal and full rule that shortens vowels in closed syllables. forms. The spelling is based on the pausal The difference of vowel length between the forms rather than full forms, and a word is indicative yaqùmu and the jussive yaqum is spelled identically, whether in medial position maintained in their pausal forms, which are or in pause. To be precise, the basic spelling, respectively yaqùm and yaqum. The vowel in composed of letters of the alphabet, represents the pausal indicative yaqùm does not shorten. A the pausal form, even in medial position, special case concerns words like qà∂i-n [judge- while the optional diacritics that augment the Nom/Gen-Abs], which is derived from /qà∂ì-n/ basic spelling with additional phonological by the same vowel-shortening rule. In pause, information, including short vowels, represent where the -n is deleted, the word may have the the full form, even at the end of a sentence.
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 2 88/22/2007/22/2007 8:56:108:56:10 PMPM pausal forms 3 This practice brought about the creation of two consonant cluster, metathesis (naql) might take orthographic features that exist specifically to place instead, so that bakr-u-n [Bakr-Nom-Abs] represent the full-pausal alternation. One is the would become bakur (Rabin 1951:39). representation of the absolute-state suffix -n, In Modern Standard Arabic, speakers follow which is deleted in pausal forms. Rather than the same three basic rules: deletion of final being written with the letter
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 3 88/22/2007/22/2007 8:56:108:56:10 PMPM 4 pausal forms preceding phrase and determine whether the and others less so; some can apply it ‘on the reciter is to stop, to continue with what follows, fly’, while others need to think before deciding or back up to bridge a break in meaning or on the appropriate form. Using pausal forms syntax” (Nelson 1985:19). It is the syntax that eliminates the need to make case and mood determines the potential for the use of pausal distinctions in many types of words. In the most forms as well as for appropriate actual pauses formal situations, i.e. the reading of a prepared in speech. script by a skilled reader such as a professional For classical Arabic prose we do not know newscaster or some public speakers, pausal where pausal forms occur, because the spelling forms are relatively few and found almost does not differentiate between full and pausal exclusively before actual pauses, i.e. silences forms. For classical poetry, the rhyme makes or phrase-final intonation contours at the ends it clear that words at the end of a line, which of clauses and sentences. At the opposite end always coincides with the end of a major of the spectrum, speaking spontaneously, all syntactic constituent, are to be read in pausal or nearly all words may be in their pausal form, and the meter shows that full forms forms. Thus, full forms connote formality, appear in midline. In the Qur±àn, too, rhyme and pausal forms in medial position connote frequently shows where pausal forms are called informality (Diem 1974:37), and for this reason for. For example, in sura 1, al-Fàti™a, the full forms occasionally occur even before actual words ar-ra™ìmi, al-≠àlamìna, ad-dìni, nasta≠ìnu, pauses. The grammatical function of the word mustaqìma, and a∂-∂àllìna nearly rhyme as ar- may also influence whether a pausal or a full ra™ìm, al-≠àlamìn, ad-dìn, nasta≠ìn, mustaqìm, form is used. Holes (2004:63–68) observes a∂-∂àllìn. According to the rules of ¤ tajwìd, the following tendencies in the retention of which prescribe the phonetics of Qur±ànic full forms in pausal position, in recent news recitation, pausal forms are used at the end of broadcasts: nouns but not adjectives tend to a verse and at certain spots in midverse, which keep the -i-n marking the genitive absolute; and are indicated by special symbols in the standard the accusative absolute -a-n is generally retained editions of the Qur±àn, but reciters have some in adverbs like taqrìb-a-n, may be retained or discretion as to which of the indicated pausal dropped in masculine nouns and adjectives, locations to actually apply. An enlightening but is dropped in feminine nouns after the -at-, discussion of the rules and esthetics is found which becomes -a(h). Extensive transcripts of in Nelson (1985:27–31). Pause is obligatory speech in different registers, showing varying following “semantic and syntactic independence degrees of use of full and pausal forms, are of what precedes or follows. . . . [where] the given by Diem (1974). phrase preceding the obligatory pause is self- contained and takes the form of an epigram 4. From Old Arabic to the or summation and most commonly signals the modern vernacular dialects end of a subject matter” (1985:28), while, if there is “syntactic and semantic dependence Most scholars accept that in Old Arabic full and on what follows, yet it may be a complete pausal forms alternated just as the indigenous phrase . . . [p]ause . . . is permitted here, especially grammarians described, and that over the if the phrase is the first half of a conditional course of time the full forms (and with them sentence, or similarly syntactically dependent the case and mood distinctions, represented on the following phrase” (1985:29). predominantly by short final vowels) were lost In the oral use of Modern Standard Arabic, from ordinary speech, surviving only in the the practice is complex and variable (Meiseles learned use of Classical Arabic, so that the 1977; Schulz 1981; Diem 1974:36–37 and forms that appear in the modern vernaculars passim; Holes 2004:63–68). Since the Standard are derived from the Old Arabic pausal forms. Arabic system of cases and moods, which are Diem (1991) shows that the high degree of marked for most words by short-vowel suffixes, redundancy of the case and mood marking in is not part of any modern vernacular dialect, Classical Arabic, followed by the loss of the Arabs must expend significant effort as part of syntactic-semantic categories of case and mood, their schooling to learn it, and, like all school must have preceded, and been conducive to, subjects, some individuals master it thoroughly the elision of the final short vowels, that is,
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 4 88/22/2007/22/2007 8:56:108:56:10 PMPM pausal forms 5 the generalization of pausal forms to medial The first involves the suffix -at- (pausal position. Just when the full forms dropped out -ah). Nouns with this suffix have two forms of colloquial use is a matter of dispute, with in modern vernacular Arabic, a form ending opinions ranging from a couple of centuries in -t that appears when suffixed or followed before the time of Mu™ammad to a couple immediately by another noun in a ‘construct of centuries after. The issue of timing has phrase’ (¤ ±i∂àfa), as in sayyàrt-i ‘my car’ or profound implications for the understanding of sayyàrit i†-†abìb ‘the doctor’s car’, and the other how natural or artificial the language of early form, without -t, that appears everywhere else: Arabic poetry was, and hence to what extent sayyàra. The structure of sayyàrit i†-†abìb derives that poetry should be relied on as a basis for from an Old Arabic genitive structure like interpreting the Qur±àn. As important as the sayyàrat-u a†-†abìb-i ‘the doctor’s car [car-Nom question of timing is, it does not affect the Art-doctor-Gen]’. When final short vowels were general picture of the history of the language, lost, Old Arabic full forms (like sayyàratu) were for it is clear that the short-vowel case endings replaced by pausal forms (sayyàrah) wherever were features of Proto-Semitic, having cognates possible, i.e. at the end of every phrase, but the in other Semitic languages, and that the pausal t was retained in a construct because the first forms correspond closely to the forms in the element of a construct phrase is not a noun modern dialects; the transition from the former phrase but a single noun. Furthermore, the t type to the latter in Arabic could not have been has survived as a fossil in two types of words: abrupt, so there must have been a period of as a ‘connective t’ in a few vowel-initial noun variation. plurals after numerals, such as xams-t-iyyàm However, Owens (1998; 2006) has challenged ‘five days’, from Old Arabic xams-at-u ±ayyàm- this picture, suggesting that varieties of Arabic i-n, and as † in the numerals from 13 through with full forms, case, and mood, and varieties 19, for instance xams†a≠š(ar) from xams-at-a without them, coexisted for many centuries, and ≠ašar-a. These examples are from Syrian Arabic in fact that such coexistence may have predated (Cowell 1964:170–171), but the forms are the rise of Arabic as a distinct Semitic language. similar in many dialects. If so, they must have remained in close contact Alternations like sayyàra ~ sayyàrit are found and shared in the many innovations of grammar in all modern dialects. The second relic exists and vocabulary that characterize Arabic as a in only a few (Blau 1965:187–202; Fischer whole vis-à-vis other Semitic languages. In this and Jastrow 1980:120–121; Owens 1998:215– view, Classical Arabic represents the variety 217). In these dialects, an indefinite noun may that retained case and mood endings, while the be linked to a following adjective, prepositional modern dialects descend from a form that lacked phrase, or relative clause by a suffix consisting them, and not from Classical Arabic at all. of a short vowel plus n. This is evidently a relic There are, however, good reasons to believe of the Old Arabic absolute state suffix -n, which that pausal and full forms coexisted within is deleted in pausal forms (¤ nunation). Thus, a single variety of Arabic and that such a an Old Arabic structure like nàs-u-n kaμìr-u-n variety is the ancestor of the modern dialects ‘many people [people-Nom-Abs many-Nom- (Hoberman 1995:162–164). The grammarians, Abs]’, or in pause nàs-u-n kaμìr, became nàs- especially Sìbawayhi, took pains to describe the en ∑eμìr [people-Linker many], where the -n language as they observed it in use, and would survives internally but not at the end of the not have invented as pervasive a phenomenon construction. Furthermore, this accounts for as the pausal/full alternation; Classical Arabic the n in such relic forms as Baghdadi šinu poetry can be parsed for meter and rhyme only ‘what’, from (±ayyu) šay±-i-n huwa [(which) if words in midline are in their full forms but thing-Gen-Abs it]. at line-ends are in pausal forms; there is also The survival of the t and n medially, but not evidence from non-Classical spellings in the finally, in these two modern productive alter- early Islamic period and from a transcription of nations (sayyàra ~ sayyàrit and nàs ~ nàs-en) an Arabic text in Greek letters (Hopkins 1984). and in the fossilized relic forms shows that at Another sort of evidence comes from relics some time pausal and full forms coexisted in the of the alternations that survive in the modern same variety of Arabic, in synchronic alternation dialects. These are of two kinds. very much as the grammarians described it.
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 5 88/22/2007/22/2007 8:56:118:56:11 PMPM 6 pausal forms 5. Pausal phenomena in Diem, Werner. 1974. Hochsprache und Dialekt im modern vernacular dialects Arabischen: Untersuchungen zur heutigen arabi- schen Zweisprachigkeit. Wiesbaden: F. Steiner. ——. 1991. “Vom Altarabischen zum Neuarabischen: The ends of phrases in modern Arabic dialects, Ein neuer Ansatz”. Semitic studies in honor of Wolf as in all other languages, are marked by slight Leslau, ed. Alan S. Kaye, 297–308. Wiesbaden: phonetic modifications of the sounds of words. In O. Harrassowitz. Fischer, Wolfdietrich and Otto Jastrow. 1980. Hand- the simplest case, there might be a prolongation buch der arabischen Dialekte. Wiesbaden: O. Har- of some sounds. Thus, in Damascus Arabic, “the rassowitz. end of a phrase is often signaled by drawling Fleisch, Henri. 1990. Traité de philologie arabe. I. out what comes after the accent. . . . With certain Préliminaires, phonétique, morphologie nominal. 2nd ed. Beirut: Dar El-Machreq. kinds of intonation – in questions, for instance – Goldenberg, Gideon (ed.). 1977. Israel Oriental the phrase-end drawl is often exaggerated so studies, VII. Tel Aviv: Tel Aviv University. that a post-tonic short vowel is as long as or Hoberman, Robert D. 1995. “Subtractive morphology longer than a true long vowel in other positions. and morpheme identity in Arabic pausal forms”. Yearbook of morphology, 1995, ed. Geert Booij In the question kìf ™àlak? ‘how are you?’, the and Jaap van Marle, 161–174. Dordrecht: Kluwer last a may actually be longer than the à in the Academic Publishers. preceding syllable” (Cowell 1964:17). In some Holes, Clive. 2004. Modern Arabic: Structures, func- tions, and varieties. Washington, D.C.: Georgetown dialects, the phrase-end effects are greater. In University Press. the dialect of the Negev Bedouin, in pause, Hopkins, Simon. 1984. Studies in the grammar of stressed a is followed by a glottal stop (medial Early Arabic. Oxford: Oxford University Press. mišá, pausal mišá± ‘he went’), and stressed Meiseles, Gustav. 1977. “Restitution of ‘word- endings’ in Modern Literary Arabic”. Goldenberg long vowels may be followed by [h] (medial (1977:173–195). damm\, pausal damm\h ‘my blood’), among Nelson, Kristina. 1985. The art of reciting the Qur±an. other effects (Blanc 1970:119, 122–123). Final Austin: University of Texas Press. (Repr., Cairo: ì and ù become diphthongs in some dialects of American University in Cairo Press, 2001.) Owens, Jonathan. 1998. “Case and Proto-Arabic”. Maltese and Lebanese and Palestinian Arabic Bulletin of the School of Oriental and African (Borg 1977). In parts of Egypt (including Cairo Studies 61.51–73, 215–227. until the end of the 19th century), final a is ——. 2006. A linguistic history of Arabic. Oxford: replaced by e in pause (Blanc 1973–1974). Oxford University Press. Rabin, Chaim. 1951. Ancient West-Arabian. London: Some of these effects are strikingly reminiscent Taylor’s Foreign Press. of those described for Classical Arabic. Retsö, Jan. 1994. “The treatment of final syllables in the classical Arabic metres: The linguistic Bibliographical references background”. Arabic prosody and its applications in Muslim poetry, ed. Lars Johanson and Bo Utas, Birkeland, Harris. 1940. Altarabische Pausalformen. 99–106. Uppsala: Swedish Research Institute in Oslo: Dybwad. Istanbul. Blanc, Haim. 1970. “The Arabic dialect of the Negev Roman, André. 1982. Etude de la phonologie et Bedouins”. Proceedings of the Israel Academy of de la morphologie de la koinè arabe. Marseille: Sciences and Humanities 4:7.112–150. Université de Provence. ——. 1973–1974. “La perte d’une forme pausale dans Schulz, David E. 1981. Diglossia and variation in le parler arabe du Caire”. Mélanges de l’Université Formal Spoken Arabic in Egypt. Ph.D. thesis, Saint-Joseph 48.375–390. University of Wisconsin-Madison. Blau, Joshua. 1965. The emergence and linguistic Wright, W. 1898. A grammar of the Arabic language, background of Judaeo-Arabic: A study of the II. 3rd ed. Cambridge: Cambridge University Press. origins of Middle Arabic. Oxford: Oxford Univer- (Repr., 1971.) sity Press. Borg, Alexander. 1977. “Reflexes of pausal forms in Robert D. Hoberman (Stony Brook University) Maltese rural dialects?” Goldenberg (1977:211– 225). Carter, Michael G. 1990. “Qà∂ì, qà∂i, qà∂: Which is the odd man out?”. Studies in the history of Arabic grammar, II: Proceedings of the 2nd Symposium on the History of Arabic Grammar, Nijmegen, 27 April–1 May 1987, ed. Kees Versteegh and Michael G. Carter, 73–90. Amsterdam and Philadelphia: J. Benjamins. Cowell, Mark W. 1964. A reference grammar of Syrian Arabic (based on the dialect of Damascus). Washington, D.C.: Georgetown University Press.
EEALLALL 3_pausal3_pausal forms_1-0.inddforms_1-0.indd 6 88/22/2007/22/2007 8:56:128:56:12 PMPM