‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 140 Al-ʿUṣūr al-Wusṭā 33 (2025): 139-148 ‘Masʾalatun or Mas’ʾalatun? That Is the Question! The Implications of Traditional Persian Metrics for Classical Arabic Stress AryA TAbibzAdeh University of Tehran Dem lieben Professor Ewald Wagner Abstract Word stress in Classical Arabic (ClAr) is usually reconstructed through cross-dialectal and diachronic analyses of a number of varieties of Arabic and other Semitic languages. However, this paper introduces a new source of data on ClAr stress—namely, the traditional recitation of metra in metrical sequences by Persian prosodists. It shows that the reconstructed pattern of ClAr stress is also observed in such traditional recitation. This stress pattern has nothing to do with Persian phonological or metrical rules, and it should rather be considered an artificial performance practice whose purpose was to imitate the original pronunciation of the metra in Arabic metrics. In view of the oral nature of this practice, I suggest that it has persisted in Persian over a long period (maybe over centuries), and it can therefore be a reliable source of data on ClAr word stress. I also provide some supporting evidence for my hypothesis from medieval Persian texts. On the basis of this new source, I propose that in words with a final heavy syllable preceded by two or more light syllables, the antepenultimate stress pattern was common in many varieties of early Islamic Arabic, including ClAr, and that it was perhaps even more prestigious than the initial stress pattern. This hypothesis may allow us to trace the antepenultimate stress in this syllable structure back to the Proto-Arabic stage. Introduction Classical Arabic (henceforth ClAr) is a collection of linguistic variations described by early Arab grammarians and used in the Quranic reading traditions as well as in pre-Islamic poetry and many early manuscripts in the Islamic period.1 From a linguistic perspective, the defining feature of ClAr is its full preservation of a declensional ending system similar to that of Proto-Arabic, along with nunation.2 Its other phonological, morphological, and syntactic characteristics, however, comprise a diverse amalgamation of linguistic variations drawn from multiple different sources and varieties. Consequently, ClAr cannot be considered 1. M. van Putten, Quranic Arabic: From Its Hijazi Origins to Its Classical Reading Traditions (Leiden: Brill, 2021), 15–47, 227–30; A. Tabibzadeh, review of Quranic Arabic, by M. van Putten, Der Islam 100, no. 2 (2023): 570–73, at 571. 2. See A. Al-Jallad with R. Vollandt, Damascus Psalm Fragment (Chicago: Oriental Institute of the University of Chicago, 2020), 69–77; van Putten, Quranic Arabic, 18. © 2025 Arya Tabibzadeh. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License, which allows users to copy and distribute the material in any medium or format in unadapted form only, for noncommercial purposes only, and only so long as attribution is given to the original authors and source. https://isac.uchicago.edu/research/publications/lamine/lamine2 Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 140 a unitary and homogeneous variety.3 As a phonological feature, ClAr word stress cannot be directly inferred from the abovementioned corpora, as written documents preserve no evidence of word stress and early grammarians say nothing about this feature in the varieties they describe. Hence, modern scholars have reconstructed the ClAr stress pattern through methods such as analysis of stress patterns in various modern dialects and the nondialectal pronunciation of Modern Standard Arabic as well as diachronic comparisons with other Semitic languages.4 However, since these reconstructed patterns are mostly derived from fairly late data, the degree to which they accurately represent older varieties is questionable. Moreover, as already noted, ClAr is an amalgamation of multiple linguistic variations, so it is probable that more than one stress pattern was often accepted in ClAr, even though many patterns were ruled out in later normative forms of ClAr. In this paper, I introduce traditional Persian metrics as a new source of data on ClAr stress. I show that the conventional reconstructed ClAr stress pattern is also observed in the traditional recitation of ajzāʾ (plural of juzʾ, lit. portion = metron) in Persian, when posited in a metrical sequence.5 In the first section I discuss the reconstructed ClAr stress pattern, and in the second section I introduce word stress in Persian. The third section is devoted to an analysis of word stress in the traditional recitation of ajzāʾ in Persian and its comparison with the ClAr and Persian stress patterns. I conclude that the traditional ajzāʾ recitation displays the same stress pattern as ClAr does. This stress pattern cannot be explained by Persian phonological and metrical rules and has nothing to do with Persian stress. In the fourth section of the paper, I argue that the unusual stress pattern observed in the traditional recitation of the metra should be considered an artificial performance practice that has been preserved in Persian for centuries, and that its purpose is to imitate the original pronunciation of the ajzāʾ in Arabic metrics. Finally, the fifth section demonstrates that this traditional recitation practice may solve some problems in the reconstructed stress pattern of ClAr (as well as Proto-Arabic6), especially in the case of words with a final heavy 3. Marijn van Putten has shown this by examining the early grammarians’ works and Quranic recitation traditions (Quranic Arabic, 15–47, 227–30). Of course, centuries-long processes of standardization gave rise to normative varieties of ClAr that endorsed just a subset of ClAr linguistic variations and deemed others nonstandard; see, for example, the homogeneous linguistic systems of textbook ClAr as well as Modern Standard Arabic, which developed in the last stages of the standardization process. See ibid., 227–31. 4. R. Kager, “Stress,” in Encyclopedia of Arabic Language and Linguistics, ed. K. Versteegh, 4:344–53 (Leiden: Brill, 2006–11), 349–50; J. Watson, “Word Stress in Arabic,” in The Blackwell Companion to Phonology, ed. M. van Oostendorp, C. J. Ewen, and K. Rice, 5:1–29 (Oxford: Wiley-Blackwell, 2011), 13–14. 5.  See A. Tabibzadeh, “An Idiosyncratic Pattern of Word-Stress in the Persian Prosody Tradition” [in Persian], Iranian Journal of Comparative Linguistic Research 14, no. 27 (2024): 183–91, 186–88. The Persian prosodic system uses the same word paradigm system as traditional Arabic prosody does to represent metra in metrical sequences, relying on mnemonic words such as مســتفعلن mostafʾelon, مفعــوالُُت mafʾūlāto, مفتعلــن moftaʾelon, and so on. This method of representing metra was introduced by Khalīl b. Aḥmad al-Farāhīdī or al-Farhūdī (ca. 100–170/718–87). See G. Weil, “ʿArūḍ,” in EI2, 1/667–77 (Leiden: Brill, 1968), 669; L. Elwell-Sutton, The Persian Metrics (London: Cambridge University Press, 1976), 10–13; F. Thiesen, A Manual of Classical Persian Prosody (Wiesbaden: Otto Harrassowitz, 1982), 119–21. 6.  See A. Al-Jallad, “Was It Sūrat al-Baqárah? Evidence for Antepenultimate Stress in the Quranic Consonantal Text and Its Relevance for صلوه Type Nouns,” Zeitschrift der Deutschen Morgenländischen Gesellschaft 167, no. https://rjhll.basu.ac.ir/article_5684.html?lang=en 141 • AryA TAbibzAdeh Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 142 syllable preceded by two or more light syllables. In this case, scholars have reconstructed the stress pattern in two different ways, with initial or antepenultimate stress. However, the Persian recitation tradition features only the antepenultimate stress, which may have been the most prestigious stress pattern in ClAr for some time. In the very least, this finding indicates that it was a permissible and common stress ClAr pattern, even if it coexisted with the initial stress pattern. The Reconstructed Pattern of ClAr Word Stress Since ClAr has no native speakers today, its stress pattern has been reconstructed on the basis of various sources, including evidence from modern dialects and diachronic analyses and comparisons of other Arabic varieties and other Semitic languages.7 Modern scholars generally believe that ClAr had phonologically predictable word stress, which was quantity- sensitive and unbounded. They reconstruct the ClAr stress pattern in a relatively consistent way. Here, for example, is Janet Watson’s reconstruction:8 a. Stress a pre-pausal superheavy (CVVC, CVVGG or CVCC) syllable: [ki’tāb] ‘book’, [‘mādd] ‘stretching (MASC SG)’, [ša’ribt] ‘I/you (MASC SG) drank’. b. Otherwise, stress the rightmost non-final heavy (CVV, CVC or CVVG) syllable (up to the antepenult): [da’rasnā] ‘we learnt’, [ṣā’būnun] ‘soap (NOM)’, [‘maktabah] ‘library’, [‘māddun] ‘stretching (NOM)’, [‘maktabatun] ‘library’ (non-pause) (or [mak’tabatun]). c. Otherwise, stress the leftmost CV syllable (or antepenult): [‘kataba] ‘he wrote’, [‘katabatuhū] ‘his scribes’ (or [kata’batuhū]). There is debate among scholars about the reconstruction of the syllabic patterns b and c (i.e., when a final heavy syllable is preceded by two or more light syllables). Some scholars claim that such words feature an initial stress,9 whereas others argue for an antepenultimate stress.10 Watson, as seen above, accepts both possibilities. The disagreement may reflect the influence of modern dialects’ stress patterns. If this is true, only one of the two patterns should be considered the original ClAr pattern; the other pattern would be derived from a non-ClAr source. However, it is also possible that ClAr already featured an optional stress pattern alongside the primary one in medieval times, as a result of the influence of other, non-ClAr varieties of Arabic; again, it should be remembered that ClAr was a heterogeneous collection of linguistic characteristics originating in multiple early Arabic varieties,11 and it would thus not be surprising to find traces of other varieties in the stress patterns of 1 (2017): 85–89, at 86; M. van Putten, “The Development of the Triphthongs in Quranic and Classical Arabic,” Arabian Epigraphic Notes 3 (2017): 47–74, esp. 64–71. 7. Kager, “Stress,” 349; Watson, “Word Stress in Arabic,” 13–14. 8. Watson, “Word Stress in Arabic,” 13–14. 9. Kager, “Stress,” 349. 10. See, for example, K. Versteegh, The Arabic Language, 2nd ed. (Edinburgh: Edinburgh University Press, 2014), 88. For other sources, see Kager, “Stress,” 350. 11. Van Putten, Quranic Arabic, 15–46. https://scholarlypublications.universiteitleiden.nl/handle/1887/47177 Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 142 ClAr. For a more precise reconstruction, then, premodern evidence would be helpful. The following sections show that the stress patterns observed in the traditional recitation of metra in Persian may provide us with such premodern evidence. Word Stress in Persian Modern Iranian Persian word stress is phonologically predictable. It is quantity-insensitive and bounded, and it is applied on the final syllable, as in de’rakht ‘tree’, khā’ne ‘house’, and bīmāres’tān ‘hospital’.12 There is no direct evidence of word stress in Early New Persian (i.e., the Persian spoken between the Islamic conquests of Persia and the premodern period), but the word stress patterns of earlier stages of the language—Middle Persian and Old Persian— as well as Modern Persian indicate that Early New Persian probably featured a quantity- insensitive and bounded final stress just like Modern New Persian does. It is commonly accepted that Middle Persian word stress was quantity-sensitive, bounded, and final.13 Many Middle Persian words with heavy final syllables became shortened in New Persian, but they have retained their final stress in Modern New Persian; for example, nā’mag ‘letter, book’ (Middle Persian) > nā’ma (Early New Persian) > nā’me (Modern New Persian). This indicates that Early New Persian likely had a quantity-insensitive and bounded final stress as well. The Traditional Recitation of the Metra in Persian In the previous section, I showed Modern Iranian Persian has final word stress and that this is likely true of Early New Persian, too. There is, however, a particular context in Persian that exhibits idiosyncratic nonfinal stress patterns: the metra formulae as traditionally recited by Persian prosodists. Persian prosody traditionally uses the same word formulae as traditional Arabic prosody does to represent the metra, including مســتفعلن mostafʾelon,14 ,mafʾūlāto, along with a few nonce words with similar shapes مفعــوالُُت fāʾelāton, and فاعالتــن e.g. فاعلییاتــن fāʾelīyāton.15 The unusual stress patterns of the metron names in traditional 12. M. Eslami, Phonology: Analyzing the Intonation System of Persian [in Persian] (Tehran: Samt, 2005), 79–84; M. Bijankhan: “Phonology,” in The Oxford Handbook of Persian Linguistics, ed. A. Sedighi and P. Shabani- Jadidi, 111–41 (Oxford: Oxford University Press, 2018), 133. 13. M. Maggi and P. Orsatti, “From Old to New Persian,” in Sedighi and Shabani-Jadidi, Oxford Handbook of Persian Linguistics, 7–51, at 20. The most important evidence for this reconstruction is the apocope of the Old Persian final declensional endings in Middle Persian. It is generally believed that these endings were omitted because they were preceded by the stressed word-final syllable. 14. Note that the Arabic ʿayn is realized as a glottal stop in Persian. The vowels in the metron names are transcribed in accordance with their form in modern Iranian Persian. There is no quantity distinction in the modern Iranian Persian vowel system, but in Persian metrics, the vowels ā, ū, and ī are considered long and a, e, and o are considered short (here, ā and a stand for the low back and low front vowels, respectively). 15.  Modern scholars, most notably A. Nadjafi, have invented other word formulae, including مســتفاعلن mostafāʾelon and مفاعیــلنت mafāʾīlaton; see A. Ghahramani Moghbel, “Pentasyllabic Feet in Arabic and Persian Prosody and Their Function in Persian Prosody” [in Persian], Persian Language and Iranian Dialects 5, no. 1 (2020): 45–65, at 52; O. Tabibzadeh, Persian Quantitative Meter: An Analysis Based on Rhythmical Scansion and Prosodic Phonology [in Persian] (Tehran: Ketāb-e Bahār and Institute for Humanities and Cultural Studies, 2023), 64–71; idem, “Correspondence Rules and Classification of Persian Quantitative Poetic Meter,” in Springer https://zaban.guilan.ac.ir/article_3856.html?lang=en https://zaban.guilan.ac.ir/article_3856.html?lang=en 143 • AryA TAbibzAdeh Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 144 recitation cannot be explained by Persian phonological rules. Most notably the name of a metron in Persian takes the normal final stress when it is used as a single word in a sentence, as in rokn-e mostafʾe’lon yek rokn-e sobā’ī-st16 ‘the metron mostafʾelon is a septuple one [i.e., it consists of seven letters]’. However, when recited within a metrical sequence the metron name has a different stress. For example, a metrical sequence consisting of four mostafʾelon metra would be traditionally recited as mos’tafʾelon mos’tafʾelon mos’tafʾelon mos’tafʾelon (i.e., with an antepenultimate stress in each metron). It is interesting that nearly all metra take a nonfinal stress in traditional recitation. To test this observation, I asked Iranian specialists in Persian linguistics, Persian metrics, and classical Persian literature to recite various metrical sequences written on a piece of paper.17 As expected, all the specialists employed the same stress patterns— mostly nonfinal—in their recitations, but they also consistently applied a final stress to the metra names when using them as single words in sentences as opposed to reciting them in metrical sequences. The metrical sequences recited by the specialists along with their stress patterns are as follows:18 1. mafā’ʾīlon mafā’ʾīlon mafā’ʾīlon mafā’ʾīlon 2. mafā’ʾīlon mafā’ʾīlon fa’ʾūlon 3. fāʾe’lāton fāʾe’lāton fāʾe’lāton ‘fāʾelon 4. fāʾe’lāton fāʾe’lāton ‘fāʾelon 5. ma’fāʾelon faʾe’lāton ma’fāʾelon ‘faʾelon 6. ma’fāʾelon faʾe’lāton ma’fāʾelon faʾe’lāt 7. maf’ʾūlo fāʾe’lāto mafā’ʾīlo ‘fāʾelon 8. maf’ʾūlo fāʾe’lāton maf’ʾūlo fāʾe’lāton 9. maf’ʾūlo mafā’ʾīlon maf’ʾūlo mafā’ʾīlon 10. mof’taʾelon mof’taʾelon mof’taʾelon mof’taʾelon 11. mof’taʾelon ma’fāʾelon mof’taʾelon ma’fāʾelon 12. mof’taʾelon fāʾe’lāto mof’taʾelon ‘faʾ 13. mof’taʾelon mof’taʾelon ‘fāʾelon 14. faʾe’lāton faʾe’lāton faʾe’lāton ‘faʾelon 15. maf’ʾūlo mafā’ʾīlo mafā’ʾīlo ‘faʾal Handbook on Phonetics and Phonology of Iranian Languages, ed. R. Falahati (Berlin: Springer, in press). .رکن مستفعلن یک رکن سباعی است  .16 17. My deep thanks to Professors Mohammad Afshin-Vafai, Milad Azimi, Aliasghar Ghahramani-Moghbel, Behrooz Mahmoudi-Bakhtiyari, Mostafa Mousavi-Rad, Mehdi Olyaei-Moghadam, Ahmad-Reza Qaemmaqami, Mohammad Rasekh-Mahand, Aladdin Tabatabaei, and Omid Tabibzadeh for participating in this exercise. 18. In Persian metrics, each line of a piece of poetry consists of two half-lines that are of equal metrical value (in contrast to Arabic metrics, where the half-lines of a line can be unequal). Therefore, I represent each sequence in this list with the length of a half-line rather than a complete line. It should be noted that with the exception of meters 22 and 23, the meters named in the list are the most popular meters in Persian poetry; see Elwell-Sutton, Persian Metrics, 87; A. Nadjafi, The Persian Metrics: A Textbook [in Persian], ed. O. Tabibzadeh (Tehran: Niloufar, 2016), 127–47. The exceptional meters (22 and 23) are included in the list because they are relevant to our discussion of their special syllabification pattern below. Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 144 16. maf’ʾūlo mafā’ʾīlo mafā’ʾīlo fa’ʾūlon 17. maf’ʾūlo ma’fāʾelon mafā’ʾīlo ‘faʾal 18. mota’fāʾelon mota’fāʾelon mota’fāʾelon mota’fāʾelon 19. mofā’ʾalaton mofā’ʾalaton mofā’ʾalaton mofā’ʾalaton 20. fa’ʾūlon fa’ʾūlon fa’ʾūlon fa’ʾūlon 21. fa’ʾūlon fa’ʾūlon fa’ʾūlon ‘faʾal 22. fa’ʾelaton fa’ʾelaton fa’ʾelaton fa’ʾelaton 23. mota’faʾelon mota’faʾelon mota’faʾelon mota’faʾelon An examination of the stress patterns in this list reveals that the position of the stress in traditional recitation of the metra does not follow Persian phonological or metrical rules.19 Instead, it depends on the quantity and order of the syllables; in other words, it behaves in a quantity-sensitive manner, like it does in the ClAr word stress system. Indeed, the stress patterns in traditional recitation and ClAr are identical (compare the Persian metra fa’ʾūlon, mota’fāʾelon, maf’ʾūlo, faʾe’lāton, etc., with the ClAr fa’ʿūlun, muta’fāʿilun, maf’ʿūlu, faʿi’lātun, etc.)—with three exceptions. As already noted, scholars have proposed reconstructing ClAr words with the syllable structures muftaʿilun, faʿilatun, and mutafaʿilun with either an initial or an antepenultimate stress. But in the Persian prosodists’ recitation, these patterns consistently feature only an antepenultimate stress, as shown in the list above (patterns 10–13, 22, and 23). The complete correspondence in stress patterns between ClAr and the traditional recitation of the metra in Persian beyond these exceptions cannot be due to mere chance, and it leaves no doubt that the two systems share the same word stress patterns. When and How Did the Traditional Recitation Crystallize? It is noteworthy that Persian prosodists and specialists use the ClAr stress pattern in their recitation of the metra apparently unconsciously.20 The fact that they do so was first noticed by Omid Tabibzadeh, who pointed out the unusual stress pattern without further elaboration.21 But how did Persian prosody come to borrow this stress pattern from ClAr? To answer this question, it is important to keep in mind that Iranian prosodists learn the traditional recitation of the metra not from any written source but by listening to their teachers and memorizing the stress patterns they apply to the metra. The teachers, in turn, received this recitation tradition by hearing it from their own teachers. Therefore, we are dealing here with a long-standing, purely oral tradition. This tradition reflects an artificial performance practice whose original goal was to imitate how the metron names were pronounced by early Arabic-speaking prosodists. But how far back can we trace this imitative practice? It is not exactly known when the traditional Persian prosody theory was 19. See A. Tabibzadeh, “Idiosyncratic Pattern,” 186, 189–90. 20. Interestingly, even those who are not familiar with the Arabic language use the ClAr stress pattern in their recitation of the metra. 21. O. Tabibzadeh, “The Comparison of Word Stress in Persian and Arabic” [in Persian], Language and Linguistics (in press). 145 • AryA TAbibzAdeh Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 146 established, but the evidence from early reports as well as early Persian poetry22 suggests that it existed as early as the eleventh century CE distinct from Arabic prosodic theory. Thus, although complete certainty is impossible, it is likely that the practice of reciting the metra with ClAr word stress patterns was a feature of Persian prosody from its very beginning in the eleventh century. Consequently, we can use the stress pattern observed in this traditional practice today to produce a more accurate reconstruction of ClAr word stress, as I propose to do in the next section. It should be noted that with the exception of a few varieties of Persian used in southern Iran near Arabic-speaking populations, Persian varieties had little if any contact with non-ClAr varieties of spoken Arabic after the fall of the Abbasids.23 Therefore, the stress patterns of Persian prosodists’ recitation may be an even more reliable source for reconstructing ClAr word stress than native Arabic speakers’ nondialectal pronunciations of Standard Arabic are. ‘Masʾalatun or Mas’ʾalatun? As noted earlier, words containing a nonstressed heavy final syllable preceded by two or more light syllables constitute a point of debate in the reconstruction of ClAr word stress. Whereas modern scholars have proposed either an initial or an antepenultimate stress for such words, the Persian recitation of the metra offers only one option for this pattern, an antepenultimate stress, as observed in the metrical sequences 10–13, 22, and 23 above (i.e., the sequences containing the metra mof’taʾelon, fa’ʾelaton, and mota’faʾelon). From this evidence, we can infer that ClAr did not originally have an initial stress in the syllable pattern under discussion, and its use by native Arabic speakers in their nondialectal pronunciation of Standard Arabic reflects the influence of later non-ClAr dialects.24 However, it must again be remembered that ClAr as defined in the introduction to this paper is but an amalgamation of features from multiple Arabic vernaculars. In other words, dialectal influence is an intrinsic part of ClAr, and we cannot assume that the initial stress pattern was never used in these syllable structures in ClAr varieties. In my view, then, the most likely scenario is that ClAr accommodated (at least) two possible stress patterns in such cases, initial and antepenultimate, but that Persian prosodists chose to endorse only the antepenultimate stress for some reason; perhaps it was more prestigious or more common than the initial stress.25 It is noteworthy that Quranic Arabic (i.e., the language of 22. For example, Nāser Khosrow in the eleventh century used the same termini and scansions as the later Persian prosodists did in his poems; see Nāser Khosrow, The Dīvān of Nāser Khosrow [in Persian], ed. M. Minovi and M. Mohaghegh (Tehran: Tehran University Press, 1978), 1:487. 23. See A. Tabibzadeh, “The Arabic of the Islamic Conquests vs. Classical Arabic in Early Persian Texts: Arabic Loanwords in the Early Persian Texts with Non-Classical Arabic Origin” [in Persian], Iranian Journal of Comparative Linguistic Research 11, no. 22 (2022): 95–160, at 98–99; idem, “Perso-Arabica Arabic Loanwords in Early Persian Texts: Different Origins, Different Norms,” Arabica (in press). 24. For examples of dialectal influence on the pronunciation of ClAr, see T. Mitchell, “Prominence and Syllabication in Arabic,” Bulletin of the School of Oriental and African Studies 23, no. 2 (1960): 369–89; B. Hayes, Metrical Stress Theory (Chicago: University of Chicago Press, 1995), 130–32. 25. This conclusion accords with Watson’s reconstruction of ClAr stress, which considers both initial and antepenultimate stress patterns possible (Watson, “Word Stress in Arabic,” 13–14). https://rjhll.basu.ac.ir/article_4363.html?lang=en https://rjhll.basu.ac.ir/article_4363.html?lang=en Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 146 the Quranic consonantal text) also has an antepenultimate stress in words of the faʿilatun and mutafaʿilun types, and the same may be true of muftaʿilun words.26 The most important implication of the traditional Persian tradition of reciting the metra, in my opinion, is that the antepenultimate stress was common in words of the types discussed here in varieties of early Islamic Arabic, and this evidence may help us reconstruct this stress pattern in the stage of Proto-Arabic as well.27 Conclusion In this paper, I have sought to show that Persian prosodists traditionally apply the word stress patterns characteristic of ClAr when reciting metron names in a metrical sequence, even though they pronounce the names with the normal Persian stress on the final syllable when using them in regular sentences. I have argued that the traditional recitation reflects an artificial practice aimed at imitating the original pronunciation of the metra in Arabic. This practice is part of a purely oral tradition of Persian prosody, which seems to have persisted over many centuries. Accordingly, the tradition may constitute an appropriate source for the reconstruction of ClAr word stress patterns. Modern scholars have posited two probable stress patterns in their reconstruction of ClAr words in which two or more light syllables precede a heavy final one. However, the stress pattern observed in the traditional recitation of the metra in Persian appears to indicate that the antepenultimate stress was common in such cases in ClAr and other early Islamic Arabic varieties, and it was perhaps even more prestigious than the initial stress pattern was. 26.  See Al-Jallad, “Was It Sūrat al-Baqárah?,” 86; van Putten, “Development of the Triphthongs,” 64–71. Although my hypothesis may seem speculative, it may be supported by some evidence from medieval Persian texts suggesting that ClAr word stress in Arabic loanwords seems to be correlated with vowel lengthening in Early New Persian, as in مالطفــه mulātafa ‘short letter’ from the ClAr ملّطّفــة muláṭṭafa, جماهــز mujāhiz ‘equipping’ from ClAr جمهــز mujahhiz, etc. See A. Sadeghi, “Gemination in Persian” [in Persian], Lexicography 5–6 (2013): 3–43, at 36; A. Tabibzadeh, “Perso-Arabica.” A notable piece of evidence is the variable spelling of the loanword in an eleventh- or twelfth-century manuscript of a Persian translation of the َمَســاله and َمَســله question’ as‘ مســأله Quran; see A. Rajayi Bokharayi, A Bridge between the Syllabic and Quantitative Persian Poetry in the first Islamic Centuries [in Persian] (Tehran: Iranian Cultural Foundation, 1974), 110, 132, 138, 153, 159, 169; see also A. Tabibzadeh, “Arabic of the Islamic Conquests,” 120; idem, “Perso-Arabica.” Since the scribe of the text placed the diacritical marks in their expected places (i.e., either above or under the letter), these variable spellings should logically indicate the pronunciations masala and masāla, respectively. If the vowel lengthening in masāla also reflects the practice of adapting ClAr word stress to Persian vowel lengthening, the spellings show the scribe struggling to render the ClAr pronunciation with an antepenultimate stress. 27. Ahmad Al-Jallad (Damascus Psalm Fragment, 63–67) has argued that until the late Umayyad period, some Arabic varieties of Hijazi origin were used in Muslim societies as high registers, and these Hijazi varieties were probably descendants of Quranic Arabic. Phillip Stokes grouped these varieties along with other prestigious medieval non-ClAr varieties into the category “para-Classical Arabic”; see P. Stokes, “In the Middle of What? A Fresh Analysis of the Language Attested in the Judaeo-Arabic Commentary on Pirqe ʾĀvōṯ (the Sayings of the Fathers), Middle Arabic and Implications for the Study of Arabic Linguistic History,” Journal of Semitic Studies 66, no. 2 (2021): 379–411, at 405; see also idem, “Key to the Kingdom: Variations as a Key to Understanding the Arabic Gospel Manuscripts,” Al-ʿUṣūr al-Wusṭā 31 (2023): 1–44, at 38–41. One wonders whether the antepenultimate stress pattern in ClAr might have been a para-ClAr influence on ClAr that was gradually replaced by the initial stress pattern in some later normative ClAr varieties (I owe this point to an anonymous reviewer for the journal). On para-ClAr borrowings in Persian, see A. Tabibzadeh, “Perso-Arabica.” https://journals.library.columbia.edu/index.php/alusur/article/view/uw31stokes https://journals.library.columbia.edu/index.php/alusur/article/view/uw31stokes 147 • AryA TAbibzAdeh Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 148 Acknowledgments I would like to thank the two anonymous reviewers for the journal who offered many helpful comments that improved the quality of this paper. I thank also my dear professors and friends Marijn van Putten, Ahmad Al-Jallad, and Hamza Khwaja, who read the first draft of the paper and provided me with insightful feedback. Of course, I am solely responsible for any errors that remain in the paper. Bibliography Al-Jallad, A. “Was It Sūrat al-Baqárah? Evidence for Antepenultimate Stress in the Quranic Consonantal Text and its Relevance for صلــوه Type Nouns.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 167, no. 1 (2017): 81–90. – – – with R. Vollandt. Damascus Psalm Fragment. Chicago: Oriental Institute of the University of Chicago, 2020. Bijankhan, M. “Phonology.” In The Oxford Handbook of Persian Linguistics, edited by A. Sedighi and P. Shabani-Jadidi, 111–41. Oxford: Oxford University Press, 2018. Elwell-Sutton, L. The Persian Metrics. London: Cambridge University Press, 1976. Eslami, M. Phonology: Analyzing the Intonation System of Persian [in Persian]. Tehran: Samt, 2005. Ghahramani-Moghbel, A. “Pentasyllabic Feet in Arabic and Persian Prosody and Their Function in Persian Prosody” [in Persian]. Persian Language and Iranian Dialects 5, no. 1 (2020): 45–65. Hayes, B. Metrical Stress Theory. Chicago: University of Chicago Press, 1995. Kager, R. “Stress.” In Encyclopedia of Arabic Language and Linguistics, edited by K. Versteegh, 4:344–53. Leiden: Brill, 2006–11. Maggi, M., and P. Orsatti. “From Old to New Persian.” In The Oxford Handbook of Persian Linguistics, edited by A. Sedighi and P. Shabani-Jadidi, 7–51. Oxford: Oxford University Press, 2018. Mitchell, T. “Prominence and Syllabication in Arabic.” Bulletin of the School of Oriental and African Studies 23, no. 2 (1960): 369–89. Nadjafi, A. The Persian Metrics: A Textbook [in Persian]. Edited by O. Tabibzadeh. Tehran: Niloufar, 2016. Nāser Khosrow. The Dīvān of Nāser Khosrow [in Persian]. Edited by M. Minovi and M. Mohaghegh. Tehran: Tehran University Press, 1978. Rajayi Bokharayi, A. A Bridge between the Syllabic and Quantitative Persian Poetry in the first Islamic Centuries [in Persian]. Tehran: Iranian Cultural Foundation, 1974. https://isac.uchicago.edu/research/publications/lamine/lamine2 https://zaban.guilan.ac.ir/article_3856.html?lang=en https://zaban.guilan.ac.ir/article_3856.html?lang=en Al-ʿUṣūr al-Wusṭā 33 (2025) ‘Masʾalatun or Mas’ʾalatun? That Is the Question! • 148 Sadeghi, A. “Gemination in Persian” [in Persian]. Lexicography 5–6 (2013): 3–43. Stokes, P. “In the Middle of What? A Fresh Analysis of the Language Attested in the Judaeo- Arabic Commentary on Pirqe ʾĀvōṯ (the Sayings of the Fathers), Middle Arabic and Implications for the Study of Arabic Linguistic History.” Journal of Semitic Studies 66, no. 2 (2021): 379–411. – – –. “Key to the Kingdom: Variations as a Key to Understanding the Arabic Gospel Manuscripts.” Al-ʿUṣūr al-Wusṭā 31 (2023): 1–44. Tabibzadeh, A. “The Arabic of the Islamic Conquests vs. Classical Arabic in Early Persian Texts: Arabic Loanwords in the Early Persian texts with Non-Classical Arabic Origin” [in Persian]. Iranian Journal of Comparative Linguistic Research 11, no. 22 (2022): 95–160. – – –. “An Idiosyncratic Pattern of Word-Stress in the Persian Prosody Tradition” [in Persian]. Iranian Journal of Comparative Linguistic Research 14, no. 27 (2024): 183–91. – – –. “Perso-Arabica Arabic Loanwords in Early Persian Texts: Different Origins, Different Norms.” Arabica, in press. – – –. Review of Quranic Arabic, by M. van Putten. Der Islam 100, no. 2 (2023): 570–73. Tabibzadeh, O. “The Comparison of Word Stress in Persian and Arabic” [in Persian]. Language and Linguistics, in press. – – –. “Correspondence Rules and Classification of Persian Quantitative Poetic Meter.” In Springer Handbook on Phonetics and Phonology of Iranian Languages, edited by R. Falahati. Berlin: Springer, in press. – – –. Persian Quantitative Meter: An Analysis Based on Rhythmical Scansion and Prosodic Phonology [in Persian]. Tehran: Ketāb-e Bahār and Institute for Humanities and Cultural Studies, 2023. Thiesen, F. A Manual of Classical Persian Prosody. Wiesbaden: Otto Harrassowitz, 1982. van Putten, M. “The Development of the Triphthongs in Quranic and Classical Arabic.” Arabian Epigraphic Notes 3 (2017): 47–74. – – –. Quranic Arabic: From Its Hijazi Origins to Its Classical Reading Traditions. Leiden: Brill, 2022. Versteegh, K. The Arabic Language. 2nd ed. Edinburgh: Edinburgh University Press, 2014. Watson, J. “Word Stress in Arabic.” In The Blackwell Companion to Phonology, edited by M. van Oostendorp, C. J. Ewen, and K. Rice, 5:1–29. Oxford: Wiley-Blackwell, 2011. Weil, G. “ʿArūḍ.” In Encyclopedia of Islam, 2nd ed., 1:667–77. https://journals.library.columbia.edu/index.php/alusur/article/view/uw31stokes https://journals.library.columbia.edu/index.php/alusur/article/view/uw31stokes https://rjhll.basu.ac.ir/article_4363.html?lang=en https://rjhll.basu.ac.ir/article_4363.html?lang=en https://rjhll.basu.ac.ir/article_5684.html?lang=en https://scholarlypublications.universiteitleiden.nl/handle/1887/47177