Article

1. In studies of Sūrat al-Baqara as a literary unity, vv. 1–39 have almost unanimously been described as the sura’s “introduction”. The main exception to this trend being synchronic analyses of the Qur’an that utilise ring theory as their organising principle. Thus, for example, Raymond Farrin classifies Q 2:1–20.285– 286 as Ring A/A’ of a concentrically ordered al-Baqara, Q 2:21–29.254–284 as Ring B/B’. See Raymond K. Farrin, “Surat al-Baqara: A Structural Analysis”, Muslim World 100 (2010).
2. Despite this dual emphasis on the structural integrity of the opening verses of al-Baqara and on the function of Q 2:1–39 in contributing to the coherence of the sura as a literary unit, diachronic breakdowns of the sura suggest a number of alternative categorisations of this material. This system was analysed in a lengthy article by Behnam Sadeghi, who corroborated a basic developmental trajectory from shorter verses to longer via three separate groups of multivariate markers: if the Qur’an is rearranged into seven “phases” in accordance with the mean verse length of its textual blocks, Sadeghi’s three groups of multivariate markers were found to exhibit a similarly smooth pattern of variance. The opening verses of al-Baqara Bazargan divides into three sections. He categorises vv. 1–20 as one unit, part of his Block 113; the longer mean verse length of vv. 21–29 then places this subsection into Bazargan’s Block183; the Adam narrative in vv. 30–39, meanwhile, falls into Block 139. Taking this data to its logical extreme, this would suggest that vv. 1–20 form part of the very earliest layer of Sūrat al-Baqara, while vv. 21–29 are contemporary with much of the central, yāayyuhā lladhīna āmanū section of the sura. Vv. 30–39, meanwhile, are similar in mean verse length to only one other textual block, namely, vv. 190–195, which instruct the believers to first “fight” and then “spend in God’s cause” (qātilū fī sabīli llāh … anfiqū fī sabīli llāh …).

More specifically, Sinai observes that the isolated letters and the statement “This is the Scripture” that open Sūrat al-Baqara are typical not of Medinan material but of suras that pre-date the hijra (See Nöldeke, History, 173, and Nicolai Sinai, The Qur’an: A Historical-Critical Introduction (Edinburgh: Edinburgh University Press, 2017), 100). While the marked similarities between Q 2:1–7 and the opening verses of Suras 27 and 31 are strongly indicative of the possibility that “the beginning of surah 2 recycles older Meccan material that was subsequently adapted to serve as the prelude to a Medinan surah”. In his placing a shift to a more typically Medinan register at the precise juncture of vv. 20/21, Sinai is here relying upon Bazargan’s division of the corpus into diachronic blocks; the presence of the phrase “in their hearts is sickness” in Q 2:10 (cf. the Medinan suras Q 5, 8, 9, 22, 24, 33, 47, and the Medinan insertion Q 74:31) is taken as one specific indication that earlier material might here have been reworked, and not inserted verbatim into the sura.The matter of the lexical and phraseological overlaps between Q 2:1–7, 27:1– 6, and 31:1–5 is, however, less straightforwardly indicative of any contemporaneity than it might at first seem. As is apparent in Table 2.1 below, a large number of elements of the middle Meccan Q 27:1–5 are echoed in the late Meccan Q 31:1–5;11 a large number of elements of the late Meccan Q 31:1–6 are then echoed in the Medinan Q 2:1–8; and of all the repeated elements it is only the inna lladhīna opening formula of Q 27:4 and 2:6 that cannot be traced as having feasibly come to al-Baqara via the late Meccan Q 31:1–5. In more specific terms: as documented by Table 2.1, (i) there are 14 words that overlap between Sūrat al-Naml (Q 27) and Sūrat Luqmān (Q 31), plus the morphologically cognate parallels al-ḥakīm/mubīn and li-l-muʾminīn/li-l-muḥsinīn, so arguably 16 parallels in total; (ii) there are 9 words that overlap between Sūrat al-Baqara and Sūrat al-Naml; and (iii) there are 19 words that overlap between Sūrat al-Baqara and Sūrat Luqmān, plus the morphologically cognate parallels li-l-muḥsinīn/li-l-muttaqīn.

Lexical overlaps in the opening verses of Q 27, Q 31, and Q 2

The theory that Qur’anic lexicon does indeed move in diachronically directional steps was recently raised in computerised form by Andrew Bannister, whose Oral-Formulaic Study of the Qur’an posits a clear shift in language use between the Meccan and the Medinan parts of the corpus, as defined by Theodor Nöldeke.14 Bannister tagged his computer database such that a label of either “Meccan” or “Medinan” was attached to any particular three-word sequence that occurred with more than 67% frequency in either Nöldeke’s Meccan or his Medinan Qur’an, and with less than 33% frequency in the other. From the data his search provided, Bannister concluded that the Meccan Qur’an consisted of 69.45% “predominantly Meccan”, 15.62% “predominantly Medinan”, and14.93%“predominantly universal”formulae (the latter beingformulae that show no discernible tendency to occur in one corpus rather than the other). The Medinan Qur’an, meanwhile, was found to be built of 0.53% “predominantly Meccan”, 90.95% “predominantly Medinan”, and 8.72% “predominantly universal” formulae. In other words, 90.95% of the language in the Medinan Qur’an was found to have a strong tendency to be more commonly utilised in the Medinan part of the corpus than in the Meccan, suggesting a more lexically coherent Medinan Qur’an than one might otherwise imagine to be the case, and a vague, but inarguably present, divide between the way the Meccan Qur’an and the Medinan Qur’an utilise language. Sadeghi’s findings, that the mvl of large Qur’anic textual blocks varied in parallel with each block’s use of three separate groups of multi-variate markers, also indicate a diachronic shift in language usage (See Sadeghi,“Chronology”).

A careful analysis of the specific language utilised in Q 2:1–7 in fact yields the following statistics: the discrete phrases that occur in vv. 1–7 of al-Baqara are nine times reflected in middle Meccan suras, 29 times reflected in late Meccan suras, and 46 times reflected in Medinan suras (including al-Baqara itself; see the data provided in Table 2.2 below). This analysis differed from that undertaken by Bannister on a number of fronts. First and foremost, the corpus was divided into all four of Nöldeke’s suggested stylistic phases, investigating the possibility of linguistic development across the Meccan side of the Qur’anic corpus in addition to any difference between the Meccan and the Medinan phases of Muhammad’s prophetic career. In addition, the analysis was carried out on phrases of variable length, and not merely three-word sequences. Finally, due to this being a hand-crafted analysis, it was possible to adjust a number of phrases to encompass near parallels or expansions that the computer would not automatically observe. Thus, for instance, the expression alladhīna yuʾminūna bi-l-ghayb, although in itself a Qur’anic hapax, is arguably formulaically and theologically parallel to such phrases as alladhīna yakhshawna rabbahum bi-l-ghayb (Q 21:49, 35:18, 67:12), man khashiya l-raḥmāna bi-l-ghayb (Q 36:11, 50:33), li-yaʿlama llāhu man yakhāfuhu bi-l-ghayb (Q 5:94) or even liyaʿlama llāhu man yanṣuruhu wa-rusulahu bi-l-ghayb (Q 57:25), which occur in four middle Meccan, one late Meccan, and two Medinan suras.

Q 2:1–7 occur a total of 46/84 times in Medinan suras (55%), 29/84 times in late Meccan suras (35%), 9/84 times in middle Meccan suras (11%), and not at all in any of the early Meccan suras (0%).17 The overall Meccan/Medinan split (38 occurrences vs 46 occurrences, 45% vs 55%) is not as pronounced here as in the conclusions to Bannister’s study, however, and it is crucial to note that Bannister’s hypothesis that there is some sort of lexical rupture between two stylistic sides of a corpus is not evidenced here.18 What appears in its place is the suggestion of a steady drift in lexical usage from middle Meccan through late Meccan to Medinan. Another interesting phenomenon is evident, moreover, in the repetition of opening and closing phrases throughout the first 39 verses of al-Baqara. These would appear to connect the initial verses of the sura to a wider literary fabric that extends as far as the end of the sura’s introduction. The repeated opening and closing phrases in the first 39 verses of al-Baqara have been listed in Table 2.3 below. Other repetitions, however, would appear to serve a punctuatory function. The ulāʾikas of vv. 5, 16, 27, and 39, for instance, mark the borders of the text units Q 2:1–5, 8–16, 26–27, and 1–39. The verse-initial references to the disbelievers in vv. 6 and 39 (first inna lladhīna kafarū and then wa-lladhīna kafarū) occur as thefinal statements of the text groups comprised of vv.1–7 and vv.1–39. The recurrence of the closing formula wa-lahum ʿadhābun ʿaẓīm (v. 7) in v. 10’s wa-lahum ʿadhābun alīmun bi-mā kānū yakdhibūn anticipates the presence of a major caesura at the wa-mina l-nāsi man formula that occurs at vv. 7/8,21 while a minor structural border within the text block Q 2:8–16 is created by the shift to the wa-idhā qīla lahum formula at vv. 10/11. The termination of the text unit vv. 8–20 is marked by the closing phrase inna llāha ʿalā kulli shayʾin qadīr; a very similar khātima occurs at the end of the text unit vv. 21–29, which closes wahuwa bi-kulli shayʾin ʿalīm. The description of paradise as eternal in v. 25 (hum fīhā khālidūn) anticipates a thematic shift to a discussion of God’s parables at vv. 25/26, along with a concurrent pronominal shift from the second person singular/plural to the third person singular/plural. The description of the sojourn in the fire as eternal in v. 39 (hum fīhā khālidūn) marks the end of al-Baqara’s introduction. On the basis of these repetitions, paragraphs and sub-paragraphs (or sections) can be proposed in Q 2:1–39 as follows (Table 2.4).

Despite the presence of, for example, two laʿallakum closers and four verses that terminate wa-hum/fa-hum lā … in Sūrat al-Naml,23 or the many inna llāha … closers of Sūrat Luqmān,24 a similar feature is not in evidence in either Q 27 or Q 31. Repeated openers and closers are not spaced across the openings or the bodies of these suras such as to provide added emphasis to discernible thematic shifts. A similar phenomenon is, however, noticeable in the opening 32 verses of Sūrat Āl ʿImrān. The repeated phrases that unify the sura’s introduction include lā ilāha illā huwa, which occurs at v. 2, and is extended to lā ilāha illā huwa lʿazīzu l-ḥakīm in vv. 6 and 18. The grammatical structure ulū l-albāb of v. 7 is echoed in the li-ulī l-abṣār of v. 13, and in the wa-ulū l-ʿilm of v. 18. The structural unit that is thereby created (Q 3:2–18) is then extended by the repetition in v. 20 of the closer to v.15,wa-llāhu baṣīrun bi-l-ʿibād, a phrase that isfurthermore echoed in v. 30’swa-llāhu raʾūfun bi-l-ʿibād. A coherent introductory unit that runs from v. 1 to v. 32 therefore seems more plausible than a shorter unit running only to the close of the last lā ilāha illā huwa and ulū l- … repeated elements in v.18.Within this, Q 3:1–32 can be broken down into twelve component sections: v. 1, vv. 2–3, 4–5, 6–7, 8–9, 10–13, 14–18, 19–20, 21–25, 26–28, 29–30, and 31–32. Among these sections, there is a strong argument for the overall cohesion of vv. 1–9 (§1; see Table 2.6 below), which feature four references to God’s “sending down” (nazzala/anzala at vv. 3 and 7) and two references to God’s “signs” (āyāt at vv. 4 and 7), and which transforms its description of alladhīna fī qulūbihim zayghun in v. 7 into a plea to lā tuzigh qulūbanā in v. 9.28 There is similarly evidence for the thematic cohesion of vv. 10–20 (§2): vv. 10, 12, 13, and 19 are connected through four references to those who disbelieve (kafarū/yakfur/kāfiratun); the sins (dhunūbihim) of v. 11 are mirrored in the sins (dhunūbanā) of v. 16; and vv. 19 and 20 are united by four references to islām. The arguments for the lexical cohesion of vv. 21–32 (§3) are less robust, perhaps, but the repeated qul statements of vv. 26, 29, 31, and 32, along with the return to v. 25’s discussion of yawmin lā rayba fīhi in v. 30’s allusion to yawma tajidu kullu nafsin mā ʿamilat min khayrin, create a degree of audible coherence.

1. Paragraphing Medinan Suras (Q 48, 49, 57, 58, 60, 61, and 66)
2. The first of the short Medinan suras to be analysed for the purposes of the present study is Sūrat al Fatḥ (Q 48). Its twenty-nine verses contain a high number of lexical echoes and repetitions. There is the suggestion of an encompassing inclusio in the reflection of the initial innā fataḥnā laka fatḥan mubīnā in the closers first to v. 18 (wa-athābahum fatḥan qarībā) and then to v. 27 ( fajaʿala min dūni dhālika fatḥan qarībā). The final words of the sura, meanwhile (wa-ajran ʿaẓīmā, v. 29), are a repetition of the closing words of v. 10 (ajran ʿaẓīmā) and reflect the verse-medial ajran ḥasanan of v. 16. The sura is similarly united by its references to the straight path (ṣirāṭan mustaqīmā; see vv. 2 and 20), to God’s reassurance (al-sakīna; vv. 4 and 18), and to imagery related to hands (see vv. 10, 20, and 24). Sūrat al Fatḥ is, at the same time, regularly punctuated by a series of wa-kāna llāhu … closers (vv. 4, 7, 14, 19, 21, 24, and 26), which further contribute to its stylistic unity.

1. This article initially set out to establish whether the remarkably short mean verse length of Q 2:1–7 (56.29 tlpv), coupled with the noticeable lexical overlaps between these verses and the opening verses of the middle Meccan Q 27 and the late-Meccan Q 31, indicated the presence of older material within a Medinan sura. A careful inspection of the precise language of Q 2:1–7 revealed the lexicon of al-Baqara’s opening verses to be predominantly Medinan. Its discrete phrases, including such statements as alladhīna yuʾminūna bi-mā unzila ilayka wa-mā unzila min qablika and wa-lahum ʿadhābun ʿaẓīm, were found to be reflected 46 times in Medinan suras, 29 times in late-Meccan suras, 9 times in middle Meccan suras, and not at all in the earliest stratum of the corpus. While the verses of Q 2:1–7 utilised vocabulary, verse lengths, and structural paradigms that are evidenced in Meccan suras, they frequently did this in a way that is also typical of Medinan Qur’anic material. The wider tendency of the verses of al-Baqara’s introduction to contain repeated opening or closing formulae, especially at what would appear to be structurally significant junctures of the sura’s introduction, also seemed remarkable. The phrase ulāʾika ʿalā hudan min rabbihim wa-ulāʾika humu lmufliḥūn, for example, which occurs in the opening verses of both Sūrat alBaqara (at Q 2:5) and Sūrat Luqmān (at Q 31:5), is reflected at one further location of Sūrat Luqmān: Q 31:6 terminates with the related clausula, ulāʾika lahum ʿadhābun muhīn. Sūrat al-Baqara (Q 2), in contrast, contains three subsequent ulāʾika statements in its introduction alone: ulāʾika lladhīna shtarawu l-ḍalālata bi-l-hudā in v. 16, ulāʾika humu l-khāsirūn in v. 27, and ulāʾika aṣḥābu l-nāri hum fīhā khālidūn at the very close of the introduction, in v. 39.
2. Similar observations were made regarding, for example, the repeated phrases inna lladhīna kafarū (see vv. 6 and 39, and cf. v. 28), wa-lahum ʿadhābun ʿaẓīm (v. 7; cf. v. 10), in kuntum ṣādiqīn (see vv. 23 and 31), and wa-hum fīhā khālidūn (see vv. 25 and 39). A comparable phenomenon was then found to be apparent in the introduction to Sūrat Āl ʿImrān (Q 3:1–32), with the phrase lā ilāha illā huwa,for instance, occurring in vv. 3, 6, and twice in v. 18, wa-llāhu baṣīrun/raʾūfun bi-l-ʿibād terminating vv. 15, 20, and 30, and the expression inna lladhīna kafarū bi-āyāti llāhi falling at the opening of vv. 4, 10, and 21 (with further echoes in vv. 12 and 19). Indeed, it was possible to posit the entirety of Q 3:1–32 as being united around the dominant Leitwort kāfir/kafarū, culminating in v. 32’s final statement: fainna llāha lā yuḥibbu l-kāfirīn. The body of this article, therefore, was dedicated to an investigation of the use of repetition within the Medinan stratum of the Qur’anic corpus. Even within this small subset of supposedly Medinan suras, moreover, there is stylistic variation. The overlapping inclusios of Q 3:1–32 and the presence of an underlying ring structure to Q 58 are a far cry from the straightforward thematic paragraphing of Q 2:1–39 or of Sūrat al-Ḥujurāt (Q 49), for example. Likewise, some suras were found to exhibit repeated openers and closers (Q 48, for example, or Q 57), while others (Q 61, for example, or Q 66) display repetition only at the beginnings of their thematic units. The bare-bones repetitions of suras that rely heavily on vocative addresses to provide structure (Q 60, for example, and Q 61) may speak to a slightly different rhetorical pattern altogether, one that is more heavily inclusio-based. There is a variety of ways of classifying this material, and a certain flexibility in the erection of the boxes and categories in which we place stylistic features, and the suras that exhibit them, continues to be necessary. Indeed, without in any way wishing to disturb the forward momentum created by recent developments in the field, I would argue that caution remains a desideratum in all attempts to date Qur’anic material to a certain stylistic period. It was demonstrated in the opening pages of this essay that Medinan suras may have utilised Meccan language and Meccan stylistic paradigms for Medinan purposes.60 This should prevent us from drawing misleading conclusions from Sinai’s tentative observation that the opener to Sūrat al-Baqara, a series of isolated letters followed by the statement dhālika l-kitāb (Q 2:1–2), is more typically Meccan than Medinan.