Hallel, Braginsky Collection, B57, Zurich, ff. 149v–150r (adapted). E-Codices
Hallel: How Many Psalms?
The answer depends on the manuscript and tradition we read: Psalms are segmented differently from as early as the Dead Sea Scrolls and the Septuagint and continuing through the medieval Hebrew manuscripts of the Masoretic Text.
Published: October 21, 2024
|
Last Updated: September 7, 2026
I use the term “Hallel” to refer to these Psalms, but they are not the only text so called. Psalms 145–150 have also been referred to at some points in history as the Hallel and Psalm 136 is sometimes called the Great Hallel. In this article, “Hallel” always refers to the Passover Hallel. For more on the Hallel, see Marc Zvi Brettler, “A Woman’s Voice in the Psalter: A New Understanding of Psalm 113,” TheTorah (2019).
Siddur R. Saadia Gaon, ed. Israel Davidson, Simcha Assaf, and Issachar Joel (Jerusalem: Sion, 1963), קנג; and see Rashi’s comment on b. Taanit 28b.
Moses Maimonides, Mishneh Torah: Megillah and Hanukkah, 3:8. Further discussion of the halakha appears in William Yarchin, “Were the Psalms Scrolls at Qumran True Psalters?” Journal of Biblical Literature 134 (2015): 775–789, here 783.
On the division of the psalms in the modern Bible, which he refers to as the textus receptus, see Yarchin, “Were the Psalms Scrolls at Qumran True Psalters?,” 776; and idem, “Is there an Authoritative Shape for the Hebrew Book of Psalms? Profiling the Manuscripts of the Hebrew Psalter,” Revue Biblique 122 (2016), 355–370, here 357–358. The dominance of this segmentation and numeration can be traced to the popularity of the Second Rabbinic Bible, produced by the printing house of Daniel Bomberg in 1525, which was used as the base text of many subsequent editions of the Hebrew Bible.
On scribal traditions for copying the scrolls, see Emanuel Tov, “The Torah Scroll: How the Copying Process Became Sacred,” TheTorah (2023).
In linguistics, segmentation refers to the subdivision of any linguistic string into smaller units of meaning—works into paragraphs, paragraphs into sentences, sentences into words, words into morphemes, morphemes into sounds, and so on.
Though numbering was practiced by Hebrew-language scribes from early in the extant medieval witness (cf. Codex Leningradensis, 11th c. C.E., which was numbered by the original scribe throughout the Psalms but nowhere else), it was not universal practice. For further history and variation in the chapter divisions of early print editions, see Jordan S. Penkower, “The Chapter Divisions in the 1525 Rabbinic Bible,” VT 48 (1998): 350–374. On the history of the verse divisions, see Jordan S. Penkower, “Verse Divisions in the Hebrew Bible,” VT 50 (2000): 379–393.
For more on the MT, see Emanuel Tov, “The Bible and the Masoretic Text,” TheTorah (2017).
See table in Eric J. Harvey, “Sing to the Lord a New(-ish) Song: The Psalms of the Egyptian Hallel Across Two Thousand Years,” (Ph.D. diss. Brandeis University, 2020), 103–105. The data set was originally compiled by William Yarchin, with analysis published in Yarchin, “Is there an Authoritative Shape for the Hebrew Book of Psalms?” and idem, “Were the Psalms Scrolls at Qumran True Psalters?” Some numbers given here will be intentionally vague, since a few of the manuscripts in the set turned out to be duplicates, ancient rather than medieval, or of uncertain identity. Results from cleaned and expanded data will appear in future publications.
From the 13th c. Italian Psalter Parm 1871. See Harvey, “Sing to the Lord a New(-ish) Song,” 124–132.
In the Ashkenazic Tanakh Berlin ms Ham 80, dated 1290, Psalms 115 and 116 are segmented into two psalms each (115:1–11, 115:12–18, 116:1–11, and 116:12–19), and Psalm 118 is segmented into four psalms (vv. 1–4, 5–24, 25, and 26–29). See Harvey, “Sing to the Lord a New(-ish) Song,” 133–137.
Some manuscripts divide the Hallel into six psalms, but differently from the modern Psalter. One example is the six-psalm Hallel of the LXX, below, but the Hebrew manuscript tradition also includes several additional six-psalm segmentations.
From there, representation of individual segmentations drops off precipitously. Another nine segmentations appear five or more times in the manuscripts, while seventeen segmentations appear in two to four manuscripts, and thirty appear in only one manuscript.
The earliest manuscripts date from the common era. Joachim Schaper, “The Septuagint Psalter,” in Oxford Handbook of the Psalms (ed. William P. Brown; Oxford: Oxford University Press, 2014), 173–174.
Psalms 9 and 10 together make a full acrostic, leading many scholars to conclude that they were originally one poem that was split apart at some later point.
In contrast to the absence of psalm numbers in many Hebrew manuscripts, numbering in Greek manuscripts is common. Of 51 Greek Psalters dating from the 3rd to the 15th centuries examined by William Yarchin (personal communication) 42 contain numbers that are original to the manuscript, while seven had numbers added later, and in two cases it is not specified. There does not seem to be a meaningful correlation between time period and original enumeration—the six oldest manuscripts, dating from the late-3rd to 6th centuries, were all numbered by their original copyists.
This segmentation predominates among extant Greek manuscripts and was adopted by many translations into Latin and other European languages.
See discussion and references in Harvey, “Sing to the Lord a New(-ish) Song,” 230–233. This segmentation may have appeared in some early Latin translations, but across the Mediterranean basin it was eclipsed by the LXX segmentation and was carried forward only in the Coptic and Sahidic translations of the Psalms transmitted in Upper Egypt and Ethiopia.
In addition to 4QPsalmsb (4Q84) and 4QPsalmso (4Q96), portions of the Hallel Psalms are also extant on 4QPse (4Q87), 11QPsa (11Q5), 11QPsb (11Q6), and 11QPsd (11Q8). Of these, 4Qpse, 11Qpsa, and 11QPsb differ so dramatically from the Masoretic Psalter that they have been identified as manuscripts of a variant (or even competing) Psalter compiled by the Qumran community. See discussion and references to literature in Harvey, “Sing to the Lord a New(-ish) Song,” 5–9, 175–221
The text is published in Eugene Ulrich, et al., Qumran Cave 4, XI: Psalms to Chronicles (DJD XVI; Oxford: Clarendon, 2000), 23–48. For discussion, see Harvey, “Sing to the Lord a New(-ish) Song,” 177–186.
For, cola, lines, and the structure of Hebrew poetry, see Jason M. H. Gaines, “Poetic Laws,” TheTorah (2017).
Psalm 117 may also have been written with the longer lines, but this cannot be proven since it is entirely missing owing to the destruction of the beginning of column XXXIV.
However, note that three Tannaitic sages describe three different methods for a leader and a congregation to recite Hallel as call and response or in unison, none of which implies a differentiation between Psalm 118:1–4 and 25–26 and the rest of the text (b. Sot 30). See Stefan C. Reif, Problems with Prayers: Studies in the Textual History of Early Rabbinic Liturgy (Berlin: De Gruyter, 2006), 85.
For text, see Ulrich, et al., Qumran Cave 4, XI: Psalms to Chronicles, 139–142. For discussion, see Yarchin, “Were the Psalms Scrolls at Qumran True Psalters?,” 780, and Harvey, “Sing to the Lord a New(-ish) Song,” 189–190.
DJD 16, 140, cf. pl. XVIII. Reconstructions follow Qumran orthographic traditions, with the chapter and verse numbers added for clarity in comparing the text of the manuscript fragment to the MT.
The second fragment of 4QPsalmso further indicates that Psalm 116 may have been segmented as two psalms. Again, the preservation is partial and the reconstruction is not ironclad, but the layout suggests that one psalm ended with Psalm 116:9. Though line 9 of the fragment has substantial room for additional text following verse 9, the scribe started the next verse on line 10 (see DJD 16, 140, cf. pl. XVIII).
See David Willgren, The Formation of the Book of Psalms: Reconsidering the Transmission and Canonization of Psalmody in Light of Material Culture and the Poetics of Anthologies, FAT II 88 (Tübingen: Mohr Siebeck, 2016).
See footnote 26, above.