10. Historical layers: Biblical, Mishnaic, Medieval and Modern Hebrew

How Modern Hebrew differs from the Hebrew of the Bible and of the Mishna, and what it took from each: the verb system, relative clauses, possession, pronouns, negation, phonology, word order, and the influence of the European languages of the revivers. It also sketches the stages in between — Medieval Hebrew and the Hebrew of the Haskala — and says what the post-Rabbinic UD treebank (PostRab) shows about where a Modern Hebrew parser breaks on older texts. For Biblical Hebrew in depth see the Ancient Hebrew manual (Ancient Hebrew (hbo) — a reference manual for Mova, if present in this checkout).

← README · ← 09 Pragmatics and politeness · 11 UD analysis →

Abbreviations: BH Biblical Hebrew (Classical, roughly 10th–2nd c. BCE in its texts), LBH Late Biblical Hebrew (Chronicles, Ezra-Nehemiah, Esther, Daniel), QH the Hebrew of the Dead Sea Scrolls, MH Mishnaic (Rabbinic) Hebrew (Tannaitic, c. 1st–3rd c. CE, spoken; Amoraic, later, literary), MedH Medieval Hebrew, ModH Modern Hebrew. Biblical forms are given in a simplified Tiberian transliteration in italics with macrons only where the vowel quality matters; Modern forms in the manual's transliteration (02).

Contents

• 10.1 The stages
• 10.2 What Modern Hebrew is built from
• 10.3 The verb system
• 10.4 Relative clauses: אשר and ש
• 10.5 Possession: construct, של and suffixes
• 10.6 Pronouns and demonstratives
• 10.7 Negation
• 10.8 Phonology: from Tiberian to Israeli
• 10.9 Word order
• 10.10 The lexicon across the layers
• 10.11 European influence on Modern Hebrew
• 10.12 Medieval Hebrew and the Haskala
• 10.13 Older Hebrew in the treebanks: PostRab and PTNK
• 10.14 One sentence through the ages
• References

10.1 The stages

The periodisation is standard (Kutscher 1982; Sáenz-Badillos 1993 — not held); the date by which Hebrew stopped being spoken in daily life is debated (Kutscher placed it around 200 CE; others argue for later pockets). Whether LBH and its relation to Classical BH reflects date or style is also debated (Rezetko and Young 2014, local hbobook-2014-rezetko-historical-linguistics-biblical-hebrew; Hornkohl 2024, local hbobook-obp0433-hornkohl-diachronic-diversity-classical-biblical-hebrew).

10.2 What Modern Hebrew is built from

A rough map of what each layer contributed (Kutscher 1982; Sáenz-Badillos 1993; Glinert 1989 — not held; details below):

The revival's policy is often summarised as "Mishnaic grammar for speech, Biblical vocabulary and style for elevation" — a simplification, but it captures why Modern Hebrew uses MH structures (participle as present, she-, shel) with a lexicon drawn from every layer.

10.3 The verb system

Biblical Hebrew

BH has two finite conjugations — the suffix conjugation (qatal, "perfect") and the prefix conjugation (yiqtol, "imperfect") — and two "consecutive" forms with a prefixed w-: wayyiqtol (narrative past: "and he did") and weqatal (future, habitual, procedural sequence: "and he will do"). The forms are traditionally described as aspectual (perfective vs imperfective) rather than tense-based, and their analysis remains one of the most debated questions in Semitic linguistics (Driver 1892, local pd driver-tenses-1892; GKC 1910 §§ 106–112, local pd gesenius-kautzsch-cowley-1910; Cook 2012 — not held; Isaksson 2021, local hbobook-obp0414-isaksson-verb-classical-hebrew; Andrason 2011, local hbo-2011-andrason-biblical-hebrew-wayyiqtol-dynamic-definition).

ויאמר אלהים יהי אור ויהי אור          (Gen 1:3)
wayyōmer ʾĕlōhîm yǝhî ʾôr wayyǝhî ʾôr
and.say.WAYYIQTOL.3MS God be.JUSS.3MS light and.be.WAYYIQTOL.3MS light
'And God said, "Let there be light," and there was light.'

The participle in BH is a verbal adjective, used for ongoing or imminent action (hinnê ʾānōkî bāʾ "here I am coming"), not yet a tense.

Mishnaic Hebrew

MH reorganises the system into tenses (Segal 1927, A Grammar of Mishnaic Hebrew, local but cited only; Kutscher 1982; Heijmans (ed.) 2020, local CC BY):

• qatal = past;
• the participle = present (and immediate future);
• yiqtol = future and modal (wishes, commands, purpose);
• the consecutive forms wayyiqtol/weqatal disappear;
• היה + participle expresses the past habitual and the counterfactual: hāyâ ʾômēr "he used to say" (the Mishna's frame for sayings of the sages, Avot).

Modern Hebrew

ModH takes the MH system almost unchanged (05):

Two further MH → ModH changes: the nitpa'el (MH) as the past of the reflexive-passive beside hitpa'el, which survives only in literary style (UD IAHLT tags it NITPAEL, 21 tokens; 11); and the passive of pa'al (BH qal passive), which MH lost in favour of nif'al — ModH keeps only its participle (katuv "written").

10.4 Relative clauses: אשר and ש

BH relativises mainly with the free word אשר ʾăšer; the prefix še- (ša-) appears in a few, mostly late or northern, texts (Song of Songs, Ecclesiastes, Judges 5) (GKC § 36, § 138). MH uses only še-. ModH uses she- everywhere and asher in formal writing (07).

The article as relative (ha- + participle or, rarely, + finite verb) is BH (hā-ʿōśê "who does"; even ha-hôlîd "who begot" in Gen 21:3) and continues in formal ModH with participles.

Counted in UD 2.18 (2026-10-17), per 1 000 words: asher 0.9 in IAHLTwiki, 0.16 in IAHLTknesset, 3.5 in PostRab; she- (SCONJ) 18.6, 33.1 and 41.8. Formal written ModH keeps a little asher; speech almost none; the post-Rabbinic texts use both, with she- already dominant (MH influence).

10.5 Possession: construct, של and suffixes

של is a Mishnaic grammaticalisation of the relative še- + the preposition lǝ- "to, of" (Kutscher 1982; Segal 1927). In ModH it is a free word, inflecting with suffixes (sheli, shelkha, shelo…; 06). The construct remains productive for compounds (beyt séfer "school"), which is why UD makes it a compound:smixut rather than an ordinary genitive (11).

10.6 Pronouns and demonstratives

MH forms (ánu, zo, élu) survive in ModH as formal doublets beside the BH forms used in speech — one of the layered features a tagger must know (04).

10.7 Negation

The MH participial negation with inflected ʾên became the formal ModH eyno yodéa, while speech generalised lo to the participle. Counted in UD 2.18 (2026-10-17), the lemma אין against לא: PostRab 81 : 93, IAHLTwiki 78 : 459, IAHLTknesset 152 : 980 — the older texts negate with eyn almost as often as with lo, Modern Hebrew six times less. (Knesset eyn is mostly the existential "there isn't".)

10.8 Phonology: from Tiberian to Israeli

The pointed Bible preserves the Tiberian reading tradition of the Masoretes (8th–10th c. CE). Khan (2020) reconstructs it in detail from medieval sources: a system of seven vowel qualities with distinctions of length tied to stress and syllable structure, the pharyngeals ḥ and ʿ, the emphatics ṭ and ṣ and q, the spirantization of bgdkpt after vowels, gemination (dagesh forte), and a reduced vowel (shewa) with rules for when it was pronounced (Khan 2020, vol. 1, local CC BY khan-tiberian-pronunciation-v1-2020). Israeli pronunciation differs at almost every point (03):

The revivers chose a Sephardi-type pronunciation as the model — five vowels, qameṣ as [a], final stress — over the Ashkenazi tradition of their own prayers ([o] for qameṣ, [s] for undotted ת, penultimate stress; Zuckermann 2003, not held). In practice their European native phonetics shaped the result: uvular r, loss of pharyngeals and emphatics. The Mizrahi pronunciation keeps the pharyngeals and is closer to the Tiberian consonantism (01, 03). Other living reading traditions — Yemenite, Samaritan, the medieval Babylonian and Palestinian pointings — are studied in Khan and Hornkohl (eds.) 2020 (local CC BY khan-hornkohl-semitic-vocalisation-2020).

10.9 Word order

Whether BH is basically VSO or SVO with verb-first narrative forms is itself debated (Holmstedt 2011, local hbo-2011-holmstedt-typological-classification-hebrew-genesis-subject; Kaajan 2019, local hbo-2019-kaajan-biblical-hebrew-non-configurational-language). The shift to SVO in MH and ModH is usually attributed partly to Aramaic contact (MH) and partly to European contact (ModH).

10.10 The lexicon across the layers

See 08 § 8.6. In brief:

• Doublets with a register split: BH ets / MH ilan "tree"; BH shémesh / MH khama "sun"; BH lakákh / MH natal "take"; BH ánu / MH… the pronoun pairs above. ModH usually keeps the BH word as neutral and the MH one as literary — or the reverse (ilan is literary, but lehatkhil "to begin", hif'il of MH origin, is neutral).
• Aramaic words (ába, íma, agav, dávka, bérez, mada) enter through MH and the Talmud.
• Greek and Latin loans of MH (sanhedrin, pruzbul, kategor, sandal) — Heijmans (2020) studies vowel reduction in Greek loanwords in the Mishna, showing how they were adapted to Hebrew patterns.
• Medieval abstract vocabulary from the translators of Arabic (08).

10.11 European influence on Modern Hebrew

The revivers spoke Yiddish, Russian, Polish, German, French, Ladino, Arabic and other languages; their children spoke Hebrew with the phonetics and much of the idiom of their parents' languages. Influences commonly cited (Zuckermann 2003; Glinert 1989; Wexler 1990 — not held; see 01 § 1.2):

Arabic influence is mainly lexical and in slang, plus the Mizrahi phonetics (08); English influence is lexical, phraseological and orthographic (loan spellings with geresh, acronyms).

10.12 Medieval Hebrew and the Haskala

Medieval Hebrew was not one language but several written registers:

• Liturgical poetry (piyyut, from the 5th–6th c.: Yannai, Qillir) — daring new forms, many from MH, invented derivations.
• The Spanish school (10th–12th c.: Dunash, Shmuel ha-Nagid, Ibn Gabirol, Yehuda ha-Levi) — poetry in pure biblical Hebrew with Arabic metres; Martínez Delgado (2023) introduces Andalusi Hebrew metrics (local CC BY hbobook-obp0351-delgado-introduction-andalusi-hebrew-metrics).
• Philosophy and science translated from Arabic (the Ibn Tibbon family, 12th–13th c.) — the source of much abstract vocabulary (08).
• Rabbinic Hebrew of commentary, responsa and codes (Rashi, Maimonides' Mishne Tora in MH style, the Shulhan Arukh) — MH grammar with Aramaic admixture.
• Grammatical science: the medieval Hebrew grammarians (Saadia, Hayyuj, Ibn Janah, Ibn Ezra, Radak) worked in dialogue with Arabic grammar. Kantor (2023) studies the standard language ideology of the Hebrew and Arabic grammarians of the ʿAbbasid period — how both traditions came to regard one reading (Tiberian for Hebrew, the ʿArabiyya for Arabic) as the pure and correct language (not quoted: the book is CC BY-NC; local kantor-standard-language-ideology-2023). The ideology of a correct, classical standard against everyday usage reappears in the Modern debate between the Academy's norms and speech (08 § 8.12).

The Aramaic dalet ד־ "that, of" (from Aramaic dī/də-) is used in many rabbinic and later Hebrew texts beside Hebrew she- and shel. Tal, Shmidman and Shmidman (2026) built a BERT model for historical Hebrew with all prefixes segmented, used it to tell the dalet's two functions (subordinator vs possessive) apart with F1 ≈ 0.89, and charted the choice between dalet and Hebrew alternatives across regions and periods in more than 300 million words (local: aclbulk-2026-tal-modeling-dalet-clitic-historical-hebrew). Modern Hebrew keeps the dalet only in fixed Aramaic phrases (de-fákto is Latin; genuine ones: kol de-hu "anything at all", b-de'avad "after the fact").

The Haskala (Berlin from the 1780s, then Galicia and Russia) used Hebrew for secular prose, journalism, science popularisation and the first Hebrew novels (Abraham Mapu's Ahavat Tsiyon, 1853). Its style, melitsa, wove biblical phrases into new contexts and avoided MH; it was criticised by the next generation (Mendele Mokher Sforim, from the 1880s) who created the nusakh, a synthesis of all historical layers — the immediate literary ancestor of Modern written Hebrew (Sáenz-Badillos 1993; Harshav 1993 — not held).

10.13 Older Hebrew in the treebanks: PostRab and PTNK

Two UD treebanks hold pre-modern Hebrew:

• UD Ancient Hebrew PTNK (hbo; Biblical Hebrew; CC BY-NC-SA) — a separate language in UD; Mova measures its MACULA-based Biblical trees against it (tools/macula, test LAS 82.23 on 2026-10-17; 12).
• UD Hebrew PostRab (he; CC BY-SA; UD 2.18 has 296 test sentences, 8 029 words) — post-Rabbinic texts: Rashi's Torah commentary, the Shulhan Arukh, Sefer ha-Hinukh, Sefer Maharil, the cosmographical Igeret Orhot Olam, and a miscellany (treebank README; Tal et al. 2025).

PostRab is filed under he but its language is MH-based rabbinic Hebrew with Aramaic. The UD 2.18 counts (2026-10-17; 11) show the difference from Modern Hebrew:

The texts have no punctuation (sentence boundaries are an editorial decision), more clitics per token, fewer analytic possessives (shel), three times the pronouns of Wikipedia, more eyn, and Aramaic forms (the dalet in דבאוכלין "that in food", sentence 2 of the test file). Relations specific to PostRab — compound:svc (serial verbs) and compound:redup (reduplication) — do not occur in the Modern treebanks. A parser trained on Modern Hebrew will therefore stumble on: missing punctuation, Aramaic words and clitics, the inflected eyn, verbless and participial clauses of legal style, and abbreviations of rabbinic references.

10.14 One sentence through the ages

"The man said to his son: Go to the field."

BH:    ויאמר האיש אל בנו לך השדה
       wayyōmer hāʾîš ʾel bǝnô lēk haśśādeh
       and.say.WAYYIQTOL.3MS DEF-man to son-3MS.POSS go.IMP.2MS DEF-field
MH:    אמר האיש לבנו: לך לשדה
       ʾāmar hāʾîš libnô lēk laśśādeh
       say.PST.3MS DEF-man to-son-3MS.POSS go.IMP.2MS to.DEF-field
ModH:  האיש אמר לבן שלו: לך לשדה / תלך לשדה
       ha-ish amar la-ben shelo: lekh la-sade / telekh la-sade
       DEF-man say.PST.3MS to.DEF-son of.him go.IMP.2MS / go.FUT.2MS to.DEF-field

The BH line opens with the narrative wayyiqtol (VSO) and uses ʾel "to"; the goal of motion can be a bare noun (an accusative of direction), as here. MH has the plain past, l- "to". ModH has SVO, the analytic possessive shelo (or formal li-vno), and a choice of imperative or future-as-imperative (05, 09).

References

Works marked not held are not in the local library; see «wanted» in Sources.

• Andrason, Alexander. 2011. Biblical Hebrew wayyiqtol: a dynamic definition. Journal of Hebrew Scriptures 11. Local: hbo-2011-andrason-biblical-hebrew-wayyiqtol-dynamic-definition.
• Cook, John A. 2012. Time and the Biblical Hebrew Verb. Winona Lake: Eisenbrauns. Not held.
• Martínez Delgado, José. 2023. Introduction to Andalusi Hebrew Metrics. Cambridge: Open Book Publishers. Local: hbobook-obp0351-delgado-introduction-andalusi-hebrew-metrics.
• Doron, Edit. 2000. VSO and left-conjunct agreement: Biblical Hebrew vs. Modern Hebrew. In A. Carnie and E. Guilfoyle (eds.), The Syntax of Verb Initial Languages. OUP. Not held.
• Driver, S. R. 1892. A Treatise on the Use of the Tenses in Hebrew. 3rd ed. Oxford. Public domain; local driver-tenses-1892.
• Gesenius, W., E. Kautzsch and A. E. Cowley. 1910. Gesenius' Hebrew Grammar (GKC). 2nd English ed. Oxford. Public domain; local gesenius-kautzsch-cowley-1910.
• Glinert, Lewis. 1989. The Grammar of Modern Hebrew. CUP. Not held.
• Harshav, Benjamin. 1993. Language in Time of Revolution. University of California Press. Not held.
• Heijmans, Shai (ed.). 2020. Studies in Rabbinic Hebrew. Cambridge: Open Book Publishers. CC BY 4.0. doi.org/10.11647/OBP.0164 our card — local heijmans-studies-rabbinic-hebrew-2020.
• Holmstedt, Robert D. 2011. The typological classification of the Hebrew of Genesis: subject-verb or verb-subject? Journal of Hebrew Scriptures 11. Local: hbo-2011-holmstedt-typological-classification-hebrew-genesis-subject.
• Hornkohl, Aaron D. 2024. Diachronic Diversity in Classical Biblical Hebrew. Cambridge: Open Book Publishers. Local: hbobook-obp0433-hornkohl-diachronic-diversity-classical-biblical-hebrew.
• Isaksson, Bo. 2021. The Verb in Classical Hebrew. Open Book Publishers. Local: hbobook-obp0414-isaksson-verb-classical-hebrew.
• Kaajan, Marianne. 2019. Is Biblical Hebrew a non-configurational language? Reconsidering the evidence from discontinuous phrases. HIPHIL Novum 5(2). Local (read only): hbo-2019-kaajan-biblical-hebrew-non-configurational-language.
• Kantor, Benjamin Paul. 2023. The Standard Language Ideology of the Hebrew and Arabic Grammarians of the ʿAbbasid Period. Cambridge: Open Book Publishers. CC BY-NC 4.0. doi.org/10.11647/OBP.0382 — local kantor-standard-language-ideology-2023 (summarised, not quoted).
• Khan, Geoffrey. 2020. The Tiberian Pronunciation Tradition of Biblical Hebrew. 2 vols. Cambridge: Open Book Publishers. CC BY. doi.org/10.11647/OBP.0163 our card — local khan-tiberian-pronunciation-v1-2020, …-v2-2020/.
• Khan, Geoffrey and Aaron D. Hornkohl (eds.). 2020. Studies in Semitic Vocalisation and Reading Traditions. Open Book Publishers. Local: khan-hornkohl-semitic-vocalisation-2020.
• Kutscher, Eduard Yechezkel. 1982. A History of the Hebrew Language. Jerusalem: Magnes; Leiden: Brill. Not held.
• Rezetko, Robert and Ian Young. 2014. Historical Linguistics and Biblical Hebrew. Atlanta: SBL. Local: hbobook-2014-rezetko-historical-linguistics-biblical-hebrew.
• Sáenz-Badillos, Angel. 1993. A History of the Hebrew Language. Trans. J. Elwolde. CUP. Not held.
• Segal, M. H. 1927. A Grammar of Mishnaic Hebrew. Oxford: Clarendon. Local (logs and citation only): segal-grammar-mishnaic-1927. Also Segal 1909, Mišnaic Hebrew and its Relation to Biblical Hebrew and to Aramaic (JQR 20), local segal-misnaic-hebrew-1909.
• Tal, Rachel, Cheyn Shmuel Shmidman and Avi Shmidman. 2026. Modeling the Dalet Clitic in Historical Hebrew Texts: A New Prefix-Segmented BERT Model and Stylistic Analysis. Local: aclbulk-2026-tal-modeling-dalet-clitic-historical-hebrew.
• Tal, Rachel et al. 2025. A New Hebrew Universal Dependency Treebank: The First Treebank of Post-Rabbinic Historical Hebrew. TLT 2025. aclanthology.org/2025.tlt-1.11 our card
• UD 2.18 treebanks UD_Hebrew-PostRab, -IAHLTwiki, -IAHLTknesset (CC BY-SA 4.0). Local ud-2.18 (counts 2026-10-17).
• Wexler, Paul. 1990. The Schizoid Nature of Modern Hebrew. Harrassowitz. Not held.
• Zuckermann, Ghil'ad. 2003. Language Contact and Lexical Enrichment in Israeli Hebrew. Palgrave Macmillan. Not held.

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/languages/he/10-historical-layers/