Elamur Research⌕
☰
home / papers / 2026 / Iʿrāb, Tanwīn and Status Constructus: Some Thoughts on the History of the Nominal Declension in Semitic and Arabic

Iʿrāb, Tanwīn and Status Constructus: Some Thoughts on the History of the Nominal Declension in Semitic and Arabic

Summary

The paper asks how much of the Classical Arabic noun declension is truly inherited from Common Semitic. Arabic keeps case endings and nasal endings. This looks like Akkadian-style conservatism, but some parts may be later systematization. Stempel compares endings for the singular, dual and plural, and for the status constructus, across Akkadian, Hebrew, Syriac, Arabic and Geʿez. He then reconstructs the proto-forms and tests them against sound laws and typology.

His main claims are:

  • The Arabic singular is essentially inherited.
  • The Arabic dual and plural carry an extra final vowel shared with Proto-Hebrew.
  • The fully case-inflected construct singular is an Arabic innovation. Originally the construct was an endingless casus indefinitus, as still seen in Akkadian and inferable for Hebrew. The likely model was a genitive pattern built with a relative pronoun.
  • Nunation (tanwīn) first carried no meaning. It came to mark indefiniteness because its n-less sandhi variant was frequent in definite genitive contexts.

The work matters because it shows that apparently archaic morphology can be a reorganized system.

Key points

  • Reconstructed Common Semitic paradigm:
    • Singular: nom. -um, acc. -am, gen. -im.
    • Dual: nom. -ān, oblique -ayn.
    • Two plurals: a general one (-ū/-ī) and a distributive one (-ūnu/-īni) for individual members. Akkadian šarrū means kings as a group; šarrūnu means individual kings.
  • Singular and dual sound changes:
    • Arabic -un/-an/-in arise from the common change of final -m to -n.
    • Akkadian gradually lost mimation, and forms with and without it alternated freely.
    • Geʿez dual -ē continues *-ayn after losing the nasal.
  • Arabic plural: -ūna/-īna is analysed as a back-formation. Once the absolute form was felt to equal the construct plus a nasal, a nasal was added to the construct forms -ū/-ī.
  • Hebrew evidence for an endingless construct: dəbar-ka 'your word' can only continue a form without a case vowel. This matches the endingless Akkadian construct singular.
  • Possessive models and Ethiopic:
    • The genitive pattern {noun + relative pronoun + genitive} is attested in several languages: Akkadian ša, Syriac d-, Geʿez za-, and Moroccan Arabic dyāl.
    • Egyptian Arabic uses a possession noun instead, bitāʿ.
    • Ethiopic behaves as a marginal language that preserves archaisms.
    • The PDF header gives the issue as Folia Orientalia 44, 2008.

Interesting ideas

  • I1. Proportional back-formation: When one paradigm cell is felt to equal another cell plus an affix, speakers fill gaps by applying that proportion. A learner could induce such morphology relations instead of memorizing endings. — For Mova (tokens): The en lemmatizer could mine proportions like base : base+affix across paradigms to guess lemmas and forms of unseen words.
  • I2. Unmarked forms where a word is not syntactically bound: Words outside government relations surface without grammatical marking, as in vocatives, the first member of compounds, and predicates. This gives a structural cue for why marking disappears. — For Mova (tree): In English noun–noun compounds the first member usually lacks plural marking. The parser could use this as a feature to separate compound modifiers from heads and to recognize vocatives.
  • I3. A sandhi variant can acquire meaning by context frequency: A phonetic variant common in one construction (definite genitives) gradually becomes a marker of that construction's meaning. — For Mova (learn): When building silver data, track which surface variants cluster with which constructions, so that contractions or reduced forms are treated as possible signals rather than noise.
  • I4. Group plural versus individuals plural: The paper separates a collective plural from a distributive one that counts individual members. — For Mova (world): The world model could mark plural entities as collective or distributive, which affects how quantities are summed in maths word problems.
  • I5. Proper names stay conservative and inherently definite: Names resist analogical change and are definite regardless of their marking. — For Mova (frames): Coreference could treat proper names as definite anchors even without an article, and give them priority as antecedents.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Akkadian and cuneiform NLP

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{stempel2026tanwin,
  title = {{Iʿrāb, Tanwīn and Status Constructus: Some Thoughts on the History of the Nominal Declension in Semitic and Arabic}},
  author = {Stempel, Reinhard},
  booktitle = {Folia Orientalia 44},
  year = {2026},
  url = {https://wydawnictwo.pan.pl/index.php/fo/article/view/6630}
}

References

14 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Benveniste, E. 1974. Problemes de linguistique generale. Paris: Gallimard.[Here cited: Chapitre VI. Structure de la langue et structure de la societe, pp. 91-102, originally published in 1970]. Fischer, W.; 0. find
  2. Jastrow (eds.) 1980. Handbuch der arabischen Dialekte. Wiesbaden. (Porta Linguarum Orientalium, N.S. 16). GAG = W. von Soden. 1969. Grundriss der akkadischen Grammatik. Roma2; W.von Soden; W.R. find
  3. Mayer. 1995. Grundriss der akkadischen Grammatik, 3., erganzte Auflage. Roma: Pontificium Institutum Biblicum. (Analecta Orientalia 33). find
  4. Goetze, A. 1946. The Akkadian masculine plural in anii/i: and its Semitic background. Language 22. 121-131. find
  5. Gruber-Miller, A. M. 1990. Loss of Nominal Case Endings in the Modem Arabic Dialects: Evidence from Southern Palestinian Christian Middle Arabic Texts. PAL I: 235-253. find
  6. Heath, J. 2002. Jewish and Muslim Dialects of Moroccan Arabic. London, New York: Routledge-Curzon. (Routledge-Curzon Arabic Linguistics Series). Hehr.Gr. = W. Gesenius. 1991. Hebraische Grammatik vollig umgearbeitet von E. Kautzsch. 6. Nachdruckauflage der 28. vielfach verbesserten und vennehrten Auflage. Leipzig 1909. Lizenzausgabe Dannstadt: Wissenschaftl. Buchgesellschaft. find
  7. Noldeke, Th. 1966. Kurzgefasste syrische Grammatik. Bearbeitet von Anton Schall. Dannstadt: Wiss. Buchgesellschaft. /CRAB, Tanwfn and Status Constructus: Some Thoughts on the History ... find
  8. PAL I= Mushira Eid (ed.) 1990. Perspectives on Arabic Linguistics. Papers from the First Annual Symposium on Arabic Linguistics. Amsterdam Philadelphia: Benjamins 1990. (Amsterdam Studies in the Theory and History of Linguistic Science. Ser. IV Current Issues in Linguistic Theory. Vol. 63). find
  9. Solmsen, F. 1922. Indogermanische Eigennamen als Spiegel der Kulturgeschichte. Heidelberg: Winter. find
  10. Stempel, R. 1999. Abriss einer historischen Gramrnatik der semitischen Sprachen. Frankfurt a.M. et al.: Lang. (Nordostafrikanisch/Westasiatische Studien 3). find
  11. Stempel, R. 2000. Akzent und Vokalismus im Hebraischen. In: P. Nagel (ed.) Dankesgabe fiir Heinrich Schiitzinger zum 7 5. Geburtstag. 179-188. Halle. (Hallesche Beitrage zur Orientwissenschaft 29/2000). find
  12. Stempel, R. 2008. Zur Geschichte des bestimrnten Artikels in den semitischen Sprachen. In: B. Huber, M. Volkart, P. Widmer (eds.) Chomolangma, Demawend und Kasbek. Festschrift fiir Roland Bielmeier zu seinem 65. Geburtstag. Band II. 551-560. Halle. find
  13. Stempel, R. 2009. The genitive construction in Morrocan Arabic: an archaism. In: L. Abu-Shams (ed.) Actas de! III Congreso Internacional de Arabe Marroqui estudio, ensefianza y aprendizaje. (Vitoria-Gasteiz 2008). Bilbao: Servicio Editorial de la UPV. 177-182. find
  14. Zaborski, A. 2000. Inflected Article in Proto-Arabic and Some Other West Semitic Languages. Asian and African Studies 9. 24-35. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/stempel-2026-irab-tanwin-status-constructus-some/