I.2. Afroasiatic

I.2.1. Afrasian

It is very difficult to reconstruct ancient Proto-Afroasiatic (PAA) vocabulary, and still more difficult to reconstruct a common morphosyntax. These are some known features common to more than one branch, though (Hodge 1971; Lecarme, Lowenstamm, and Shlonsky 2000; Frajzyngier and Shay 2012):

·       Verb initial (Egyptian, Semitic, some Berber, Central Chadic). Clause-final position (Omotic, Cushitic, Akkadian) considered a product of contact with other languages.

·       Case marking (Semitic, Berber, Cushitic, Omotic, maybe also Egyptian): subject and object less overtly marked. Probably similar to Proto-Semitic:

o   Nominative in *-u,

o   Genitive in *-i,

o   Focalised element (object, vocative) in *-a.

o   Construct state in *-Ø (in Semitic, Cushitic).

o   Plural in ablaut to *-a- (Semitic, Berber, Cushitic, Chadic), suffix *- (Semitic, Berber, Cushitic, Chadic), or ending *-t (the only one in Omotic).

o   Feminine in *-(a)t.

·       Maybe adverbial (locative?) in *-iš as found in Semitic and Egyptian.

·       Adjective forming suffix (from genitive) *-(a/i)-.

·       Apparently, verb stems ended in consonant; pronoun and indeclinable stems could end in vowel; and nouns and adjectives were distinguished by the so-called ‘terminal vowel’.

·       Verbal system:

o   Two tense/aspectual systems (Chadic, Egyptian, Semitic, Cushitic).

o   Vowel alternation or Ablaut codes a variety of functions.

o   Prefix-conjugation as the most ancient one (Pre-PAA), continued by suffix conjugation and subject agreement.

o   Extensions of the verb: causative in *s, inchoative/denominative * (and *), non-finitive *n, durative *t, stative/intransitive *m, middle voice *dl, amplificative *h, complementive *ɣw, etc.

o   No initial vowels or consonant clusters.

o   One or more negative markers depending on the position of the verb in the clause.

o   Sequential clauses marked by markers, converbs (Cushitic, Omotic, Semitic, Chadic) or by tense forms (Berber, Egyptian).

bagu ʕar-da

bagu ma hai ĉaʕara,

araˀ ʕaar;

aḳul tu ĉaˀa ḳ(w)urī,

aʒub tu ḳama gadī,

tu nasa pidiš.

akaˀ bagu ʕaar:

adum ani libbu,

raˀi ahim nasu ʕaar.”

akaˀ ʕar: “gur baga!

adum n libbu raˀi

apal nasu, ˀadu, kur

ĉaʕar-bagi kica sirfī,

ma aha bagu ĉaʕarī.”

ta kina abuḳ bagu ʔaqa


·       For ‘and’, the reconstructed ‘comitative’ case *-dV / *-Vd ‘along with, together with, in addition to’ is used (hypothesised to be an ancient postposition).

·       For plural forms, the ending *- is used for nominative, ablaut in *-a- for the accusative (although this was most likely not the case, and these forms alternated, since cases were probably not marked in the plural).

·       There is no reconstructed word for ‘hear, listen’, hence *gur- ‘sound, voice’ is used as a verbal stem for the ‘imperative’, and *kin- ‘know, learn’ for ‘having heard’.

I.2.2. Semitic

These are some general features reconstructed for Proto-Semitic (Weninger et al. 2011):

·       General VSO order.

·       Noun:

o   PSem. feminine is marked by *-at.

o   Singular case paradigm nom. *-u, gen *-i, acc. *-a.

o   Plural m. nom. *ū, gen/acc. -ī, pl. fem. nom. -ātu, gen-acc. -āti. Internal plurals (not occurring in Akkadian) are assumed to be a secondary feature spread by areal diffusion, hence not central to a PSem. reconstruction (although potentially quite old).

o   Less clearly reconstructible for PSem. (and probably specialised in East Semitic) are a locative *-u (or *-ū), and a terminative *-.

·       Simple syntax of preposition + genitive. Formation of prepositions from the construct state of the accusative, cf. Classical Arab bana ‘between’ < banun ‘separation; interval’. Reconstructible PSem. preposition include *ˀad ‘until, to’.

·       Common prepositions include *a ‘and’, *a ‘or’, and *šimmā ‘if’.

·       Adverbial ending Akk. *-, Syr. -(ˀ)īϑ, or (indefinite) accusative in Akkadian (-am), Classical Arab (-an).

·       Interrogatives generally in *ˀa-, also *man ‘who’, *matī/a ‘when’; negations in *ˀal/ˀul, *ˀiV, *lā.

·       Verbal system:

o   Finite forms show a great stability over millennia, with the usual paradigm being:

§  Singular: 3m. *i-_-Ø, 3f. *ta-_-Ø; 2m. *ta-_-Ø, 2f. *ta_-ī; 1c. *ˀa-_-Ø;

§  Plural: 3m. *i_ ū, 3f. *i_-ā, 2m. *ta-_-ū, 2f. *ta_-ā, 1c. *ni-_-Ø.

o   The ‘perfect’ is an innovation of West Semitic (WS), and seems to have evolved from the Akkadian stative, which did not have an ending for the 3sg.m. (although cf. *-a in WS, considered as from the absolutive). In this fable, the 3sg. form is left without ending.

o   Many afformatives in nominal derivation can be reconstructed for PSem. The most relevant ones are *ma-, *mi-, *mu-, *ta-, *ti-, *ˀa-, *ˀi-, *ˀu-, *-ān.

raḫilu a muhrū

raḫilu lā ihu šipāti,

iˀumr muhrī;

iṣaar ða arkaba kabita,

iϑ̣akam ða aϑ̣kama rabb,

ða ˀinša nidiš.

iḳul raḫilu muhrī:

iagaʕ-nī libbu,

ḥizai irkaba ˀinšu parašī.”

iḳlū muhrū: “šmuʕ raḫilu!

iagaʕ-nā libbu ḥizaī

iʕabad ˀinšu, baʕlu, tāṉ

šipātu-ṣ̌aˀni kista šaḫn,

a lā ihaa ṣ̌aˀnu šipāti.”

hā šimaʕ ibluṭ raḫilu ˀad šadai.


·       For individual ‘sheep’ *raḫil is selected. For the plural (collective) forms, *ṣ̌a’n seems more appropriate.

·       PSem. *muhr, ‘foal’, has been selected as the oldest reconstructible stem for ‘horse’. The other widespread ancestral root is *ss-, which seems a borrowing from an Indo-European source, akin to Luw azzuwa, or PIAr. *áśvās (via Mitanni). Maybe the term for ‘foal’ was the original, although it is more likely that the donkey was the usual Middle Eastern/African animal before the introduction of the horse.

·       For ‘having’, a head noun ‘wool’ in construct is used.

·       For ‘chariot’, a form of the root *rkb ‘ride, drive, mount’ is used, as in Akkadian narkabtum ‘carriage’.

·       PSem. *ṣr ‘carry (on shoulders)’, and *ṯ̣km ‘carry on shoulders/back’.

I.2.3. Northwest Semitic

Features of WS, some of them innovations of a previous Central Semitic (CS) stage include (Weninger et al. 2011):

·       General VSO, possessor possessed, and noun-adjective order, as in PSem.

·       PSem. initial *- became *- (except for conjunction *a- where *-- was probably interpreted as word-medial).

·       PSem. *n- became assimialted to an immediately following consonant except for *h in several cases.

·       NWS determinative-relative pronoun *ðū inflected in case, number, gender.

·       Deixis in *h-, *n-, *ð-, *l-, *k-.

·       ‘Triptotic’ declension in the singular, ‘diptotic’ in the dual and plural.

·       Adjectives regularly agree with nouns, inflect in masc. /fem. gender and number.

·       Prepositions can be reconstructed for WS: *bi- ‘in’, *la- ‘to’, *ʕala ‘on’, *min ‘from’, etc.

·       Conjunction *pa- ‘and, then’ as Central Semitic innovation.

·       Verbal system:

o   Imperfect or ‘prefix conjugation’ with an older and a younger type:

§  The “short imperfect” or “jussive”, (cf. 3.m.sg. *aktub-Ø), akin to Akkadian preterite, expresses deontic modality and punctual past.

§  The “long imperfect” (cf. 3.m.sg. *aktub-u, with long vowels of the 2f.sg. and 3/2pl. expanded by *nV), is used in CS for present-future tense, durative or iterative past, circumstancial events etc.

o   “Perfect” or “suffix conjugation” (cf. 3m.sg. *kataba for active verbs) expresses different types of past tense or completed action, from a previous stage denoting timeless qualities or mental stages (cf. *kabida ‘he is heavy’)

o   Subjunctive (CS) acts as volitive, and maybe indicates subordination, cf. 3m.sg. *aktub-a.

o   “Energetic” mood forms in *-(a)nna (also *-(u)nna?), an innovation of Central Semitic.

o   Imperative formed on the base of the ‘short imperfect’.

o   Active participle *kātib- (for sound fientic roots) inflects initially like a noun.

o   Passive participle forms *katīb and *katūb.

o   Different levellings of the verbal system, e.g. drop of initial * in verbs I (originally I); assmilation of *n or *l in “imperfect”; etc.

o   Internal passives introduced before NWS (cf. *kutiba vs. *kataba)

o   Levelling of vowels in verb prefixes: Proto-Semitic, informed by Akkadian, showed *ˀa-, *ta-, *ni-, *i-.


raḫilu a parašū

raḫilu lā ahua šapāti,

ˀamara parašī;

ṣāiru ðū markabta kabita,

ϑ̣ākimu ðū ϑ̣aʕina rabba,

ðū ˀanāša ida.

qaul raḫilu parašī:

aguʕu libbu-nī,

ḥāzii rākiba ˀanāšu parašī.”

qaulū parašū: “ušmuʕ raḫilu!

aguʕu libbu-nā ḥāziī

apʕulu ˀanāšu, baʕlu, tāṉ

šapātu-raḫili kista šaḫana,

a lā ahuu raḫilu šapāti.”

hā šamīʕ abruḥ raḫilu la-dabri.


·       For ‘horse’, the West Semitic *paraš has been used.

·       Active participle *ϑ̣aʿīn ‘loaded with’ = carrying as in Syriac.