Wikipedia / Proto-Circassian language
Wikipedia / Proto-Circassian language
Proto-Circassian (or Proto-Adyghe–Kabardian) is the reconstructed common ancestor of the Adyghean and Kabardian languages.
Phonology
Note on Orthography: This article employs Cyrillic characters alongside IPA to assist readers familiar with the Circassian alphabet. However, standard Adyghe orthography contains inconsistencies; notably, the digraph кӀ represents the sound [t͡ʃʼ] despite visually suggesting a velar ejective [kʼ]. To ensure phonetic precision, this article utilizes the following distinctions: чӀ for [t͡ʃʼ], чӀъ for [ʈʂʼ], and кӀь for [kʲʼ].
Consonants
The consonant system is reconstructed with a four-way phonation contrast in stops and affricates, and a two-way contrast in fricatives.
Shifts
The following table shows how Proto-Circassian consonants diverged and evolved in the Western and Eastern dialects.
Specific Shifts
Aspirated, plain and tense consonants
In Proto-Circassian each voiceless obstruent could appear in up to three manners: an aspirated (lax) variant, a plain variant, and a tense (geminate) variant.
The aspirated series survived in the Shapsug and Bzhedug dialects, while it became plain in the other dialects (Abzakh, Chemgui, Kabardian).
The tense series survived in the western dialects (Shapsug, Hatuqay, Bzhedug), merged into plain voiceless consonants in Chemgui (Temirgoy), and became voiced in the eastern dialects (Besleney and Kabardian). In linguistics literature the tense consonants are often written with a colon (e.g. ) to denote gemination/tenseness, and in some Cyrillic orthographies with double letters (e.g. пп, тт, цц) or a colon (e.g. п:, т:, ц:).
In the tables below, dialects are grouped under West Circassian (Shapsug, Bzhedugh Hatuqay, Chemgui, Abzakh) and East Circassian (Besleney where attested, and Kabardian). Bzhedugh and Hatuqay share a single column. The subsections group each consonant's aspirated and tense reflexes together.
Bilabial p
Aspirated: →
Tense: →
Dental t
Aspirated: →
Tense: →
Front affricate t͡s
Aspirated: →
Tense: →
Middle affricate t͡ɕʷ (labialized)
Aspirated: →
Tense: →
Palatal affricate kʲ → t͡ʃ
Aspirated: →
Tense: → /
Labialized velar kʷ
Aspirated: →
Tense: →
Uvular q
Aspirated: → (Kabardian )
Tense: → (Adyghe); (Kabardian)
Labialized uvular qʷ
Aspirated: → (Kabardian )
Tense: → (Adyghe); (Kabardian)
Postalveolar fricative ʃ
Aspirated: → /
Retroflex variant: →
Postalveolar fricative ʃ (additional lax–tense reflexes)
The book records further lax and tense back-sibilant reflexes, including a glottalized member.
Back affricates t͡ʃ / t͡ʂ (tense)
Tense palato-alveolar: → /
Tense retroflex: → /
Back affricates t͡ʃ (lax and plain, full dialect reflexes)
The lax (aspirated) and plain back affricates show distinct outcomes across all dialects, including Besleney.
Front affricate t͡s (additional lax–tense reflexes)
The book records further front-affricate examples, including tense reflexes that surface as in Kabardian.
Back spirant ʒ (voiced)
Voiced back spirant: → Adyghe and Besleney : Kabardian .
The Two Major Shifts
The consonant system of the modern Circassian dialects looks the way it does because of two sound changes that swept through the language at different times. To understand them, it helps to start with the ancestor language, Proto-Circassian, and see what it had that the modern dialects have since reshaped.
Proto-Circassian had a rich set of stops and affricates. (An affricate is a single sound that begins like a stop and releases like a fricative — think of the "ch" in English church or the "j" in judge.) Crucially, Proto-Circassian distinguished three groups of sounds that would later become tangled together:
Postalveolar affricates (made with the tongue just behind the ridge behind the teeth):
Retroflex affricates (made with the tongue curled further back):
Palatalized velars (k- and g-type sounds pronounced with the tongue raised toward the hard palate, giving them a "soft", y-coloured quality):
The key thing to notice is that the palatalized velars (гь, кь, кӀь) were originally completely separate from the affricates (дж, ч, чӀ). They sounded different and behaved differently. The two sound changes described below are what eventually caused some dialects to merge these two groups, which is why the modern picture can look confusing at first.
The two phases are: Spirantization (Phase 1), in which old affricates turned into fricatives, and Velar Palatalization (Phase 2), in which the palatalized velars turned into affricates. They did not happen everywhere, and they did not happen in the same dialects — which is exactly why comparing dialects lets us reconstruct the original system.
Phase 1: Spirantization (Affricate → Fricative)
Spirantization simply means "turning into a fricative" (a spirant is just another word for a fricative — a hissing sound like sh, or zh, where air flows continuously instead of being briefly stopped). In Phase 1, the original Proto-Circassian postalveolar and retroflex affricates lost their initial "stop" portion and became plain fricatives.
In other words, a sound like (the "j" in judge) simply dropped its stop beginning and was left as (the "s" in measure). The same thing happened to the whole series.
Affected Dialects: Abzakh, Modern Standard Kabardian, and archaic Kabardian dialects.
Unaffected Dialects: Shapsug, Bzhedug, Chemguy, and Besleney. (These dialects kept the original affricates intact, which is why they are so valuable for reconstructing the older system.)
The individual shifts were:
→
→ /
→
→ /
→
Notice that the plain affricates could end up as either / or depending on the dialect (compare Abzakh, which favours the voiceless ш/шъ, with Kabardian, which often voices these to ж). The ejective affricates (чӀ, чӀъ — the ones with a "popping", glottalic release marked by Ӏ) both collapsed into the single ejective fricative .
There is also one dialect-specific quirk worth noting. In Abzakh, the ejective frequently did not become a fricative at all, but instead weakened all the way to a palatalized glottal stop — essentially a catch in the throat rather than a hiss.
A few worked examples make the pattern concrete:
Proto-Circassian чӀэкӀьын "to come out from under" → щӀэкӀьын in Proto-Kabardian, but ӀекӀьын in Proto-Abzakh (showing the чӀ → Ӏь weakening).
Proto-Circassian чӀэгъуэжьын "to regret" → щӀэгъуэжын in Kabardian, Ӏегъуэжьын in Abzakh.
Proto-Circassian пачӀэ "mustache" → пащӀэ in both Kabardian and Abzakh.
Proto-Circassian чӀалэ "boy; young man" → щӀалэ in Kabardian, but Ӏелэ in Abzakh.
Proto-Circassian чӀымахуэ "winter" → щӀымахуэ in Kabardian, щӀымафэ in Abzakh.
Proto-Circassian чӀэ "new" → щӀэ in both Kabardian and Abzakh.
There is one important catch that sets up Phase 2. In the dialects affected by Phase 1 only (some archaic Kabardian dialects) or by both phases (Standard Kabardian and Abzakh), the change applied to the original affricates but left the original palatalized velars untouched — for the time being. So at this stage, words built on the old velar кӀь were still pronounced with : кӀьапсэ "rope", кӀьэрахъуэ "gun", and кӀьэ "tail" all still had a genuine k-sound. They had not yet been pulled into the affricate series. That only happened later, in Phase 2.
The following table demonstrates the Phase 1 shift (affricate → fricative), grouped by individual sound change. It shows how Abzakh and Standard Kabardian innovated, while Bzhedug, Chemguy, and Besleney stayed conservative for these particular consonants.
Phase 2: Velar Palatalization
Phase 2 came later, and in a sense it is the mirror image of Phase 1. Where Phase 1 turned affricates into fricatives, Phase 2 turned the leftover palatalized velars into affricates. The technical name is velar palatalization (or affrication of velars): the "soft" k- and g-sounds (гь, кь, кӀь) shifted their place of articulation forward and became palato-alveolar affricates — дж , ч , and чӀ respectively.
This is the change that finally merges the two groups in the dialects that underwent both phases. Once the old affricates had vacated their "slots" by becoming fricatives (Phase 1), the velars moved in and filled those slots by becoming the new affricates (Phase 2). The result is that a modern Standard Kabardian чӀ is not the original чӀ at all — the original one became щӀ in Phase 1 — but rather a brand-new чӀ created out of the old кӀь.
Affected Dialects: Abzakh, Bzhedug, Chemguy, and Modern Standard Kabardian.
Unaffected Dialects: Shapsug, Besleney, and archaic Kabardian dialects. (In these, the soft velars кь, гь, кӀь survive to this day.)
The individual shifts were:
→
→
→
This is why words that were still pronounced with a velar at the end of Phase 1 are now pronounced as affricates in these dialects. The words чӀапсэ "rope", чӀэрахъуэ "gun", and чӀэ "tail" — which had genuine кӀь a moment ago — are now said with .
Summary of Dialectal Evolution
Because the two phases were independent, each dialect can be classified simply by asking two yes/no questions: did it undergo Phase 1, and did it undergo Phase 2? The four possible answers give four neat groups:
Shapsug & Besleney — neither phase. These are the most conservative dialects. They preserve both the original affricates and the original palatalized velars as separate sounds, which is precisely why linguists lean on them to reconstruct Proto-Circassian.
Bzhedug & Chemguy — Phase 2 only. They kept the original affricates but shifted their velars into affricates. Because the old affricates were still present, the new ones from the velars simply merged in with them.
Some Kabardian dialects (e.g. Uzunyayla) — Phase 1 only. They turned the original affricates into fricatives, but their soft velars never advanced, so words like кӀьэ "tail" still keep their .
Abzakh & Modern Standard Kabardian — both phases. First the original affricates spirantized (Phase 1), and then the velars moved in to become the new affricates (Phase 2). This two-step "chain shift" is why a modern чӀ in these dialects descends from an old velar кӀь, not from the old affricate чӀ (which is now щӀ).
The cleanest way to see all of this is to follow two words side by side: one built on an original affricate (so it only ever reacts to Phase 1) and one built on an original palatalized velar (so it only ever reacts to Phase 2). The table below does exactly that.
Introduction of Labiodental Fricatives
Proto-Circassian lacked the labiodental fricatives: the voiceless labiodental fricative and the voiced labiodental fricative . These sounds developed independently in the Western (Adyghe) and Eastern (Kabardian) dialects from completely different phonological sources.
In summary:
Western Dialects derived [f] from the velar .
Eastern Dialects derived [f] and [v] from labialized sibilants (, , ).
Western Shift (Velar to Labiodental)
In Western Circassian dialects (such as Abzakh, Bzhedug, Chemguy, and Shapsug), the Proto-Circassian labialized voiceless velar fricative shifted to the voiceless labiodental fricative .
In Eastern dialects (Kabardian and Besleney), this consonant remained a velar .
→
Eastern Shift (Sibilant to Labiodental)
In Eastern Circassian dialects, a series of Proto-Circassian labialized postalveolar consonants (, , , ) shifted to become labiodental consonants.
In Western dialects, these consonants retained their original sibilant nature (pronounced as , , , ).
The specific shifts in Eastern dialects were:
→
→
→
→
→
Divergence Example: "Light" vs. "Blind"
An interesting consequence of these opposing phonological shifts is the divergence of the words for "light" and "blind".
Proto-Circassian possessed two distinct words:
Light: (containing the velar )
Blind: (containing the sibilant )
Due to the different sources of the labiodental in the dialects, these words shifted in opposite directions:
In Western (Adyghe): The velar became . Therefore, "light" became . The sibilant was retained, so "blind" remained .
In Eastern (Kabardian): The sibilant became . Therefore, "blind" became . The velar was retained, so "light" remained .
As a result, the word is a "false friend" between the dialects: it means light in Adyghe, but blind in Kabardian.
Morphological Impact: Benefactive vs. Malefactive Prefixes
The phonological shifts also affected grammatical prefixes, creating distinct forms for the Benefactive and Malefactive prefixes in the modern dialects.
In Proto-Circassian, these prefixes were distinct:
Benefactive ("for"): — containing the velar .
Malefactive ("to the detriment of"): — containing the ejective sibilant .
Due to the shifts:
In Western (Adyghe): The Benefactive became , while the Malefactive was retained as .
In Eastern (Kabardian): The Benefactive was retained as , while the Malefactive shifted to .
Examples of verbs utilizing these prefixes:
Glottalized uvular qʼ
The Common Circassian glottalized uvular *qʼ became a plain glottal stop in most Circassian dialects, while it was preserved as in the Hakutsw (Hakuchey) subdialect of Shapsug.
Grammar
Dynamic prefix *wa-
In Proto-Circassian, the dynamic prefix is reconstructed as *wa- (уэ-). This verbal morpheme spans across several grammatical categories, strictly indicating that a verb is operating in the positive, dynamic, present tense. As the languages diverged, this prefix underwent distinct phonological shifts:In some specific verb roots, the East Circassian (Kabardian) branch preserves the ancestral *wa- (уэ-) more transparently than West Circassian. For instance, in the dynamic verb for "standing/spending time," Kabardian retains the form щ-уэ-ты-р, whereas West Circassian fully reduces the prefix to щ-э-ты.
The historical reduction of the ancestral *wa- (уэ-) to the modern -э- in West Circassian is clearly visible in the conjugation of monovalent intransitive verbs without prefixes (such as "to go"). Through a process of vowel dropping and assimilation, the original forms simplified:
1st Person Singular: сэ-кӀо (I go) < *с-уэ-кӀо < *сы-уэ-кӀо
2nd Person Singular: о-кӀо / уэ-кӀо (You go) < *у-уэ-кӀо < *уы-уэ-кӀо
1st Person Plural: тэ-кӀо (We go) < *т-уэ-кӀо < *ты-уэ-кӀо
2nd Person Plural: шъо-кӀо / шъу-э-кӀо (You all go) < *шъу-уэ-кӀо < *шъуы-уэ-кӀо
The presence of the dynamic prefix is highly restricted. It is completely dropped from the verb complex under the following grammatical conditions:
Non-Present Tenses: It is omitted in the past and future tenses, allowing personal pronouns to attach directly to the verb root or relational prefixes. For example, while the dynamic prefix is active in the present tense, it is entirely absent in past forms such as сы-кIуагъ (I went), у-кIуагъ (You went), кIуагъэ (He went), and с-тхыгъэ (I wrote that).Negative Polarity: It is dropped in negative forms across all tenses, including the present tense.Conditionals and Conjunctions: The addition of suffixes such as -мэ (if), -ми (even if), or adverbial suffixes like -у (while) strips the verb of its dynamic prefix.
The мэ- (ma-) Mutation
The morphological placement of the dynamic prefix typically falls between the personal pronoun prefixes and the verb root. However, a significant structural mutation occurs in monovalent intransitive verbs when conjugated for the third person (he/she/it/they).
Because the third-person absolutive index in Circassian is a null prefix (∅-), the dynamic prefix is forced into the absolute word-initial position. Since Circassian phonotactics do not permit the dynamic vowels (-э- or -о-) to stand alone at the beginning of a word without a preceding consonant, the prefix mutates into мэ- (ma-) in both the West and East branches (e.g., ∅-э-кӀо mutates to макӀо).
This mutation is exclusively a feature of the positive present tense, as the dynamic prefix drops out entirely under other conditions. When the dynamic prefix is omitted, the мэ- mutation naturally disappears alongside it, leaving the root exposed:
Non-Present Tenses:Negative Polarity:Conditionals and Conjunctions:
Intransitive Verbs with Spatial Prefixes
When a monovalent intransitive verb includes a spatial or directional prefix (such as къ- meaning "hither" or щ- indicating location), the dynamic prefix is sandwiched between this spatial prefix and the verb root. In these environments across all grammatical persons, the ancestral *wa- consistently reduces to -э- in modern West Circassian:
Bivalent Intransitive Verbs
Bivalent intransitive verbs take an absolutive subject and an oblique indirect object. Because the dynamic prefix sits between the personal pronouns and the verb root in these cases, it is no longer in the word-initial position. Consequently, it never mutates into мэ- (ma-). Instead, it acts as a phonetic glue that often "swallows" or blends with adjacent vowels.
In Proto-Circassian, the underlying dynamic prefix *wa- (уэ-) clustered with the preceding indirect object pronouns. The ancestral combinations for "I look at him", "I look at you", and "I look at y'all" were reconstructed with the prefix sequence absolutive + oblique + dynamic prefix + root. These Proto-Circassian clusters effectively functioned as:
I look at him: *сы-й-уэ-плъ
I look at you: *сы-у-уэ-плъ
I look at y'all: *сы-шъу-уэ-плъ
As the branches diverged, these sequences evolved based on their respective dynamic prefix shifts. In West Circassian, the *wa- simplified to -э-, while in East Circassian, the *w+a contraction resulted in the rounded -о- (creating the intermediate/Eastern forms сы-й-о-плъ, сы-у-о-плъ, and сы-шъу-о-плъ).
Distinguishing the Tense Marker from the Pronoun
Because the dynamic prefix drops out in the past tense, it is crucial not to mistake phonetic pronoun contractions for the present tense dynamic marker.
For example, look closely at the East Circassian (Kabardian) past tense for "I looked at you" (сыноплъащ). The "о" here is not the dynamic prefix. The present tense dynamic prefix was dropped. Instead, the "о" in the past tense is the phonetic contraction of the 2nd person singular pronoun уэ blending with the preceding relational prefix (сы-ны-уэ-плъ-ащ).
If we compare the 2nd person singular past tense with the plural form, we can clearly see that the plural form lacks the "о", confirming it was just a pronoun contraction and not a tense marker:
Transitive Dynamic Verbs
The dynamic prefix is equally present in transitive dynamic verbs operating in the present tense. Just as with intransitive verbs, the ancestral *wa- (уэ-) sits between the personal prefixes and the verb root, undergoing reduction and blending with adjacent sounds:
1st Person Singular: с-э-тхы (I write that) < *с-уэ-тхы
2nd Person Singular: о-тхы / у-э-тхы (You write that) < *у-уэ-тхы
3rd Person Singular: е-тхы / й-э-тхы (He/she writes that) < *й-уэ-тхы < *йы-уэ-тхы
Function in Static and Dynamic Counterparts
In Circassian grammar, verbs are fundamentally divided into static verbs (expressing a fixed state of being) and dynamic verbs (expressing an action or transition). The dynamic prefix is the central morphological tool used to derive dynamic counterparts from static roots, nouns, and adjectives.
Deriving from Nouns and Adjectives
The dynamic prefix can be attached to words expressing the quality of an object (adjectives) or a profession (nouns) to create active dynamic verbs. When conjugated in the third-person present tense, these take the мэ- mutation:
From Nouns: The noun пхъашIэ (carpenter) forms the static phrase ар пхъашIэ (he is a carpenter). Adding the dynamic prefix creates ар мэпхъашIэ (he works/acts as a carpenter).
From Adjectives: The adjective бзаджэ (evil) forms the static phrase ар бзаджэ (he is evil). Adding the dynamic prefix creates ар мэбзаджэ (he is becoming evil).
Terminal Vowel Retention
The presence of the dynamic prefix also affects the phonetic structure of the verb's ending. Static verbs in the first and second-person present tense typically drop their terminal vowels (э and ы). In contrast, dynamic verbs bearing the dynamic prefix retain these final vowels.
Static: сэ сы-щыт (I stand - conveying a state of being)
Dynamic: сэ сы-щэты (I am standing / I spend time standing - conveying an action)
Transitive Static Verbs
While most transitive verbs in Circassian are inherently dynamic, the verb Iыгъын (to hold) is a unique bivalent transitive static verb. In its standard present-tense form (с-Iыгъ, "I hold"), it lacks the dynamic prefix because it expresses a state. However, it can take the dynamic prefix to form a dynamic variant, с-э-Iыгъы (derived from *с-уэ-Iыгъы), which instead conveys a habitual or ongoing action ("I hold it generally/often").
Dynamic present suffix *-r
Proto-Circassian featured a dynamic present tense suffix *-r. It is crucial to note that this suffix attached exclusively to dynamic verbs (actions or events). Stative verbs (expressing continuous states, such as "to stand") did not take this suffix in Proto-Circassian, and therefore lack any reflex of *-р in their modern paradigms. In the modern normal positive present tense, this suffix has been completely elided in the word-final position in the West Circassian (Adyghe) branch. However, the historical *-r surfaces in both branches under specific morphological conditions where it is protected by a subsequent suffix, preserved in fossilized forms, or kept as an optional marker in East Circassian.
In both West and East Circassian, the -r appears in the negative present form, preceding the negation suffix:
West: ар кӀорэп, сэ скӀорэп
East: ар кӀуэркъым, сэ скӀуэркъым
In West Circassian, the suffix is retained in participles, when conjunctions are added, and in certain lexicalized relicts:
Present participles: кӀорэр (the one who goes, absolutive case), кӀорэм (oblique case), кӀорэмкӀэ (instrumental case)
Coordination suffix (-i / "and"): кӀори (goes and...)
Lexical relicts: The suffix has fossilized in certain derived words, such as the noun and adjective стыр ("burn" or "spicy"), which derives from the verb стын ("to burn").
In East Circassian (Kabardian), the suffix is preserved in imperfect forms, interrogatives, participles, lexical relicts, adverbial constructions, and optionally in the positive present tense:
Optional present tense: Unlike Adyghe, modern Kabardian allows the suffix to optionally surface in standard dynamic verbs in the present tense: сэ сокӀуэ / сокӀуэр ("I go"), ар макӀуэ / макӀуэр ("he goes").
Imperfect: сэ сефэрт (I used to drink), сэ скӀуэрт (I used to go)
Interrogative: ар кӀуэрэ пӀэрэ? (does he go?)
Adverbial form: ар кӀуэурэ (going / as he goes)
Present participles: The suffix is consistently retained in infinite participle forms: макӀуэр ("the one who goes", absolutive case), сызэрэкӀуэр ("how I go").
Lexical relicts: Fossilized in certain substantivized participles (nouns and adjectives), such as пшътыр ("hot/hotness"), гъуыр ("dry"), and плъыр ("guard/watchman").
The following table demonstrates the reconstructed Proto-Circassian present tense paradigm using the verb "to go" alongside its modern descendants, illustrating the loss of the word-final *-r:
Negative prefix *mə-
In Proto-Circassian, verbal negation was primarily marked by the prefix *mə- (мы-). While this prefix survives in both the modern West Circassian (Adyghe) and East Circassian (Kabardian) branches, both languages have innovated by developing additional, distinct negative suffixes for standard indicative verb forms. West Circassian developed the suffix -ep (-эп), while East Circassian developed -q'əm (-къым).
The evolution of Circassian negation created a strict grammatical division between finite (declarative) and non-finite or dependent verbs. The modern negative suffixes entirely took over the standard indicative forms in declarative sentences across the present, past, and future tenses. However, the historical Proto-Circassian prefix *mə- (мы-) is strictly preserved in both modern branches under specific grammatical conditions where the new suffixes cannot be used. Namely, the original prefix remains mandatory in non-finite forms (such as participles and verbal nouns), dependent clauses (such as conditionals and concessives), and non-declarative moods (such as imperatives/prohibitives).
The following table illustrates the divergence toward suffixal negation across standard indicative tenses using the verb "to go" in the third-person singular (ar, "he/she/it"):As noted, despite this shift in standard declarative sentences, the historical prefix *mə- is retained in the following constructions:
Conditionals: Negation within a conditional clause retains the prefix.
Positive: кӀомэ (if he goes)
Negative: мыкӀомэ (if he does not go)
Concessives: Similarly, verbs taking the concessive suffix (-i, "even if") use the prefix.
Positive: кӀоми (even if he goes)
Negative: мыкӀоми (even if he does not go)
Imperatives: Negative commands (prohibitives) utilize the prefix alongside the second-person index.
Positive: КӀо! (Go!)
Negative: УмыкӀу! (Do not go!)
Participles and Verbal Nouns: Infinite verbal forms maintain prefixal negation.
Negative participle: мыкӀорэр (the one who does not go)
Past Tense in Proto-Circassian
In Proto-Circassian, the past tense was historically marked by the suffix *-гъэ (-ğă). Following the split of the proto-language, this suffix evolved differently across the modern Circassian branches. In West Circassian (Adyghe), the original Proto-Circassian suffix is strictly preserved. In East Circassian (Kabardian), however, the suffix underwent a phonetic shift, merging with the preceding vowel to form the long vowel -а.
Linguists N. F. Jakovlev and D. A. Ashkhamaf theorized that this past tense suffix did not originally emerge as a grammatical marker. Through etymological reconstruction, they traced the suffix *-гъэ back to independent primary noun roots related to time—specifically гъэ (meaning "year" or "summer") and гъуэ (meaning "time", "season", or "period"). Over time, these nominal roots grammaticalized, losing their independent lexical meanings to become the standard tense suffix for past actions across the Circassian language family.
The following table illustrates the divergence of the past tense suffix between the modern branches:
Future Tense in Proto-Circassian
The reconstruction of the future tense in Proto-Circassian (Common Circassian) presents a complex case due to significant divergence between the modern branches. Linguists identify the historical Proto-Circassian future tense as being marked by the suffix *-н (-n). This historical verbal base is highly conservative, phonetically coinciding with—and serving as the direct origin of—the Circassian infinitive (неопределенная форма глагола).
However, the affixes expressing the standard declarative future tense differ drastically between modern West Circassian (Adyghe) and East Circassian (Kabardian), to the point that they cannot be explained by simple phonetic derivation from a single proto-form. West Circassian innovated by developing the suffix -щт (historically -шьт), while East Circassian retained and adapted the historical *-н alongside the declarative suffix -щ (resulting in -нущ for Future I and -нщ for Future II).
It is important to note that the щ in the West Circassian future suffix -щт is etymologically unrelated to the East Circassian suffix -щ. In Kabardian, the -щ is a later morphological innovation appended to the end of verbs to mark affirmative declarative statements. In contrast, the Adyghe щ is an inherent part of the future tense marker itself. In some West Circassian dialects, this suffix is simplified to merely -т (e.g., кӀот), most likely dropping the fricative щ to ease pronunciation.
Despite the divergence in the primary future tense, the historical *-н suffix strictly survives in both branches' secondary future or obligative moods (Future II). Modern Adyghe uses -н for its Future II tense (meaning "must" or "will possibly"). Kabardian utilizes the suffix -нщ (from -н + affirmative -щ) for its Future II tense, while adapting it into the extended suffix -нущ for its primary Future I tense.
The following table demonstrates the divergence in future tense marking using the verb "to go" (кӀуэн) in the third-person singular (ar, "he/she/it"):A notable piece of evidence for the antiquity of the future tense suffix *-н is its fossilization into everyday vocabulary. Because this tense denoted an action that will happen or is intended to happen, it was historically used to form substantivized verbs (nouns). Several primary Circassian nouns trace their origins directly back to this future tense base:
шхын ("food" or "meal"): literally "that which will be eaten" (from the root шхы-, to eat).
тыны ("gift"): literally "that which will be given" (from the root ты-, to give).
щыгъын ("clothing"): literally "that which will be worn" (from the root щыгъы-, to wear).
Fossilized Noun Classes
While modern Circassian languages do not have a productive system of grammatical gender or noun classes, Proto-Circassian originally possessed grammatical class markers that distinguished between humans and inanimate objects. According to the comparative-historical research of linguist G. V. Rogava, ancient Circassian noun stems were built using specific class prefixes and determinative suffixes. Over centuries of language development, these class markers either fell off entirely or fused into the roots, becoming a "dead" (fossilized) morphological inventory.
A surviving relic of this Proto-Circassian grammatical class system is found in the interrogative pronouns, which strictly contrast the "human" class from the "thing" (inanimate) class. The initial с- / шъ- in the inanimate forms is a fossilized Proto-Circassian prefix specifically marking the "class of things":
Human Class: хэт (who?), хэти (anyone / everyone)
Inanimate Class: сыт / шъыд (what?), сыди / шъыди (anything / everything)
This fossilized inanimate prefix с- is also visible in other interrogative variations, such as the East Circassian (Kabardian) слIо (what?), formed by combining the prefix с- with the interrogative root -лIо. Furthermore, the colloquial variant лIо serves as a prime example of the gradual decay of this obsolete class system. Because the grammatical distinction between humans and things is no longer productive in modern Circassian, the fossilized с- prefix eroded over time in rapid speech, leaving the bare root лIо to function as an independent interrogative pronoun. While all these forms inquire about inanimate objects or concepts, лIо and слIо often carry a more conversational or rhetorical nuance compared to the standard, neutral сыт.
Labile Verbs
Proto-Circassian featured a distinct class of labile verbs—verbs that were completely neutral regarding transitivity. The exact same verbal root could be used in an ergative construction (as a transitive verb) and an absolutive construction (as an intransitive verb) without any morphological changes to the root. Older Circassian grammars (such as those by Z. Kerasheva and N. Gishev) hypothesized that all transitive verbs in Proto-Circassian were originally labile, but modern linguists like M. Kumakhov and K. Vamling argue against this, proving that many core transitives (like s'’ə-n, "to do", or hə-n, "to carry") were always strictly stable.
Kumakhov and Vamling positively identify a specific set of verbs that were truly labile in Proto-Circassian. These include roots like *z˚e / *ve ("to plough"), *le ("to paint/dye"), *c’ele ("to spread"), and *p’ət’ə ("to press"). This labile nature allowed the exact same root to function in two entirely different syntactic frameworks:
Transitive (Ergative construction): ЛIы-м щIыгу-р е-жъо (The man is ploughing the field). Here, the verb takes a direct object ("field" in the absolutive case) and the subject is marked with the ergative case (-м).
Intransitive (Absolutive construction): ЛIы-р ма-жъо (The man is ploughing / engaged in ploughing). Here, no direct object is permitted, and the subject is marked with the absolutive case (-р).
Vowel Gradation (Ablaut) for Transitivity
According to linguists N. F. Jakovlev and D. A. Ashkhamaf, the strict syntactic division between transitive and intransitive verbs did not exist in the earliest stages of the language. Originally, there was a single, undifferentiated syntactic construction. As the language evolved a need to separate "productive" (transitive/object-affecting) actions from "unproductive" (intransitive) actions, Proto-Circassian developed a system of vowel gradation (Ablaut) at the end of the verbal roots to mark this grammatically.
After the split of the proto-language, many previously labile verbs were forced into "stable" forms using this vowel alternation. The fundamental rule dictated that the base ending in the vowel «ы» (ə) came to designate a transitive action, while the exact same base ending in the vowel «э» (ă) designated a general, intransitive action. This resulted in modern Circassian verb pairs such as:
тхы-н (to write something, transitive) vs. тхэ-н (to write generally, intransitive)
ды-н (to sew something, transitive) vs. дэ-н (to sew generally, intransitive)
шхы-н (to eat something, transitive) vs. шхэ-н (to eat generally, intransitive)
The "Amorphous" Noun
Before the development of the modern Circassian case system (which uses -m for the ergative and -r for the absolutive), Proto-Circassian went through an "amorphous" stage where nouns lacked formal case affixes entirely. In this ancient period, a bare, uninflected noun was used for all grammatical roles (subject, direct object, indirect object).
As the formal case suffixes (-m and -r) developed later to mark specific syntactic roles and definiteness, the ancient uninflected noun did not disappear. This ancient "amorphous" Proto-Circassian noun survives today as the indefinite form of the noun. Modern Circassian still drops all case endings when a noun is used in a generic, non-specific, or indefinite sense, which is a direct syntactic relic of the proto-language's pre-case era:
Amorphous / Indefinite Subject: ЛIы мажэ (A man runs) — the noun лIы (man) takes no case marker.
Definite Subject: ЛIы-р мажэ (The man runs) — the noun takes the absolutive definite marker -р.
Amorphous / Indefinite Object: Фыз къуажэ кIуащ (A woman went to a village) — the destination къуажэ (village) is completely unmarked.
Definite Object: Фызы-р къуажэ-м кIуащ (The woman went to the village) — the destination takes the ergative/oblique definite marker -м.
Numbers
Verb Conjugation
The Circassian verb is polypersonal: it agrees with each of its core arguments by a personal prefix, filling up to three roles — absolutive (ABS), ergative (ERG) and oblique (OBL). Paradigms are grouped by valency (one, two or three arguments) and by whether the subject is absolutive (intransitive) or ergative (transitive).
This section reconstructs the Proto-Circassian verb and shows how Adyghe (West) and Kabardian (East) developed from it. Nearly all of the divergence comes down to three recurring developments, described first; every example gives Proto-Circassian beside Modern Adyghe, Proto-Kabardian (the reconstructed older eastern stage) and Modern Kabardian.
Shared developments from Proto-Circassian
The present dynamic: the prefix *-уо- and the suffix *-р
A dynamic verb (an action, as opposed to a static verb of position or state) marked the Proto-Circassian present twice: a prefix *-уо- before the root and a suffix *-р at the end of the word; with no personal prefix to lean on (a bare 3rd-person subject) the prefix was *мэ- instead. The past had neither. Adyghe dropped -р and fused -уо- into the preceding vowel; Kabardian kept the vowel as -о- but likewise dropped -р in the plain present (it survives only where something follows it — questions, the imperfect). Static verbs never had any of this, so their present is bare in both branches.
The three plural markers, and their loss in Modern Kabardian
The largest difference between the branches is number marking. Proto-Circassian marked a 3rd-person plural separately for each role, with three distinct markers: the absolutive plural suffix *-хэ (Adyghe -х), the ergative plural prefix *я- (vs. singular е-/и-) and the oblique plural prefix *я- (vs. е-). Adyghe keeps all three, and Proto-Kabardian still had them — but Modern Kabardian has lost every one: the verb always uses the singular form, and number rests on the free pronoun ( "they", "them"). Many Modern Kabardian forms are therefore syncretic — in each pair below the two eastern forms have fallen together while Adyghe keeps them apart.
Consonant voicing in Kabardian
Kabardian voices Proto's voiceless prefix consonants between vowels in the present — 1sg с- → з-, 1pl т- → д-, 2pl шъу- → ф-/в- — where Adyghe keeps them voiceless.
The verb classes
Monovalent intransitive: adyплъэн "to look"
One absolutive subject, one front prefix (1sg /, 2sg , 1pl /, 2pl ; 3rd person Ø). The past suffix is Adyghe -гъ, Kabardian -ащ.
Monovalent static intransitive: adyщысын "to sit"
The same single absolutive subject, minus all dynamic morphology: the present is bare, closed in Kabardian by the confirmation suffix -щ.
Bivalent intransitive: adyеплъын "to look at"
An absolutive subject plus an oblique object (3sg , 3pl ); no ergative. Kabardian loses the oblique plural ("at him" = "at them") and innovates an obligatory directional prefix in some configurations — a 1st-on-2nd object takes the translocative н-.
Bivalent static intransitive: adyтесын "to sit on"
A static verb with an oblique object from the possessive series: bare present (Kabardian -щ) and the same oblique-plural loss. In "they sit on them" both plurals stack: the oblique я- and the absolutive -хэ — Modern Kabardian drops both.
Bivalent transitive: adyлъэгъун "to see"
An ergative agent and an absolutive object. The agent alternates by tense (2sg /, 3sg /), and both plurals are visible — ergative on the agent, absolutive on the object. The Kabardian root also ablauts (present , past ), whence the dictionary form against Adyghe .
Trivalent ditransitive: adyтын "to give"
All three roles: an ergative agent, an absolutive theme and an oblique recipient. A linking -ре-/-ри-/-ра- appears between recipient and agent only when both are 3rd person. With a local theme ("I give you…") the theme prefix stands first, and the recipient still shows the singular/plural contrast that Modern Kabardian levels ("I give you to him" = "I give you to them").
Reflexives
When two arguments co-refer, the co-referential slot takes a reflexive prefix — зы- for an absolutive, зэ- for an oblique or recipient — and the verb adds the refactive suffix (West -жьы, East -ж/-ыж); everything else stays as in the plain verb.
Schleicher's fable
Schleicher's fable in Proto-Circassian:
χʷǝ č́ʷara-gjǝ
χʷǝ ja laśʷam mә q́ˤ:an
č́ʷara pǝʎ́an;
mǝ χwanǝta k:ʷǝm q:irǝ,
mǝ čʷǝχʷa čʷam,
mǝ ć̣ǝm pasa mǝš́ʷrǝ.
χʷǝ č́ʷara q̇́ˤan:
"źǝʁʷǝ sā ǵʷǝ,
q:ać̣am ć̣arǝ č́ʷara ḳ́ʷarǝ."
č́ʷara q̇́ˤan: "q:́ˤʷa χʷǝ!
źǝʁʷǝ š́a ǵʷǝ ć̣arǝ,
q:ać̣a, ł́a, č́ʷara laśʷam
ʎ́ʷa ḳ́ač̣ʷǝm čǝ-wǝ,
χʷiara-gjǝ laśʷam mә q́ˤ:a."
nǝ q:́aˤʷasa χʷǝ rǝq:ʷada q:ˤʷan.
See also
References
Урыс Хь. Ш. (2000). Адыгэбзэм и тхыдэ: ІыхьитIу зэхэлъу. Налшык: Эльбрус.
Nikolayev, Sergei L.; Starostin, Sergei A. (1994). A North Caucasian Etymological Dictionary (Edited by S. A. Starostin). Moscow: Asterisk Publishers. Preface.
Chirikba, Viacheslav A. (1996). Common West Caucasian: The Reconstruction of its Phonological System and Parts of its Lexicon and Morphology. Leiden, The Netherlands: Research School CNWS, School of Asian, African, and Amerindian Studies (CNWS Publications, Vol. 48). .
Kuipers, Aert Hendrik (1975). A Dictionary of Proto-Circassian Roots (PdR Press Publications on North Caucasian Languages 1). Lisse, Netherlands: The Peter de Ridder Press. PDF.