Ayāsthi (Ayāsthi: àjāsṫı), or late Thārasian vulgar Adāta (Ayāsthi: àjāsṫı ah-āmáï ah-ṫárah mìh), was the spoken language of the kingdom of Thāras approximately 800 years after the collapse of the empire of Athalē, or approximately 1,600 years after the death of Zārakātias.
Contents |
| Spoken | Written | |
|---|---|---|
| ... | Adāta | |
| 2nd century | ||
| 3rd century | Common Vulgar Adāta | Classical Adāta |
| 4th century | ||
| 5th century | Imperial Adāta | |
| 6th century | ||
| 7th century | Proto-Ayāsthi (Adhātha) | |
| 8th century | ||
| 9th century | Old Ayāsthi (Adhāsth) | |
| 10th century | ||
| 11th century | Post-Imperial Adāta | |
| 12th century | Middle Ayāsthi | |
| 13th century | ||
| 14th century | Late Ayāsthi | Neoclassical Adāta (Ayāsthi recension) |
| 15th century | ||
| 16th century | New Ayāsthi | |
| 17th century | ||
| ... | ||
The existence of Thāras (Ayāsthi: ṫárah) as an independent city-state ended with its defeat by the then-burgeoning empire of Athalē (sèriṅ ah-àtsalē, or simply sèriṅ) in 183rd year of the prophet. The unification of Rathedān (ratşejāṅ) brought increased mobility to the Dāiadak (ḋáıjaḋac) population, which in turn "smeared out" the dialectal characteristics of the spoken and written languages. In particular, the Athalē empire imposed an standard written language upon its bureaucracy in the 5th century: Imperial Adāta (àjāsṫı ah-sèriṅ). Exceptions, of course, existed: for example, the city-state of Khalanu (ċàlany) remained isolationist even after its incorporation into the empire, and its spoken language had diverged to the point of incomprehensibility even in the Proto-Ayāsthi (ēıllàjāsṫı) period. Evidence for this is found in a line in a play of the time:
The period shortly before and after the collapse of the empire in the late 10th century was a period of rapid linguistic evolution. A systematic shift in the pronunciation of vulgar Adāta plosives was noted in the early 5th century, and the first signs of something clearly not Adāta emerging in the 9th century. The following example is cited in a 11th century grammar:
The 12th century sees further development, in particular, a verbal complex began to emerge which would eventually become one of the defining characteristics of Ayāsthi. The dynamic stress of Adāta persists to this period, though already there is a correlation between stressed syllables and high pitch. The speech of the 12th century would have been recognised and understood by a 17th-century speaker of new Ayāsthi (àjāsṫı ósṫı), though he would likely comment on the odd accentuation.
In the early 15th century, the dynamic stress of Adāta was wholly replaced by pitch accent, but by this time the simple correlation between old stressed syllables and high pitch had also evolved into a somewhat more elaborate system of pitch contours over the entire word. The morphophonological phenomenon of elision also emerges in this period, with former short reduced vowels dropping out or becoming non-syllabic in various situations.
In 1289 Y.P., the printing press is invented, laying the foundation for a new age of literature. In the 14th century, scholars constructed a new written language based on classical Adāta grammar - rejecting the "corruptions" of post-imperial Adāta - but with a certain (old) Ayāsthi flavour:
However, this neoclassical Adāta (àjāsṫı ġèzoar ósṫı) was not to last: the 16th century saw the reunification of the spoken and written languages, when the bureaucracy finally officially switched to the vernacular for documentation. Yet, in practical terms this meant very little - bureaucrats certainly did not speak to each other in neoclassical Adāta. Furthermore, more often than not what was actually written down was little more than the vernacular with archaic spelling.
One of the most striking characteristics of late Ayāsthi phonology is its rich vocalism and relative poverty of consonants, in particular, its defining feature: the extreme infrequency of plosive phones. This was due to a radical sound change turning classical Adāta plosives [p t k pʰ tʰ kʰ b d ɡ] into Proto-Ayāsthi affricates and fricatives [ɸ θ x pf ts kx β ð ɣ], which remained stable, with the exception of the voiced fricatives, which turned into approximants [w j ɦ] around the late middle Ayāsthi period.
Phonemic analysis of late Ayāsthi is hindered by the fact that gaps in the distribution of certain phones do not line up nicely, making it difficult to posit a smaller phonemic inventory. One common criticism of the "reductionists" is that their analyses are too strongly motivated by diachronic considerations. This article will adopt a shallow analysis.
The consonant system is as follows:
| labial | dental | alveolar | postalveolar | palatal | velar | glottal | |
|---|---|---|---|---|---|---|---|
| nasals | m | n | (ɲ) | ||||
| plosives | p | t | k | ||||
| affricates | pf | ts | (tʃ) | kx | |||
| fricatives | ɸ · [β] | θ · [ð] | s · z | (ʃ · ʒ) | x | (h) | |
| approximants | w | l | ɹ | j | ɦ | ||
The phonemic status of the phones in round brackets in the above table was marginal: minimal pairs were generally only found adjacent to old high vowels or morpheme boundaries. In particular, [h] may be treated as a morpheme-final allophone of /s/, however, some scholars reject such analyses since "morphemic boundaries should play no part in phonemic analysis."
Phones in square brackets were used in affected pronunciations of classical Adāta words: /w z/ were usually substituted for /β ð/.
Word-final consonants may be pronounced long, though a paragogic vowel was always added after a syllable containing a long vowel. Long consonants resyllabify into geminate consonants when a vowel-initial suffix is added; but note that /zː ʒː θː/ were pronounced [sː ʃː sθ].
The vowel system is as follows:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
As with the consonants, the phonemic status of the bracketed phones is marginal. [ɨ] arose from various sources, including centralised and delabialised classical Adāta /u/, as well as reduced /i u/. [ɨː] on the other hand was almost exclusively from classical Adāta /uː/, and may be treated as an allophone of Ayāsthi /uː/ in some cases, especially after coronal consonants. [u] was a surface phone in Ayāsthi, however, it may be analysed as /wɯ/ or as an allophone of /w/; [u] was also used in affected pronunciations of classical Adāta short /u/, which is usually pronounced as /ɯ/. As for [y yː œ œː], these were all contractions of various diphthongs and are sometimes phonemicised as /u̯i u̯iː u̯e u̯eː/. Finally, regarding [əː]: this phone did not exist at all in new Ayāsthi and was only marginally present in late Ayāsthi, as a contraction of /ə.ə/.
Ayāsthi also had a full set of nasal vowel phonemes, with almost every oral vowel having a nasalised counterpart. Nasal vowels first appeared in old Ayāsthi as allophones of oral vowels before nasal consonants. Nasal vowels became phonemic when old Ayāsthi /ɴ/, itself a lenited form of classical Adāta /m n/, dropped out in middle Ayāsthi. Early in the late Ayāsthi period, (short) /m n/ in various positions disappeared, leaving behind nasalised vowels. A further development denasalised vowels in open syllables followed by nasal consonants, effectively confining nasal vowels to appear only in word-final and prevocalic positions.
Ayāsthi was mora-timed, that is, each mora had equal duration. Syllables contained exactly one vowel and had one, two or three morae depending on their structure:
For word-final syllables, slightly different rules applied:
Word-final affricates were considered heavy. There is evidence to suggest that middle Ayāsthi timing for all syllables was as described for new Ayāsthi word-final syllables, that is to say, the development of moraic codas in syllables was in the late Ayāsthi period.
A prevocalic consonant belonged to the syllable of the vowel it preceded; therefore, /elɑɹː/ had three morae, but /ellɑɹː/ had four.
Ayāsthi had a lexical pitch accent system which determines the tonal contour of each (phonological) word. The tonal contour was fully determined by the position of the primary accent. Not many words were distinguished by pitch accent alone. The precise phonological details of the pitch accent varied from dialect to dialect and from period to period; the following describes the pitch accent of standard new Ayāsthi.
Pitch accent operated on the moraic level, with classical Adāta stressed long syllables mapping to long syllables accented on the second mora. Classical Adāta stress corresponded to primary accent in new Ayāsthi, and the secondary stress which developed in middle Ayāsthi corresponded to secondary accent. Taking the natural tone to be mid, primary accent was manifested as a mora with high tone in middle Ayāsthi. The mora immediately following a primary accent or a secondary accent lowered in tone at a later stage, and eventually all mora between accents in the same word had low tone. At the same time, a final low mora was raised to mid tone if it did not immediately follow an accent. Finally, a rule developed requiring the first two morae of a word to have contrasting tones, causing some words to have a long stretch of high tone.
In words of six morae or longer, secondary accents developed on the syllable containing the fourth mora after the previous accented mora. If there was a contour within a syllable, the contour manifested on the vowel, which became especially prominent if the coda could not carry a tone (e.g. unvoiced coda consonants).
The tonal contours can be tabulated as follows:
| Accent on | 1st | 2nd | 3rd | 4th | 5th | 6th | 7th | 8th |
|---|---|---|---|---|---|---|---|---|
| 1-mora | H | |||||||
| 2-mora | H-L | L-H | ||||||
| 3-mora | H-L-M | L-H-L | L-H-H | |||||
| 4-mora | H-L-L-M | L-H-L-M | L-H-H-L | L-H-H-H | ||||
| 5-mora | H-L-L-L-M | L-H-L-L-M | L-H-H-L-M | L-H-H-H-L | L-H-H-H-H | |||
| 6-mora | H-L-L-L-M-L | L-H-L-L-L-M | L-H-H-L-L-M | L-H-H-H-L-M | L-H-H-H-H-L | L-H-H-H-H-H | ||
| 7-mora | H-L-L-L-M-L-M | L-H-L-L-L-M-L | L-H-H-L-L-L-M | L-H-H-H-L-L-M | L-H-H-H-H-L-M | L-H-H-H-H-H-L | L-H-H-H-H-H-H | |
| 8-mora | H-L-L-L-M-L-L-M | L-H-L-L-L-M-L-M | L-H-H-L-L-L-M-L | L-H-H-H-L-L-L-M | L-H-H-H-H-L-L-M | L-H-H-H-H-H-L-M | L-H-H-H-H-H-H-L | L-H-H-H-H-H-H-H |
Lexical words accented on the 4th mora or later were rare. It is noted that syllables containing accented [ə] do not occur.
For example:
| Orthography | róa | nac | zá | ra | ċā | seı | āh | ə | ’r-à | uı | tşy | şip: | əx | ’r-à | uı | ẅ | oıl | l-yṅ | ā | nàċ. | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPA | [ɹɔːnɑk] | [zɑːɹɑxɑːseɑːh] | [əɹɑwytʃɨ] | [ʃip] | [əxɹɑwyuwœllɨ̃] | [ɑːnɑx] | |||||||||||||||||||
| Accents | 2nd | 2nd, 6th | 2nd | 1st | 3rd, 7th | 3rd | |||||||||||||||||||
| Contour | L-H | L | LH-H | L | L-L | M | L-M | L | H | L | L | H | LH-(H) | H | L | L | LM-M | L | L-H | H | |||||
As the geographic area in which Ayāsthi was spoken was not very large, there was not much dialectal variation. Recorded deviations include:
Ayāsthi was written in an alphabetic derivative of the Adāta syllabary. It is transcribed as follows in this article:
Long consonants were written as geminate consonants. Apostrophes were used to mark elision, both of long consonants and of vowels. Clitics were generally written as orthographic words, except when the clitic had been elided to a single consonant, in which case the apostrophe substituted for a space. However, the possessive markers were always separated with an apostrophe only.
Where ambiguous, the second element of diphthongs will be written with a diaeresis, hence <aï> for /ɑi/ but <aı> for /æ/, <oä> for /ɔɑ/ but <oa> for /ɔ/. A note regarding <w>: it was a conventional ligature for <vu> in the native orthography, but will always be written instead of <u> for the [u] allophone of /ɯ/ in this transcription.
Nasalised vowels were written with postscript <ṁ ṅ> - initially, the choice was based on etymological considerations, but later orthographies used the rule <ṁ> after rounded vowels and <ṅ> after unrounded vowels. /ɔ̃ ɔ̃ː/ were written as if they were /õ õː/. As such, <aṅ eṅ eıṅ əṅ iṅ uṅ yṅ> but <oṁ oıṁ ūṁ uıṁ wṁ>.
In addition to the above native orthographic conventions, the following academic conventions will be adopted:
Ayāsthi remains an inflected language, but augments the inherited synthetic verbal morphology with a system of agreement clitics. Were it not for the orthographic separation of clitics, an Ayāsthi verb would have looked fantastically complex: [ɨɹˈsɑpfĩʃeːllə̃] y’r-sapfiṅ-ş’ēıll-əṅ for 1sg;nom-aff-give-aor-3;inanim-dat-3sg;obl, or I gave it to him.
On the nominal side, Ayāsthi also has proclitics, which mark case, and enclitics, which mark possession, in addition to affixes marking number. Unlike some of its neighbours, Ayāsthi has failed to develop any sort of core argument case marking on its nouns at all, though its personal pronouns retain the distinction as inherited from Ndak Ta.
The most important morphophonological process was that of elision (Ayāsthi: tàeıeṅ ah-aúr, becoming short):
Conversely, some elided final consonants may resurface, e.g. àvaṅ but àvan’aċ, àfeı but àfeıj’aċ, aṫā but aṫāh-aċ, and shortened final /ɑː/ is restored, e.g. ʒìmēa but ʒimēā’ċ.
Another pervasive alternation was that of consonant mutation (èzā’ alàuıʒell, change of consonants), which occurred at a very small number of morpheme boundaries, e.g. plural + noun and mood + verb.
The last significant alternation was that of vowel mutation (èzā’ àuıʒell, change of vowels), which occurred mainly in verbal morphology:
The declension of an Ayāsthi noun was generally quite regular, with three principal parts:
Nouns may be classified into caudal declension classes by examining patterns in the principal parts. Each noun also belonged to either the consonantal capital declension or the vocalic capital declension; within the consonantal capital declension there was a subclass of irregular nouns in /z/ which exhibited rhotacism, e.g. singular zásy, plural ārásy, a subclass in /s/ which exhibited affrication rather than voicing, e.g. singular sièıll, plural ātşièıll, and a subclass which resisted mutation, comprised of (neo)classical Adāta loans.
Number was the most central of the categories marked on the noun, attaching directly to the noun as affixes, often causing some phonological change in the stem as well, e.g. singular céllı, plural āċéllı, singulative célazy; singular ċóraṅ, plural āchóraṅ, singulative ċóranazy; singular làh, plural ālàh, singulative làsazy. Sometimes, there is instead a change in the affix, e.g. singular jásy, plural āıjásy, singulative jásizy; singular ùleıə, plural ōaẁleıə, singulative ùleıozy.
| Variants | |||
|---|---|---|---|
| Singular | The singular is unmarked. | ||
| Singulative | -zy | i.e. one of (the) ...; attached to the augmented stem, and was usually used in conjunction with the collective until the late 15th century. | |
| Plural | ā- | āı-, ōa- | Was mutually exclusive with cardinal numbers and other quantifiers. |
| Collective | ā- | āı-, ōa- | Was derivational rather than inflectional; collective nouns behaved like singulars, i.e. could take cardinal numbers or quantifiers. |
Possession (including alienable and inalienable) was marked by enclitics. Note that these were enclitic pronouns, not agreement markers - a phrase such as afeıj-aşeı aṫā fruit-poss;3;inanim forest was ungrammatical.
| Variants | |||
|---|---|---|---|
| Possessive | -’a | -’ | The possessor did not take genitive marking if the possessed was already in the possessive case. |
| 1sg | -’aı | ’-āı | i.e. my ... |
| 2sg | -’āï | i.e. your ... | |
| 3sg | -’aċ | -’ċ | i.e. his/her ...; animate referents only. |
| 1pl | -’aıc | ’-āıc | i.e. our ... |
| 2pl | -’ājoc | i.e. your ... | |
| 2pl;hon | -’ālōa | i.e. your ...; considered archaic by the early 16th century. | |
| 3pl | -’aċa | -’ċa | i.e. their ...; animate referents only. |
| 3;inanim | -’aşeı | -’şeı | i.e. its ..., their ...; number was not marked. |
Case markers in Ayāsthi were proclitics which attach to the first word of a noun phrase, whether it be a quantifier or the head noun itself. Some of these were strictly speaking derivational rather than inflectional.
| Preconsonantal | Prevocalic | Type | ||
|---|---|---|---|---|
| Direct | adverbial | Had nominative, accusative, topical functions; vocative function with the particle ō. | ||
| Genitive | ah- | ah- | adjectival | i.e. of ..., especially of (a place); partition and possession were usually indicated with the postpositional possessive markers instead. |
| Oblique | zō- zōa- | zō- zōa- | adv. / adj. | i.e. (intended) to ..., (intended) for ..., about ..., concerning ...; also had benefactive and antibenefactive functions. |
| Compositive | yṅ- | yn- | adjectival | i.e. (mass noun) of ..., made (up) of ... |
| Comitative | peṅ- | pen- | adverbial | i.e. (together) with ... |
| Instrumental | as’- aş’-, az’-, aʒ’- | ass- | adverbial | i.e. using ..., with ...; also formed adverbs from (nominalised) adjectives. |
| Comparative | neş’- nes’-, nez’-, neʒ’- | neşş- | adverbial | i.e. than ...; indicated standard of comparison. |
| Essive | y- | y- | adverbial | i.e. as ..., being ... |
| Abessive | əl- | əl- | adv. / adj. | i.e. without ..., except ... |
| Locative | ņē- | ņē- | adverbial | i.e. in(side) ... |
| Adessive | ’w- | əv- | adverbial | i.e. near ..., by ..., on ... |
| Allative | ēıl’- | ēıll- | adverbial | i.e. to(wards) ...; also had dative function. |
| Illative | ats- atş- | ats- | adverbial | i.e. into ... |
| Ablative | ġə- | ġ’- | adverbial | i.e. from ...; sometimes marked agent of a passive verb. |
| Terminative | aġērit- | aġērit- | adverbial | i.e. up to ..., until ... |
| Vialis | rapf- | rapf- | adverbial | i.e. along ..., through ..., by ... |
Pronouns were roughly equivalent to nouns in terms of syntax, though morphologically they differed from nouns in some respects; most importantly, in that personal pronouns had nominal and accusative cases. Prepositions governed the accusative case, except the genitive and possessive cases, which did not govern case, and the essive, which governed the nominative case.
| Stressed | Unstressed | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| nom sg | acc sg | nom pl | acc pl | nom sg | acc sg | nom pl | acc pl | ||
| 1sg / 1pl | ì | ìṅ | ìċ | í | y | yṅ | yċ | ī | I, we |
| 2sg / 2pl | jò | jòṁ | jòc | jó | jə | jəṅ | joċ | jō | you |
| 2;hon | láchoċsy | láchoc | láchō | Considered archaic by the early 16th century; jòċsy replaced láchoċsy as the honorific 2sg pronoun. | |||||
| 3sg / 3pl | à | àṅ | àċ | á | ə | əṅ | əċ | a | he, she, they; animate referents only. |
| 3;inanim | şèıə | āʒèıə | şeı | āʒeı | it, they, that thing (yonder), those things (yonder); formerly a distal demonstrative pronoun. | ||||
| Interrogative | ġèss | ġèzaṅ | ġèzac | ġèza | who | ||||
| Demonstrative pronouns (inanimate) | |||||||||
| Proximal | ʒéıə | āʒéıə | ʒēı | āʒēı | this thing, these things | ||||
| Medial | ġàġə | āġàġə | ġa | āġa | that thing (near you), those things (near you) | ||||
| Interrogative | ġèss | āġèss | what | ||||||
Determiners were quite unusual in the overall picture of Ayāsthi grammar, for they declined for nominative and accusative case like the personal pronouns, yet were placed after the head noun as if an adjective. Additionally, like pronouns, determiners came in stressed and unstressed forms; unstressed forms were enclitics attaching to nouns, and were more common by far in the new Ayāsthi period.
A note regarding the distal demonstrative: middle Ayāsthi and post-imperial Adāta tended to use it as a definite article, like Mavakhalan. Such usage was condemned by neoclassicists, who succeeded in eradicating its use as a definite article in late Ayāsthi. However, the use of şeıə as a generic 3rd-person inanimate pronoun rather than a distal demonstrative pronoun in late Ayāsthi was not attacked despite originating from the same semantic shift.
| Stressed | Unstressed | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| nom sg | acc sg | nom pl | acc pl | nom sg | acc sg | nom pl | acc pl | ||
| Proximal | ʒéı | ʒéıṅ | zác | zá | -ʒēı | -ʒēıṅ | -zāc | -za | this |
| Medial | ġà | ġàṅ | ġàc | ġá | -ġə | -ġəṅ | -ġac | -ġa | that (near you) |
| Distal | şì | şìṅ | şìc | şí | -şy | -şyṅ | -şic | -şī | that (yonder) |
| Interrogative | eızà | eızàṅ | eızàc | eızá | -eıss | -eızəṅ | -eızac | -eıza | which (of) |
Unlike the determiners, quantifiers retained their position before the noun. On the other hand, by analogy with quantifiers, cardinal numbers moved from postnominal position to prenominal, and as much may be considered to be quantifiers as well. On the other hand, ordinal numbers remained adjectival.
| Type | |||
|---|---|---|---|
| Paucal | fà | pl | i.e. few of ... |
| Plural | òpfə | pl | i.e. many of ... |
| Partitive | nàmm | sg / pl | i.e. some ...; took singular agreement if uncountable, plural agreement otherwise. |
| Exhaustive | èşş | sg / pl | i.e. all of ...; took singular agreement if uncountable, plural agreement otherwise. |
| Distributive | òar | sg | i.e. every ...; stacked with other quantifiers. |
Ordinal numbers were formed by prefixing ly- to the number, or the first element of the number phrase. Simple compounding was used to form the numbers from 1 to 99; higher numbers were formed by stringing together number words with the conjunction -ən- (or its contracted form -’n-), for example, jaméıllofē-ən-ceıìpfə’n-wméıllə’n-ņilìpfə’n-ēròw for 2,104,964, which was pronounced jamèıll’fē’ṅ-cèıpfə’n-wmèıllə’ņ-ņilìpfə’n-eròw in rapid speech.
| Cardinal | Ordinal | +10 | ×10 | ×100 | ×1000 | ||
|---|---|---|---|---|---|---|---|
| 0 | mà | The construct exhaustive + negative was more common. | |||||
| 1 | cè | lyċè | róċ | (ce)rò | (ceı)ìpfə | ceméıllə | |
| 2 | jà | lyjà | róıi | jarò | jaıìpfə | jaméıllə | |
| 3 | zó | lyzó | rózō | zōrò | zōıìpfə | zōméıllə | |
| 4 | ẁ | lyẁ | rów | vorò | ùipfə | wméillə | |
| 5 | jó | lyjó | rójō | jōrò | jōıìpfə | jōaméıllə | róıjō was also used instead of rójo. |
| 6 | èh | lyèh | ròleh | ērò | eʒìpfə | ēméıllə | |
| 7 | màṅ | lymàṅ | róamaṅ | māṅrò | maņìpfə | mamméıllə | |
| 8 | ġòï | lyġòï | ròlgoï | ġoïrò | ġoípfə | ġoïméıllə | The /o/ was often replaced by /œ/ in speech. |
| 9 | ņìl | lyņìl | róaņil | ņēırò | ņilìpfə | ņīméıllə | |
| 10 | (ce)rò | lyrò | roaméıllə | ||||
| 100 | (ceı)ìpfə | lìpfə | ipfiméıllə | ||||
| 1000 | méıllə | lyméıllə | méıllə by itself was a noun, and as such required the possessive marker or the use of compositive case: méıll’a ... or méıllə yṅ-... | ||||
| 1000000 | méıllofē | lyméıllofē | As with méıllə, méıllofē by itself was a noun; ceméıllofē on the other hand was a quantifier. | ||||
Ayāsthi adjectives were strictly speaking indeclinable forms, however, zero-derivation produced deadjectival nominals, which could be turned into adverbs by casting it in the instrumental or the abessive. Superlative adjectives were formed with the construct lyċè (first) + gen-adjective. For example:
However, there were some defective adjectives that resisted adverb formation and were predicated like nouns rather than like adjectives - these adjectives tended to have a "discrete" meaning. Examples included the ordinal numbers, àseıss (true), vòt (same), şìchy (separate) among others. The deadjectival forms of defective adjectives also tended to be outright nominalisations - lyċè ([the] first), vòt (truth), for example. Superlatives were also rare, and where attested tended to have idiomatic use: e.g. lyċè ah-vòt the fundamental truth. Curiously, the opposites of such adjectives were not always members of that class: for example, aṫóıssı (false), ġìseṅ (different), téw (together) all form adverbs, deadjectivals and superlatives as expected.
As previously mentioned, the spoken Ayāsthi finite verb may seem fantastically complex, as it was marked for up to eight categories:
Of these, only categories 3 to 6 were affixes inherited from Adāta. (Categories 4 to 6 were actually marked on a single affix.) Categories 1, 7 and 8 had only been obligatory since late Ayāsthi - they behaved more like clitic pronouns in middle Ayāsthi, but gradually became more like agreement affixes.
An Ayāsthi verb had four principal parts:
The citation form is, however, the reduced stem. In addition, each verb either belonged to either the vocalic or the consonantal capital conjugation. Within the vocalic capital conjugation there was a subset which caused palatalisation, and within both classes there were a subset of verbs that caused a-umlaut. As with the nouns, verbs underwent consonant mutation when mood prefixes were attached.
It is also possible to classify verbs into caudal conjugation classes by examining the patterns in the principal parts - most verbs followed a single pattern, as exemplified by nòann (to go) ← nona:
However, there were many verbs which had irregularities. For example, a large class of verbs in -àrr/-arr (← -aza) exhibited rhotacism, e.g. peılàrr (to send) ← pilaza:
Some verb stems exhibited vocalism, e.g. àï (to touch) ← ade:
The only non-finite verbal form was the participle/gerund, which was only marked for mood, voice and aspect. It was formed using the suffix -eıeṅ, which displaced any stem-final vowel and palatalised the stem consonant, except where the stem-final vowel is stressed. As a noun, it has the stem -eıen-, and possessive markers were used to indicate the subject, for example, jàʒeıen’aċa (their fighting) or nònaņeıen’aı (my having gone). The abbreviated forms -eınaı (1sg), -eınaï (2sg) and -eıeṅ’ċ (3sg) were sometimes used in rapid speech and constrained poetry. As with nouns, constructions such as -eıen’aċ máï were ungrammatical, requiring the simple possessive instead: -eıen’a máï.
Deverbal nouns, which may be regarded as infinitives, were formed by taking the bare verbal stem and treating it as a noun.
| Active | Passive | ||
|---|---|---|---|
| Habitual | [cons]-eıeṅ | [long]-leıeṅ | Caused palatalisation of stem consonant: /n ts θ s z/ → /ɲ tʃ s ʃ ʒ/. |
| Aorist | [full]-ņeıeṅ | [long]-ņeıeṅ | Caused stem vowels /o oː/ to lower to /ɔ ɔː/. |
| Imperfective | [full]-ʒeıeṅ | [long]-ʒeıeṅ |
In an unusual development, the negative prefix detached from the verb in old Ayāsthi, possibly by analogy with the affirmative particle, giving *ər aff and *əm neg. However, the consonant lenition caused by the negative prefix remained, in effect marking negation in two places. However, the initial-lenited forms were lost in middle Ayāsthi, thus completing the separation of polarity and mood, with the exception of the imperative.
| Variants | |||
|---|---|---|---|
| Affirmative | -ər- | -’r- | |
| Negative | -əm- | -əṅ-, -’ṅ-, -’m- | /m/ lenited to zero before consonants, but nasalised the preceding vowel. |
The modal system of Adāta was inherited by Ayāsthi, though its use continued to decline, with some prefixed verbs developing specialised meanings and lexicalising as verbs in their own right, e.g. zanòann (to visit a shrine or temple), ēıratsàrr (to waste time; to do unproductively), ēıràsṫ (to be unlikely to happen). The mood prefixes came in two major allomorphs, selected by the capital conjugation class, and attached directly to the verb, causing lenition of the initial stem consonant.
The antibenefactive mood, derived from the Adāta futilitive, fell out of use in middle Ayāsthi and was replaced by an indirect passive construction.
| Preconsonantal | Prevocalic | ||
|---|---|---|---|
| Indicative | Unmarked form, did not trigger lenition. | ||
| Optative | ȳ-, ū- ō-, ōa- | yċ-, wċ- oċ- | /ɨ/ occured in the affirmative, /u/ in the negative; ōa- occured when the verb was nasal-initial and triggered a-umlaut. |
| Benefactive | za- | z- ʒ- | |
| Obligative | so- soa- | s- ş- | |
| Irrealis | pū- pō-, pōa- | paz- paʒ- | pōa- occured when the verb was nasal-initial and triggered a-umlaut. |
| Imperative | ī- ēı-, ūı- | ūı- if the verb triggered u-umlaut. | |
| Prohibitive | mī- mēı-, mūı- | mūı- if the verb triggered u-umlaut. | |
Use of plural agreement to show deference to the subject of the verb is attested even in old Ayāsthi. In time, plural agreement eroded, and by late Ayāsthi the plural forms of the voice-aspect markers were only used as honorifics. It is thought that the development of obligatory subject agreement proclitics accelerated the process.
| Plain | Honorific | ||||
|---|---|---|---|---|---|
| Active | Passive | Active | Passive | ||
| Habitual | [redu] | [long]-l | [full]-tşy | [long]-tşy | |
| Aorist | [full]-n | [long]-n | [full]-w | [long]-w | The /n/ lenited to zero, nasalising the vowel, when no clitics followed. |
| Imperfective | [full]-ʒy | [long]-ʒy | [long] | [long]-ə | The /ə/ triggered a-umlaut. |
Late Ayāsthi developed obligatory subject and direct object agreement clitics, as well as dative object enclitics. It is important to note that the subject and direct object clitics for animate referents were agreement markers - it was not possible to omit them if a stressed pronoun is used or even if the referent appears in the sentence.
The /e/ in the dative clitics becomes /œ/ if it immediately follows the honorific aorist suffix /u/.
| nom sg | nom pl | acc sg | acc pl | dat sg | dat sg | ||
|---|---|---|---|---|---|---|---|
| 1sg / 1pl | y- | yċ- | -yṅ | -ī | -eıll-yṅ | -eıll-ī | |
| 2sg / 2pl | jə- | joċ- | -jəṅ | -jō | -eıl’-jəṅ | -eıl’-jō | |
| 3sg / 3pl | ə- | aċ- | -əṅ | -a | -eıll-əṅ | -eıll-a | /əː/ contracted to /ə/ in new Ayāsthi, so the 3sg;nom prefix became zero. |
| 3;inanim | Ø- | -şeı | The direct object clitic was obligatory only if a dative object clitic followed; there was no dative object clitic for 3;inanim. | ||||
The vast majority of Ayāsthi vocabulary was inherited directly from Common Vulgar Adāta, which were in turn mainly comprised of words with Ndak Ta etymology. A substantial minority of words had Gezoro etyma, including such words as mèınn (mother), ċèıarr (council) and ņèıts (bread). Fáralo loanwords were also attested, and were primarily words related to imperialism, philosophy and governance.
Some Adāta words were reimported from (neo)classical Adāta, usually with an ecclesiastical or otherwise classical semantic flavour, leading to doublets, e.g. éaḃ (husband) and éap (man), nēré (wife) and ņēré (woman), chênu (the light; good as opposed to evil) and ċêny (light), zàma (the Sun, as a deity) and zàmm (sun). Some Fáralo loanwords were also found in such doublets, most notably sèriṅ (the Empire, usually of Athalē) and şèriṅ (empire). Such reimported words exhibited lenition of voiced plosives, affrication of aspirated plosives and monophthongisation of some diphthongs, but did not have the lenition of unvoiced plosives nor the devoicing of final voiced plosives. It is likely that this was partly due to spelling-pronunciation, and partly to substitution of Ayāsthi phones for Adāta ones.
Ayāsthi employed zero-nominalisation on verbs and adjectives to form deverbal and deadjectivals, which were essentially the same as nouns. In addition, the following productive derivational morphemes are found:
The basic word order in Ayāsthi was topic verb subject object:
As for the woman, she is baking bread.
As for the bread, a/the woman baked it.
A/the woman is baking bread.
However, Ayāsthi permitted heavy adjuncts to be fronted and placed before the verb, either before or after the topic.
Adverbs were generally ordered time manner place:
I went to Nitazē and Mezaras by horse last year.
Late Ayāsthi inherited from middle Ayāsthi a somewhat elaborate system of pronominal usage, which may be summarised as follows:
| Speaker | |||
|---|---|---|---|
| Excluded | Included | ||
| Audience | Excluded | 3sg 3;inanim | 1sg |
| 3pl pl-3;inanim | 1pl = 1pl;excl 3pl = 1pl;humble | ||
| Included | 2sg 2;hon 2;hon-sg | 2pl = 1pl;incl | |
| 2pl 2;hon | |||
However, as the 2;hon pronoun (láchoc) fell out of use, the usage simplified, though one peculiarity remained: the option of using 2pl as an inclusive 1pl pronoun. New Ayāsthi usage may be summarised as follows:
| Speaker | |||
|---|---|---|---|
| Excluded | Included | ||
| Audience | Excluded | 3sg 3;inanim | 1sg |
| 3pl pl-3;inanim | 1pl | ||
| Included | 2sg 2pl-sg | 1pl 2pl = 1pl;hum;incl | |
| 2pl | |||
In addition to the personal pronouns described above, máï (person) / éap (man) / ņēré (woman) + determiner were sometimes used as deferential substitutes for 1st and 2nd person pronouns.
The noun phrase consisted of several constituents:
The case of the determiners in the possessive and genitive phrases agreed with the noun phrase.
Each preposition may only take one complement, so to construct a prepositional phrase with multiple complements, the preposition was repeated before each complement and each non-initial preposition took the proclitic ən-. For example, eıll-āmèşş ən-eıll-āmèıṅə for to the fathers and mothers. However, a possessive phrase did not repeat the enclitic -’a, so of the fathers and mothers was -’a meşş əṅ mèıṅə, not **-’a meşş ən’a mèıṅə.
The verb phrase consisted of these constituents:
Ayāsthi was a pro-drop language, so it was possible to omit even core arguments when the resulting sentence could be understood in context. As previously mentioned, Ayāsthi also permitted the fronting of any constituent in a main clause to the topic position, with the option of dropping prepositions and casting the topicalised phrase in the direct/nominative case.
Ayāsthi retained the copula of classical Adāta, though its use was restricted to making nominal predicates, and various periphrastic constructs which will be discussed later.
Some of the irregular forms of the copula lexicalised as independent verbs, resulting in the following set of regular verbs, each of which conjugated for aspect:
| Singular | Plural | |
|---|---|---|
| Present | àtş àtşe-, àtşeıeṅ | ày àï-, àjeıeṅ |
| Preterite | à à-, àeıeṅ | vè vè-, vèıeṅ |
| Future | ùċ û-, ùċeıeṅ | |
Morphosyntactically, the copulae were intransitive: they did not take object agreement markers, and the predicated noun was expressed as a subject complement in the nominative, rather than as an object in the accusative.
I am the king.
As for the king, he is I.
However, the copulae were contrasted against the verbs şì (şì-, şèıeṅ, to be) and àff (àfe-, to sit). Whereas the copulae usually expressed identity or group membership, şì and àff were used to express temporary or transient states. With animate subjects, şì had somewhat restricted usage, usually having the nuance of acting as or filling a position, while àff had more general use. àff was not typically used with inanimate subjects. However, both verbs required the predicate noun phrase to be cast in the essive (except when topicalised).
I am [currently] the first of this council [until they elect a new one to replace the previous first, who died].
The [current] first of the council is I.
Similarly, adjectives may be predicated using either the copulae or şì or àff.
He is [a] brave [man].
I am [feeling] cold.
They are hostile to me.
The active voice of Ayāsthi was fairly typical; on the other hand, the passive voice had several peculiarities which will be explored here. Most importantly, due to the topic-comment nature of Ayāsthi, the primary function of the passive voice was not to topicalise objects (patients, recipients, themes, etc.), but rather to promote experiencers (often, but not always, an oblique argument) to core argument status.
The following is an example of a sentence where the object is topicalised without the use of passive voice:
The woman, her father killed her.
Let us consider this sentence:
As for that woman, her father killed her husband.
If the speaker wanted to topicalise the husband and shift the focus to the woman, he would say:
As for that man, he was killed by the father of his wife [and she was affected by this].
There are several things to be noted here:
These properties were also present in the passive participles, however, the passive participles tended to be used as a conventional passive more often.
A finite verb phrase was nominalised by essentially replacing the finite verb complex with the participle-gerund. The subject was marked as the possessor of the gerund (i.e. y’r-àuı ↦ àwoıen’aı), and object clitics were replaced with their stressed forms. Polarity, mood and register were not marked on the gerund, though voice and aspect were.
Relative clauses were formed by extracting the head noun and its dependents from the verb phrase and placing it before the particle. A pronoun was inserted to fill the gap created if the gap so created caused ambiguity. Use of a passive participle where the relativised noun is the object was not obligatory.
I like the man who saw me.
I like the man I saw.
I like the man I saw.
The most common periphrastic idiom were those used to express tenses and perfects. Tenses were formed by combining a nominalised verb phrase with the appropriate copula. For example:
I went to the festival of Ophai.
I am going [= on the way] to the festival of Ophai [right now].
I will go to the festival of Ophai.
The formation of perfects was somewhat more complicated, with three different auxiliary verbs being used in different situations. The first type, known as the nominative perfect, used the auxiliaries şì (to be) and àff (to sit); the second type, known as the causative perfect, used the auxiliary ésṫı (to have) and expressed the causative voice. In both cases, the nominalised verb phrase immediately followed the auxiliary verb.
With the nominative perfects, the subject was redundantly marked on the auxiliary and main verbs. The situation with the causative perfect was more complicated, depending on the voice of the main verb:
In either case, the subject of the main verb was the object of the auxiliary verb.
If the main verb was in the perfective, the construct expressed the perfect of result; the auxiliary àff also carried the nuance of an experiential perfect.
I have gone to the festival of Ophai.
I have been to the festival of Ophai.
I have baked bread.
I have her bake bread.
If the main verb is in the imperfective, the construct expressed the perfect of persistent situation:
I have been waiting for you since three days ago!
He has had me waiting for him since three days ago!
Yes-no questions in Ayāsthi had exactly the same syntax as indicative sentences, but were presumably distinguished by prosody. Often questions would be reinforced by a tag such as sîra (yes) or sîma (no). Answers could be composed of just an interjection, but typically repeated the verb, or used the pro-verb ṫarr.
Similarly, other questions were formed by using the appropriate interrogatives. Interrogatives were never topicalised, and sometimes were dislocated to the end of the question.
“éılany àfeıen’aı əv-ōfáċāseı’a mèşş’aı, èşş làh ġèıl əċ-ər-vèʒy àfeıen’aċa ēıll-ìṅ as’-géċoṁ. ālàh ġèıl càzyh əċ’r-àuıṁ şìp: “mèşş’aċ ər-à seıòac pèra. ər-ésṫ’a ġòpfeıāņeıen’a ājèızaċalaha āàċāraṅ. əṅşìp ər-tàṅ y-nàċ. jàl à àfeıen’aċ əv-ōfáċāseı’a mèşş’aċ ər-àfeʒy y-jèpfə.”
“ì, àtşeıen’aı méċat’a zàma əṅ tsálo, ġòll y’r-àff əv-ōfáċāseı’a mèşş’aı, éılany nòanaņeıen’aı ats-ālàh ġèıl àfeıen’aċa ēıll-ìṅ as’-géċoṁ, y’r-zanòanaṅ ēıl’-zásy’a òpfaı. y’r-zamèızāċoan-a əṅşìp y’r-gàvafaṅ tòan’aı ēıl’-mèınn màll. y’r-àuıṁ şìp: “ó èjaċy’aı, àtşeıen’āï ċény’a āʒé, ājèızaċalah càzyh əċ-ər-vòpfōanapfanan-yṅ as’-pafàʒeıen’aċa ēıll-ìṅ y-jèpfə. əṅşìp əċ-ər-sáseṅ pàʒeıen’aċa cálah’a làhjōsin’āï, ó èjaċy’aı! īénarafaw-a á jósimmevy!”
“òpfaı ər-ràtsow leʒé’a màw’aı. ər-vòlaw-yṅ əṅşìp ər-sàpfiw vàpfoar ēıll-ìseıan’aı. y’r-pètsan-a á múċeıen’aċa ìṅ ņē-cerò lát. yċ’r-ésṫ-a pètsāņeıen’aċa. y’r-àvyjen-a āıjásī òṁ āú òṁ āġárrı, əṅşìp y’r-peılàran-a ēıl’-làh’a cáġat.”