Tsakxa noun phrases

Nouns in Tsakxa are comparatively lightly inflected. Some analyses treat them as being declined for both case and number; however, Dawkins (1999), who wrote the most widely-accepted Grammar of Tsakxa, opposes this by noting numerous clitic-like properties assigned to these suffixes, most predominantly their suffixing to adjectives. There are two 'classes' of noun in Tsakxa, distinguished by their inflection for plural.

Inflection for number

There are three numbers distinguished in the nouns of the Central dialect of Tsakxa: singular, dual and plural. Other dialects may or may not distinguish the dual. Dawkins (1999) posits that the current numerical inflection system derived from an older system of classifiers that collapsed into suffixes, with the exception of the class two nouns, which retained older plural forms. The plural suffixes apply in effect to noun phrases rather than nouns as a class; the final constituent of a noun phrase takes the suffix - adjectives are undeclined when used predicatively. The only case in which more than one constituent is marked for number is when a class 2 noun takes an adjective, in which case the -t is applied to the adjective and the noun takes its irregular plural as usual.

Class 1 nouns

Class 1 nouns, or 'regular nouns', form their plural simply using the suffix -t (derived, according to Dawkins, from a classifier plus tum, 'many'):

paʔe -> paʔet - 'child' -> 'children'

In the case of nouns already ending in a consonant that cannot form a cluster with /t/ (/ç/, /t/), the suffix is preceded by a vowel that assimilates to the final vowel of the singular form:

sat -> satat - 'house' -> 'houses'

Those ending in a nasal consonant assimilate the /t/, resulting in no change:

pan -> pan - 'wolf' -> 'wolves'

The dual is formed by -f (according to Dawkins, derived from a classifier plus fi 'two'):

paʔe -> paʔef - child -> two children

Through epenthesis, following a stop other than /p/,' -f is usually found as an infix before the final consonant:

sat -> saft - 'house' -> 'two houses'

The f also causes a final bilabial stop (i.e. /p/) to assimilate:

tap -> taf - 'eye' -> 'two eyes'

Class 2 nouns

Class 2 nouns, variously called 'irregular nouns', 'strong nouns' and 'broken nouns' by grammarians, undergo a vowel change in the plural (but not the dual). According to Dawkins, this was the original system for the formation of plurals, thanks to an old system of vowel harmony which caused vowels to assimilate with the vowel of an initial prefix which he posits to be e- or perhaps originally i-. They form the dual identically to Class 1 nouns. All Class 2 nouns have an initial consonant followed by /a/.

In nouns beginning with any other consonant than /t/ or /p/, the plural is formed by a simple vowel change from /a/ to /e/:

mama -> mema - 'human' -> 'humans'

In nouns beginning with /t/, in addition to the vowel change, the /t/ lenites to /s/:

tap -> sep - 'eye' -> 'eyes'

In nouns beginning with /p/, the /p/ lenites to /f/:

pasak -> fesak - 'cooking-pot' -> 'cooking-pots'

Adjectives

All adjectives distinguish only the singular (including the dual) and plural, and are formed regularly with the -t suffix in exactly the same way as class 1 nouns:

epe metqu -> epe metqut - 'massive hat' -> 'massive hats'

tap wasa -> sep wasat - 'blue eye' -> 'blue eyes'

Demonstratives

Demonstrative adjectives are suffixed to the noun phrase:

	Distance from speaker
this	that	yonder
-nxu	-ep	-e

Nxu roughly equates to 'this', referring to something close to the speaker, nxa to 'that', referring to something easily visible, and e to 'yonder', referring to something either far away or out of visual range.

Case

Tsakxa nouns are sometimes described as being inflected for case. It is perhaps more accurate to say that there are a number of postpositioned particles and true postpositions that have cliticised and can phonetically assimilate with preceding noun phrases. These are totally regular in their affixation and there is no change between Class 1 and Class 2 nouns.

Absolutive case

The absolutive case is marked by a null morpheme. Nominals in the absolutive case may be either grammatical patients (i.e. the things affected by a verb) or grammatical agents with no volition (the unwilling subject of a verb). For example:

Waspa ashewaq-ø qa-engxut-u ush-u - the tentflap blew outward towards me (tent door.flap-ABS DAT-blow-PERSONAL outward-CONJ)

In this case, the tentflap is what in English would be called the 'subject'. In Tsakxa, however, it is considered to have no volitional part in the action described in the verb, so it is treated as a patient. Likewise:

Waspa Nunu-ø unqa-utshu-nxa - Nuna died in the tent (that we are discussing)

In this case, Nunu is again the 'subject' of the verb utshu, 'to die'. Nevertheless, he is treated as a patient because the dying is not of his volition, contrasted with Waspa Nunu-kxa unqa-utshu-nxa, 'Nunu let himself die/committed suicide in that tent', where Nunu is marked with the ergative or active case.

Waspa Efan-akxa Nunu-ø unqa-utshu-nxa - Efan made Nunu commit suicide in the tent (that we are discussing)

In this last example, despite the translation of the verb being identical to the former translation with the ergative, Nunu is treated as being non-volitional again. This is the same in all cases of causativity; the 'agent' is reduced to a patient.

Waspa unqa-tusha-nxa Nunu-ø engqa-ø - In the tent (that we have been discussing), Nunu accidentally killed a fly/a fly accidentally killed Nunu

Note that the above sentence is ambiguous as to who did what to whom. This would be shown by context.

Ergative case

The ergative or active case, marked with the particle kxa, is used for volitional agents: that is, grammatical agents considered to be willingly engaged in the activity described by the verb. This is the only way to distinguish a number of distinct verbs in English: for example, Paʔ, Nunu-kxa utshu-nxa ('as for that, Nunu committed suicide') and Paʔ, Nunu utshu-nxa ('as for that, Nunu died') are only differentiated by the degree of volition exhibited by the agent.

Waspa unqa-tusha-nxa Nunu-nxa engqa-ø - In the tent (that we have been discussing), Nunu killed a fly

The above sentence shows the typical use of the ergative case (contrasted with Waspa unqa-tusha-nxa Nunu-ø engqa-ø, which implies an accidental killing). Omission of Nunu is possible and would result in a passive-like sentence: Waspa unqa-tusha-nxa engqa-ø - in the tent, a fly was killed.

Waspa unqa-tusha-nxa Nunu-ø engqa-ø Ivan-nxa - In the tent (that we have been discussing), Ivan made Nunu kill a fly/Ivan made a fly kill Nunu

In a causative situation, the causative party is promoted to the main agent and takes the -nxa. The actual agent of the verb, in this case Nunu, is demoted to the absolutive. Note that this produces ambiguity like that above which is disambiguated by context.

A sentence can only ever have a maximum of one agent.

To sum up the morphosyntactic roles of the absolutive and the ergative:

		Sentence type
		Causative	Transitive	Intransitive
Case	Causative	ergative/ergative	-	-
	Agent	absolutive/absolutive	ergative/absolutive	ergative/absolutive
	Patient	absolutive	absolutive	absolutive

Here, green represents volitional and red non-volitional arguments. As it can be seen, only agents in non-causative sentences may change case to show volition.

Focus case

The 'focus' case is more of a lack of case than an actual case unto itself. No matter what the actual grammatical relationship the focus has with the rest of the sentence, it takes no case marking; instead, the relationship is marked on the verb:

Nunu, utshu-nxa - Nunu committed suicide

Nunu, se-utshu-nxa - Nunu died

Dative case

The dative case, marked with the particle -(q)a, is used for indirect objects, for genitives and in a large number of idiomatic constructions:

Sesu-ø uʔ-a - as for your sister (sister-TOPIC 2ps-DAT)

Locative cases

The locative cases are really a highly comprehensive set of postpositions which phonologically assimilate with the preceding noun phrase.

ta: 'and', 'in company with', 'at', 'in' (a location), 'by', 'to' (with verbs of motion) assimilates to preceding nasal

fesak-ta - by the cooking pots, and the cooking pots

fesak wasat-ta - by the blue cooking pots

Mosaçan-na - 'in Moscow', 'near Moscow'

unqa: 'inside' (a hollow object), 'into' (with verbs of motion), causes preceding /t/ to assimilate to /s/, causes preceding vowel to assimilate to /u/

sas-unqa: 'in the house', 'into the house'

memu-unqa: 'inside people'

sat wasu-unqa: 'inside the blue houses'

ut: 'outside', 'near', 'out of', 'away from' (with verbs of motion), causes preceding /t/ to assimilate to /s/, causes preceding vowel to assimilate to /u/

sas-ut: 'outside the house', 'out of the house'

fa, 'beneath', 'below', 'under', 'down (into)', 'down in the' (with verbs of motion), causes preceding /p/ to assimilate to /f/

kutqu-fa: 'down into the cave', 'under the cave'

anqe, 'above', 'on', 'onto', 'up (out of)' (with verbs of motion)

wespa-anqe: above the tents, on the tents

te: 'in', 'within', 'during' (time)

fasha-te: 'in that year'

Other suffixes

Negative suffix

The 'negative suffix' acts to a degree like a case, although it can be combined with other cases and is the last suffix to be applied to a noun phrase, after plural. This takes the form -en and can be applied to both nouns and adjectives:

Paʔ, Nunu-n utshu-nxa - Nunu didn't die

Nunu-n keq-em-en - Nunu is not dead

Personal suffix

The so-called 'personal suffix' is used in a number of contexts, most commonly to imply a form of motion towards the speaker. When used with nouns, it implies inherent possession. It is used consistently with parts of the body belonging to the speaker. It takes the form of -(s)u for male speakers and -(s)a for female speakers. It precedes locative cases.

taf-u - 'my eyes', literally 'eyes towards me'

Focus-referential suffix

Whilst the 'topic' or 'focus' itself is unmarked except through word order, there is a suffix (or rather a clitic) whose purpose is ill-defined in most grammars. It appears to imply some kind of connection to the initial element of the sentence (i.e. the focus/topic). It is usually glossed as a possessive suffix, although its uses are far more wide-ranging than simply a marker of ownership. It is believed to have derived from a former reflexive pronoun or second third person pronoun that was used to discriminate in situations such as 'Anna gave Joan her (Anna's) book' and 'Anna gave Joan her (Joan's) book'. It is often used in the manner of a copula with stand-alone adjectives or noun phrases which modify the focussed element.

Mosaqan, metqu-m! - As for Moscow, it's vast! (Moscow-TOPIC vast-TOP)

Ifan, taf-em metqu - Ivan has large eyes (Ivan-TOPIC eye.PLU-TOP large)

Ifan, sesu-em ta-nxa - As for Ivan, I saw his sister (Ivan-TOPIC sister-TOP see-PERF)