Aus MelaTAMP

Wechseln zu: Navigation, Suche


Temporal adverbials

It appears that temporal adverbials operate differently in different languages. English yesterday generally refers to the day before the day of the utterance, it cannot simply mean the day before. By contrast, Daakaka pelyen ("tomorrow") and similar expressions are interpreted as relative to the time of utterance by default, but they can easily be shifted to a different, contextually given, time.

pelyen an ka ya-t mwe vyan tee esi too, te esi or mwe luk tetes
tomorrow DEF MOD 3P-DIST REAL go look see garden DISC see bush REAL grow again
the next day when they went to see the garden, they saw that the plants had regrown. (Daakaka.3071)

Ongoing Grammaticalization: The Case of Daakaka Verbs, Auxiliaries, and TMA Markers

Sabrina Demirović, Benni Butz

Daakaka verbs can be preceded by other verbs, auxiliaries and TMA markers, among others. The first case constitutes serial verb constructions, one type of the serial predicate constructions which Daakaka makes extensive use of (including verb-verb, verb-adjective, etc). Such constructions can e.g. specify the direction of a movement (see below), but since the choice of the predicates in a serial construction is relatively unconstricted, the possibilities of semantic expression are widely varied. Auxiliaries and TMA markers, on the other hand, both respectively form a highly restricted closed class of modifiers with generally aspectual and modal semantics. Within this part of Daakaka grammar, we looked at the verb pwer "to stay, be" and its phonologically reduced forms, pwe and bwe. According to von Prince (2015) , both pwer and pwe occur as a full verb and as auxiliaries, and bwe as a phonological contraction of the realis TMA marker mwe and pwe. However, the case may be more complex than this: We found some evidence that bwe might in fact be on its way to becoming a new TMA marker, and the distribution of pwer and pwe in full verbal and auxiliary position might also not be completely arbitrary.

Serial Verb Constructions

In the following example, two verbs are serialized, one expressing direction on the vertical axis and the other specifying direction relative to the speaker.

te ya=m du tas milye mwe seling me
DISC 3p=REAL stay sit on.top REAL descend come
So they stayed up [on the hill] and he went down (Daakaka.0138)


Below, the optional auxiliary "mas" modifies the following verb, expressing its necessity.

kuli yene ka te esi puskat te ka we mas óte puskat
dog now MOD PAST see cat DISC MOD IRR must hunt cat
Today, when the dog sees the cat, it will always chase the cat. (Daakaka.0916)

TMA Markers

TMA markers occur in two kinds of phonological forms: As consonantal clitics or as monosyllabic words. They only assume the second form if they cannot cliticise to either of the words around them for morphosyntactic or phonological reasons. Compare the clitisation to the preceding morpheme

te ka ka: na=m ka ki=p tas tene nye te ki to tas
DISC say say 1s-REAL say 2p-IRR sit wait 1s DISC 2p NEG sit
He said: I told you to wait, but you didn't wait. (Daakaka.0152)

to the clitisation to the following morpheme

ye mw=i yaapu melumlum swa
3pc REAL=COP big.man quiet one
they were calm men (Daakaka.0002)

and the monosyllabic form due to lack of surrounding host morphemes:

te ya=m du tas milye mwe seling me
DISC 3p=REAL stay sit on.top REAL descend come
So they stayed up [on the hill] and he went down (Daakaka.0138)

Some TMA markers only occur as monosyllabic words.

pwer, pwe and bwe: A Case of Ongoing Grammaticalization?

The verb pwer "to stay, be" can be a part of a serial predicate construction, both in initial and non-initial position. It can, however, also occur as an auxiliary, along with its phonologically reduced form pwe. They seem to be generally interchangeble, but there might be a preference of the form pwe when occurring as an auxiliary.

a pwe kolir ane bwye ente kyun te yaapu ka we kuone nge
and CONT sing TRANS song this just DISC god MOD IRR help 3s
She was singing this song and God would help him (Daakaka.0969)

When "pwe" occurs after the monosyllabic TMA marker "mwe", they contract into the phonological form "bwe" [ᵐbʷɛ]. (Voiced plosives are prenasalized in Daakaka, hence the sequence [mʷɛ pʷɛ] is phonetically close to [ᵐbʷɛ].)

te nat-eya bwe deng
DISC child=3d.POSS REAL;CONT cry
their child was crying. (Daakaka.3016)

Since bwe has to occur in a position that is not distinguishable from the position of the TMA marker on the syntactic surface, it is possible to view bwe itself as a TMA marker. In any case, pwer is undergoing phonological and semantic bleaching, which suggests that some new grammatical formative might be taking shape. Even though this has to be further investigated on morphosyntactic grounds, there are several reasons to believe that bwe is developing some independence from pwer/pwe.

  • there seem to be no other cases of morphemes merging phonologically in the way that gives rise to bwe
  • pwer is one half of one of the many pairs of synonymous Daakaka verbs in which one (in this case pwer) can only occur with singular referents and the other with non-singular referents. It's non-singular counterpart is du, which occurs only with non-singular referents. Since they are supposed to be synonymous apart from this restriction, we would expect them to behave similarly if pwe and bwe are not really distinct morphemes from pwer. However, du seems to behave differently both phonologically and morphosyntactically: It shows no signs of being phonologically reduced or contracting with mwe; and, while there is no occurrence of mwe occurring in its non-syllabic clitic form with preverbal pwe, it does so with du quite frequently, though there is no apparent reason why pwe should exert a stronger gravity on the TMA marker than du. Compare the following two examples, where we find ye=m du but not na=m pwe; instead, na bwe is used.
ye=m du oko melipro nyoo ma ge-te kyun
3pc-REAL stay walk lax 3p REAL be.like-DEM.MD just
they just strolled leisurely like this (Daakaka.0004)
ko=m kuowilye na na bwe ka te
2s-REAL know COMP 1s CONT say DISC
You know what I'm saying? (Daakaka.0018)

Daakaka examples

A list of examples from the language in this wiki, with suggested tags, can be found here: Daakaka examples


A list of abbreviated glosses with their meanings can be found here: Daakaka glosses