George Aaron Broadwell

State University of New York at Albany

Audio | Handout

Some Aspects of Verbal Morphology in Timucua and the Gulf Languages

Timucua is an extinct language isolate, formerly spoken in Florida. Its connections to other Native American languages are unclear. The only modern discussion of Timucua is that of Granberry (1993), who suggests that Timucua has an origin as a South American creole. In this paper, I outline several aspects of the verbal morphology which are not described in Granberry (1993). I will argue that, properly appreciated, they cast doubt on the South American creole hypothesis and instead point to a relationship with Gulf, the family of indigenous Southeastern languages that includes Muskogean, Natchez, Tunica, Atakapa, and Chitimacha.

a.) There is a suffix -la which functions to mark assertions, particularly first person singular assertions, as shown in the following example. O, haba-sota-la. ‘Yes, I accept him.’ yes accept-vsuff-la I show from an examination of a corpus of Timucua text that -la is statistically the most frequent marker of first person singular. It is striking, however, that -la is the only suffix in the system of person-markers in the language; all the others are prefixes.
b.) Granberry’s (1993) grammar of Timucua describes a system of person markers prefixed to the verb stem: nihi ‘he dies’ chi-nihi ‘you die’ die you-die However, based on re-examination of the 17th century Timucua texts, I will describe a second person marking pattern in which the person marker occurs between the verb stem and a following suffix: nihi-qe ‘if he dies’ nihi-chi-qe ‘if you die’ die-if die-you-if The auxiliaries which seem to serve as hosts for person markers include auxiliary -ka, interrogative -o, and conditional -kwa/-kwe ‘if’.
c.) The Timucua corpus contains a very frequent suffix -ta which serves to link verbs sharing the same subject together, as in the following example: Biro-leqe uqua-ta pueno ni-ca-la. male-nsuff take-ss come 1-aux-la ‘I bring males.’ (Lit. ‘I take males and come’) Examination of the corpus shows that a number of the most frequent verbs show shortened forms before this suffix. Thus mana ‘think, want’ and faye ‘go’ have the truncated forms man-ta ‘thinking, wanting’ and fa-ta ‘going’, in which a final vowel or syllable is deleted before this suffix.

All three of these aspects of the verbal morphology show similarities to verbal morphology in Gulf languages. In particular, I will argue the following:

a.) Muskogean, Tunica, and Atakapa all show person marking systems in which the first person singular is the only suffix in a system which is otherwise composed entirely of prefixes. This is an unusual pattern and not likely to have arisen independently in Timucua.
b.) Alternation between agreement on the main verb and agreement on an auxiliary is also found in Tunica, Natchez, and Proto-Muskogean (Haas 1946). Furthermore, at least one of the Timucua auxiliaries is identical to that reconstructed for Proto-Muskogean.
c.) Timucua -ta is strikingly similar to the Proto-Muskogean *-t ‘same-subject’ morpheme which links verbs together. Just as in Timucua, Proto-Muskogean must be reconstructed with a pattern in which a number of the most frequent verbs show truncated stems before *-t. Taken together, these three aspects of the verbal morphology show similarities to the morphology of the Gulf languages that are strongly suggestive of a genetic relationship between Timucua and Gulf. This evidence thus support the proposals of Swanton (1929), Haas (1951), and Crawford (1988) for genetic links between Timucua and Gulf, while casting doubt on a South American connection.