| Home | Papers | Presentations | Classes | Tools |
In this paper I argue that linear order must be determined earlier in syntax than usually assumed. The evidence comes from closest-conjunct effects in switch-reference marking in Kĩsêdjê (Jê, Brazil).
In this paper I present a new description of switch-reference in Kĩsêdjê, based off of novel data collected in November 2014. Though this new collection of data was only meant to complement the judgments on which I had based the description published in my thesis, a few of the novel judgments turned out to contradict previous judgments. In response to this turn of events, this paper updates the description of switch-reference in Kĩsêdjê that I gave in my thesis and argues that the methodology used in this new collection overcomes issues present in previous data collections.
This paper describes the main strategies used in Kĩsêdjê to name new objects and concepts: humorous phonetic matching, phono-semantic matching, calque creation and meaning extension. Though nowadays most Kĩsêdjê are fluent in Portuguese, simple loanword adaptation, though used, is rare. The main strategies are based on phonetic matching, that is, on finding a Kĩsêdjê word or phrase that approximates the source Portuguese word. The matches that are most well-accepted among speakers are those that have humorous undertones. Those that approximate the semantics of the source word also tend to be successful. Before the Kĩsêdjê were fluent in Portuguese, meaning extension was very important, and this process is also described in this paper.
In this paper I propose a refinement to the idea that nominal possessors are syntactically analogous to subjects of transitive verbs. This idea was first put forward by Abney (1987) and has received different implementations over time. The different implementations have in common the notion that nominal possessors are external arguments of possessed nouns and they differ mostly on the specific structural treatment given to external argumenthood. Here I introduce evidence that nominal possessors are actually analogous to subjects of intransitive verbs, with possessors of different kinds of nouns being analogous to subjects of different kinds of intransitive verbs. In the languages I enlist first-hand evidence from, Bororo (Brazil, Bororoan) and Kĩsêdjê (Brazil, Jêan), intransitive verbs belong either to the unergative or to the unaccusative category, and nouns belong either to the inalienably possessed or to the alienably possessable category. As I show, possessors of inalienably possessed nouns pattern morphologically with subjects of unaccusative verbs whereas possessors of alienably possessable nouns pattern with subjects of unergative verbs. Based on this evidence and on the hypothesis that subjects of unergative verbs are external arguments whereas subjects of unaccusative verbs are internal arguments (the unaccusative hypothesis, Perlmutter 1978), I propose that possessors of inalienably possessed nouns are generated as internal arguments of the noun and possessors of alienably possessed nouns are generated as external arguments of the noun.
In this paper I question the widely accepted idea that nominal possessors are analogous to subjects of transitive verbs (“external subjects”). This idea is founded upon the existence of languages whose morphology marks nominal possessors and subjects of transitive verbs alike (Abney, 1987). Here I introduce languages whose morphology marks nominal possessors and subjects of intransitive verbs alike. Unlike subjects of transitive verbs, subjects of intransitive verbs can be generated in a variety of structural positions (Hale and Keyser, 1993). This fact creates difficulties for the determination of the structure of possessed noun phrases.
In my thesis, I investigate clausal coordination. Based on a study of the clause structure of Kĩsêdjê, I argue that the exotic construction of clause chaining is a variety of coordination. This allows me to exploit a properties of the construction, switch-reference, to propose a novel characterization of the distinction between symmetric and asymmetric clausal coordination.
This work forms part of a more general project of investigating how languages combine clauses. Switch-reference is a promising window into the mechanics of clause combining, which this work only begins to exploit. Its relation to control and its use in complementation are among the understudied topics that aren't studied in my thesis.
This paper identifies a deletion phenomenon in Kĩsêdjê and discusses it within the framework of Optimal Interleaving (Wolf 2008). In order to account for the Kĩsêdjê phenomenon, I have to formalize an aspect of Wolf's system that is left vague in his original formulation, namely, the size of the domain that is input to morphophonological evaluation. I conclude that morphophonology evaluates abstract prosodic phrases.
A wealth of work in functional-typological linguistics posits a sui generis construction called “clause chaining”, attributing various specific properties to it (see for instance Dooley (2010a,b) and the works cited therein). In this paper I argue that the special notion of clause chaining is epiphenomenal. Once a few independent language-specific properties are factored out, the construction is indistinguishable from asymmetric vP coordination
Bororo has around 700 speakers, living in five villages in the region of Rondonópolis, MT. It is the last living language of the Bororoan family (the others, according to Kaufman (1994) were Umutina, whose last speaker died recently, Otuké, and a supposed dialect called West Bororo).
This dissertation begins with a brief introduction (section 1) and a summary of the activities developed in the field (section 2). After that, it concentrates on its two main subjects: in the first part (section 3), it is a descriptive grammar of Bororo and, in the second (section 4), it proposes a theoretical model based on the data from this language. At last, it includes three appendixes: a Bororo-Portuguese lexicon (Appendix 1), the sentences elicited in the field (Appendix 2) and the stories collected and analyzed (Appendix 3).
The grammar in section 3 employs mostly the terms from the grammar tradition that comes from the Greeks. Symbolism of the generative grammar is used only in some parts that it allowed to present in a more elegant a precise way.
The model presented in the second part of this dissertation is based on the theory sketched in Chomsky (2000, 2001). Given the evidence that this theory doesn't account for the phenomena of case, agreement and movement of Bororo, some revisions are proposed in order to, essentially, rescue head movement to the syntax (head movement had been left to PF by Chomsky 2001), linking it explicatively to the phenomena of case and agreement and argument movement. Besides accounting for the data of an ergative active language as Bororo, this model is extended to other systems of case and agreement (ergative and accusative alike) and explains Holmberg Generalization.
Chomsky (2000, 2001) delineates a theory of case, agreement and movement that leaves head movement to phonological computation. I propose changes to this theory that rescue head movement to the syntax, explanatorily connecting it to the facts of case and agreement (of different ergative and accusative systems) and argument movement.
This is a description of the segmental phonology of Bororo
In this squib I present a rare laringeal restriction instanced in Bororo, namely, a restriction against two voiceless obstruents ocurring in a single root as well as across Agr+root+(negation). I provide an account within the framework of the Optimality Theory.