Pavel Caha (CASTL, University of Tromsø) will give a colloquium this Tuesday, Feb. 4 at noon in the Greenberg Room.
Semi-lexical Categories Revisited
Abstract: It is standardly assumed that an item cannot simultaneously belong to two syntactic categories at the same time. Something either is – or is not – a noun/a preposition/a numeral/… The ‘single category view’ plausibly reflects the common understanding that lexical items occupy the terminals of the syntactic tree, and that the terminals of the syntactic tree have a unique label (X or Y, but not both). However, there are reasons to believe that such a view does not do justice to all the complexities of the categorization problem.
To illustrate that, I revisit some traditional observations (going back at least to Ross’ work from the early seventies) that many items stand somewhere in between the traditional (and prototypical) linguistic categories. The particular cases I will consider correspond to Luganda/Czech analogues of examples such as: in FRONT of the car (P) vs. the FRONT of the car (N); HUNDRED cars (Num) vs. HUNDREDs of cars (N). Such items have been sometimes called ‘semi-lexical’ categories (see in particular van Riemsdijk 1998).
However, the ‘items as terminals’ view has alternatives. In a theory like Nanosyntax (Caha 2009, Starke 2009), lexical items may correspond to a whole set of terminals provided that they form a constituent. If that is so, semi-lexical categories may be understood as items whose lexical specification corresponds to a complex syntactic tree containing (by definition) several terminals with distinct category labels (e.g., [ X [ Y ]]). I argue that if this view is adopted, some curious properties of the data receive an accurate and neat explanation.
Boris Harizona (UC Santa Cruz) will give the following seminar this (Thursday Feb. 6) at noon in building 320 room 106.
On the interactions between syntax and prosody in the determination of word order
Abstract: One of the central concerns of generative syntax has been to account for word order patterns crosslinguistically. Much work since the 1970s has suggested, however, that word order might be introduced only in the mapping of unordered, hierarchical syntactic structures to phonology. I explore the consequences of this thesis in the context of cliticization in two South Slavic languages.
The starting point is a striking contrast in clitic placement between the closely related languages Bulgarian and Macedonian: Macedonian clitics can be initial within root clauses, but Bulgarian clitics cannot. In this talk I first show that the clitics in both languages form complex morphosyntactic heads with the verb; these heads are syntactically atomic in a number of ways. Second, while Macedonian object clitics have been argued to be the morphophonological reflex of syntactic agreement relations (Franks 2009), I demonstrate on the basis of various syntactic diagnostics that object clitics in Bulgarian are not agreement but instead exhibit the characteristics of nominal phrases.
I leverage this syntactic difference between the two languages to explain the non-initiality requirement in Bulgarian and its absence in Macedonian. However, syntax alone is not sufficient — it is its interaction with independently motivated principles of prosodic structuring that ultimately determines the surface order of clitics in these languages. Such sensitivity of order to prosody can be understood in a model of grammar where order is imposed on the hierarchical structures of syntax as part of the mapping to phonology.
Boris Harizanov (UC Santa Cruz) will give a colloquium on Friday February 7 at 3:30PM in the Greenberg Room. A departmental social will follow.
On the mapping from syntax to morphophonology
What are the atoms of syntax and how do they correspond to words? In this talk I address this question by documenting a certain kind of mismatch between the set of objects that syntax manipulates and morphophonological words. In particular, I provide novel empirical evidence from Bulgarian denominal adjectives that certain parts of words can behave syntactically as (non-branching) phrases. The nominal component of these denominal adjectives is syntactically active in ways expected of typical nominal phrases with respect to their thematic interpretation, anaphoric properties, and interaction with syntactic movement dependencies.
However, these denominal adjectives exhibit a number of adjectival characteristics as well. I attribute this kind of mismatch to the application of Morphological Merger (cf. Marantz 1981), an operation that is part of the mapping procedure from syntax to morphophonology. Consequently, I treat denominal adjectives as underlying nominal phrases that are converted into adjectives by Morphological Merger in the course of the derivation, as part of the word formation process which combines a nominal phrase with adjectivizing derivational morphology.
This approach results in the syntactic decomposition of morphophonological words, which leads to a syntactic treatment of at least some aspects of word formation: syntactic objects realized as parts of words and those realized as autonomous words do not necessarily differ for the purposes of syntax. The present investigation contributes to a long line of research on what have traditionally been viewed as mechanisms of syntactic word formation, such as head-to-head movement (Baker 1985, 1988) and merger under adjacency (Marantz 1981, 1988).
Luc Steels is Director of the Artificial Intelligence Laboratory of the Vrije Universiteit Brussel, head of the Sony Computer Science Laboratory in Paris and a visiting researcher at the Pompeu Fabra University in Barcelona. He will give the lecture described below on Monday, February 3 at 4 pm in Margaret Jacks Hall, Bldg. 460, Room 126. His abstract is given below.
CAN ROBOTS INVENT THEIR OWN LANGUAGE?
Abstract: For more than a decade we have been doing robotic experiments to understand how language could originate in a population of embodied agents. This has resulted in various fundamental mechanisms for the self-organisation of vocabularies, the co-evolution of words and meanings, and the emergence of grammar. It has also lead to a number of technological advances in language processing technologies, in particular a new grammar formalism called Fluid Construction Grammar, that attempts to formalise and capture insights from construction grammar, and a new scheme for doing grounded semantics on robots.
This talk gives a (very brief) overview of our approach and discusses some details of the technical spin-offs that have come out of this work. The talk is illustrated with live software demos and videos of robots playing language games.
The next installment of the Cognition & Language Workshop will feature behavioral scientist Jean-Francois Bonnefon from the Centre National de la Recherche Scientifique. He is the director of the CLLE research center in Toulouse. He works on reasoning, decision-making and moral judgment. Please join us at CSLI (Cordura 100) at 4PM for Bonnefon’s talk. Refreshments will be provided!
The pragmatics of decision-making: Experiments on face-saving connectives and quantifiers
Abstract: Because connectives (if, or) and quantifiers (some, probably) are the building blocks of judgment and reasoning, any pragmatic factor that affects their interpretation also affects a broad range of inferences and decisions. In this talk, I review experiments investigating the interpretation of connectives and quantifiers that apply to face-threatening contents such as criticisms, impositions, or bad news. I show that these interpretative effects cannot be explained by frameworks that assume speakers to use language efficiently. Rather, these effects are borne out of the tension that exists between speaking efficiently and speaking kindly.
Our very own John Rickford and Sharese King will be presenting Monday, February 10 at noon at CERAS Hall 101 on prejudice against speakers of AAVE in judicial proceedings, specifically against Rachel Jeantel in the Zimmerman trial.
Rickford and King will discuss the evidence of Jeantel’s limited literacy that emerged during the trial, and the poor reading performance of African American students at her school, Miami Norland, and in the Miami-Dade public school district, neither of which came to public attention. They ask about the extent to which speakers of African American Vernacular English and other dialects are misunderstood, disbelieved, or otherwise unfairly evaluated in courts, schools, and other settings.
Paul Kiparsky will be a keynote speaker at the conference “Theoretical Issues in Contemporary Phonology Reading Tobias Scheer”, in the thematic session ”Morpho-Syntactic Interfaces and the Architecture of Grammar”, held in Paris on February 6-8.
Vera Gribanova is an invited speaker at the Berkeley Linguistic Society parasession “Approaches to the Syntax-Phonology interface” on February 7-9.
Joke courtesy of Annalee Newitz. For more on the popular Doge internet meme and its syntactic nonconcordance, see this interesting article.