24.960 Syntactic Models
Instructor: David Pesetsky Lecture: T2pm-5pm (32-D461)
The course has twin goals:
First, it gives a quick introduction to at least two “frameworks” for syntactic research that compete with the Government-Binding/Principles & Parameters/Minimalist tradition in the current syntax world: HPSG and Lexical-Functional Grammar (LFG). We work speedily through much of the HPSG textbook by Sag, Wasow and Bender, and also look at the LFG textbook by Bresnan.
Next, the class turns historical, tracing the development of generative syntax from Syntactic Structures (1957) up to the early 1980s, when HPSG and LFG first separated themselves off from the research program that became GB/P&P/Minimalism. An overarching theme of the course is the issue of derivational vs. representational views of syntax — a theme that offers some surprising observations about who said what at various points in the history of the field, but also gives the course a focus relevant to the most current work.
For a demonstration that the issue is live (including the hotly debated question of whether there even is a question), you need look no further than a recent discussion on Norbert Hornstein’s blog, featuring Omer Preminger (who taught this very class in 2011). See http://facultyoflanguage.blogspot.com/2014/08/cakes-damn-cakes-and-other-baked-goods.html, which begins with links to earlier discussion on the blog that prompted that posting, and continues with millions of comments. In fact, at the right moment (about half-way through the semester), we will use this blog debate as a springboard for our own discussion.
You can get a good sense of what the class will be like from its old Stellar pages — for example http://stellar.mit.edu/S/course/24/sp09/24.960 (and http://stellar.mit.edu/S/course/24/sp11/24.960 for Omer’s version). I plan to follow essentially the same structure (with improvements in the LFG section due to Omer) — but I will work extra hard to reserve time for a topic of your choosing at the very end.
24.964 Topics in Phonology
Instructor: Edward Flemming Lecture: R9.30-12.30 (32-D461)
This course will organized around three main topics:
Phonetic grammars It has long been known that the grammars of languages must regulate relatively fine details of phonetic realization, but relatively little is known about the form of the relevant component of grammar. We will study a model based on weighted constraints (Flemming 2001), based on case studies including coarticulation (local and long distance) and the timing and realization of tones. We will also consider the relationship between phonetic and phonological grammars in light of this model.
Morphology-phonology interactions: what are the relative roles of morphology and phonology in accounting for allomorphic variation in paradigms? Morphological paradigms sometimes show complex patterns of distribution of stem realizations, in which a given phonological form of a stem appears in a morphosyntactically arbitrary set of contexts (e.g. 1pl and 2pl pres subj, 1pl pres indic, 1pl pres imp). Such cases have been taken as evidence that morphological grammars can specify complex and arbitrary mappings from morphosyntactic specification to phonological form, as proposed by Aronoff (1994). On the other hand, if these morphosyntactically arbitrary distributions of allomorphs are the result of phonological conditioning, then these powerful morphological devices are not required (cf. Steriade in press). Proponents of purely morphological analyses have often dismissed phonological alternatives based on an impoverished conception of the possibilities for phonological conditioning of stem form. We will review cases from Latin and Romance languages in light of mechanisms such as Output-Output Correspondence constraints, phonological conditioning of allomorph selection, similarity-conditioned merger etc.
Do speakers’ grammars contain phonetically-based constraints? Phonological typology has been shown to reflect a variety of phonetically-based constraints, but it remains controversial whether these constraints play a role in individual grammars, or whether they are external to grammar, applying only through processes of sound change (e.g. Blevins 2004). We will try to clarify the empirical claims that are at issue here and examine experimental evidence that bears on those claims. (This topic was also covered last year, but we will be looking at new/different sources of evidence this year).
24.979 Topics in Semantics
Instructor: Danny Fox and Roger Schwarzschild Lecture: Mondays 2-5; 32D-461
This seminar will deal with various issues in the semantics of degree expressions (scalar adjectives, comparatives, equatives, degree questions, etc.). We will begin with a well-known puzzle pertaining to the scopal interactions in which such expressions partake. We haven’t yet decided were we will go from there, but the aim is to get to Roger’s recent work reconceptualizing degrees as segments (parts of which were presented here at MIT during IAP and parts of which were presented at SALT) and some work in progress on equatives that Danny’s been doing with Luka Crnic.
21F.514/24.946 Ling Theory & Japanese Lang
Instructor: Shigeru Miyagawa Lecture: M10-1 (32-D461)
We will look at a variety of related topics centering on Japanese but also across a number of other languages. The topics mostly relate to issues of agreement, very broadly conceived:
- agreement systems, including those that don’t appear to have agreement (Japanese, Chinese, Malayalam, Mongolian, with reference to Romance)
- topic systems and root phenomena (Japanese, English, Spanish, etc.)
- binding and agreement (Japanese, Chinese, Malayalam, with reference to Basque, etc.)
- the structure of ‘why’ (Japanese, Chinese, English)
- answer fragments and sluicing (Japanese, English, German)
- position of the subject (standard Japanese, Kumamoto dialect of Japanese)
- marking of the subject as genitive (Japanese, Mongolian, Turkish, etc.)
- passive (Japanese)
24.S95: Seminar on Computation, Biology, and Language
Instructors: Robert C. Berwick & Noam Chomsky Lecture: Fridays, 11-2, 32D-461
This seminar will cover four inter-related topics: (1) recent work in linguistic theory extending ‘Problems of Projection’; (2) evolutionary biology as it relates to the origin of language, including the background results from evolutionary population biology required to understand evolutionary modeling, as well as comparative biology, genomics, and the role of natural selection; (3) computation and generative grammar, including results on the role of strong generative capacity, the computational complexity of natural language, and implemented parsers for modern minimalist generative grammars, including principles and parameters theory, derivation by phrase, and problems of projection; and (4) learnability and the poverty of the stimulus, including the classical Gold results, the role of locality constraints in learnability, and the implications of statistical approaches such as Bayesian modeling and minimum description length. No prior knowledge of computation or evolutionary biology is assumed. Syllabus and readings for the first meeting (9/5) and subsequent meetings are posted on the stellar site.