(= special topics courses)
UCLA Linguistics Department
Non-proseminar graduate courses are listed here too whenever a new course description is issued.
Archive: [Winter 2008][Fall 2007][Spring 2007][Winter 2007][Fall 2006][Spring 2006][Fall 2005][Winter 2005][Fall 2004][Spring 2004][Winter 2004][Fall 2003][Spring 2003][Winter 2003][Fall 2002][Spring 2002][Winter 2002][Spring 2001][Winter 2001][Fall 2000][ Spring 2005]
Spring 2008
Computational Linguistics II (LING
185B)
Marcus Kracht
Spring 2008
Time: Monday and Wednesday 9 - 11 am
Place: Bunche 2150
Course ID: 253784200
Short Description of the Course
In this course we shall look at semantics of natural language expressions in a very concrete way. We shall construct a finite model M and then translate sentence into logical formula which can be evaluated in M. Possible candidates are:
1. Boolean expressions (and, or, not)
2. Modal/temporal expressions (will, possible)
3. Quantifiers (every, some)
We may also consider different ways to interpreted a formula (static versus dynamic).
The idea is stay concrete: models are finite, and manipulations are done on elements of the model. This allows to use a computer to do the bookkeeping for us. Moreover, finite structures guarantee (at least in principle) that an answer to a query can be found.
On the way we shall discover how the different types that we find in OCaml are reflected as different objects in the model. A first-order structure, for example, sees the meaning of a relational symbol as a set of tuples. A functional model sees them as higher order functions. While we would like to think of them as being the same, they are not. And so must find ways to mediate between these two.
LING 225, Language Topics: Zapotec
Syntax
Felicia Lee
MW 2:00-3:50
Public Affairs 1278
The Zapotec languages of southern
Mexico show a range of syntactic patterns that provide a rich testing ground
for current syntactic theories. Among their features are a large variety of
left-periphery phenomena (overt focus and topic movement, different question
particles with different syntactic constraints and pragmatic usages), complex verbal morphology and verb
movement behavior, and anomalous binding
and coreference patterns.
The course will begin with a
descriptive and comparative overview of the Zapotec languages, then move on to
detailed theoretical and descriptive examination of selected syntactic
constructions and morphological phenomena. As a point of departure, we will
focus primarily on a single language of the family (San Lucas Quiaviní (a.k.a.
Tlacolula Valley) Zapotec, spoken in central Oaxaca); we will also examine data
from other Zapotec languages.
Course requirements will include a
term paper, an in-class presentation of a selected reading, and a few short
assignments.
Prerequisites: LING 120B required;
LING 200B highly recommended. A basic reading knowledge of Spanish will be
helpful, but not essential.
Linguistics 251: Bantu Tonology
Michael Marlo
The Bantu languages are a huge family of 400-500 languages, spanning
virtually the whole of sub-Saharan
The study of tone in Bantu languages has made many important contributions to theoretical linguistics. In phonology proper, the theory of autosegmental phonology was developed initially to account for Bantu tone systems and later expanded as a general theory of phonological representations. More recently, a non-autosegmental theory of domain-based representation called Optimal Domains Theory was developed based on Bantu tone data; the main insights of this work have been implemented in John McCarthy's Span Theory. Throughout the years, data from Bantu tone languages have also been brought to bear on issues of locality, the relationship between tone and metrical structure, and the role of constraints like the OCP in phonological theory. However, over the past 10-15 years (since about the time three important Ph.D. theses were written at UCLA on complex Bantu tone systems), Bantu tonology seems to have been only marginally important in phonological theorizing in Optimality Theory, the dominant theoretical paradigm since the early to mid 1990s. The likely cause of this is the fact that the "classic" studies of Bantu tones systems involved multiple rule interactions and opacity effects, which are notoriously difficult to analyze in surface-oriented OT. While Bantu tone has not been discussed much, the problem of opacity has not gone away in OT, and there have been recent proposals by leading phonologists like McCarthy to re-introduce derivations into the theory. Perhaps we should see this state-of-affairs as a challenge and an opportunity.
Bantu tone languages have also played a central role in studies of the architecture of grammar and the interaction of grammatical components. Descriptions of Bantu tone systems (as well as patterns of vowel lengthening and shortening) have occupied a prominent position in the literature on phonology-syntax interactions. In particular, they have been brought to bear on the debate between "indirect reference" and "direct reference", i.e., to what extent phonological rules and constraints can refer to syntactic structure, and they have contributed greatly to the development of the theory of Prosodic Phonology. In recent years, such data have been used by some researchers to argue against the prosodic hierarchy, in favor of a phase-based theory of spell-out, following developments in syntactic theory. An understudied aspect of tone in many Bantu languages is that tonal patterns in verbs are often controlled by the syntactic structure in which the verb is realized. For example, keeping all other factors constant, a given verb may be realized with a different tonal pattern in matrix clauses, in clauses in which a subject DP is extracted, and in clauses in which an object DP is extracted.
The study of Bantu tone systems also bears on theories of phonology-morphology interactions and has played an important empirical role in evaluating prominent theories. For example, phrase-level opacity effects reported in some Bantu languages should not exist given the model of grammar espoused in the theory of Lexical Phonology and Morphology and its more recent incarnations as Stratal OT. In-depth studies of Bantu tone systems reveal that tonal processes often display a high degree of sensitivity to morphological categories and domains. For example, the location of grammatical tonal suffixes is controlled in many Bantu languages by tense-aspect-mood-negation distinctions, such that in one set of tenses a H may be located on the stem-initial mora, in another set of tenses a H is located on the second mora of the stem, in another set of tenses a H is located on the third mora of the stem, and in another set of tenses a H is located on the fourth mora of the stem (this is Kikuria, Chacha's language). In other languages, tonal processes may show sensitivity to particular morpheme types, such that, for example, Hs contributed by object prefixes, the reflexive prefix, the verb root, and inflectional suffixes are treated differently by tonal rules. In some Bantu languages, the tonal patterns of nouns and verbs operate by the same principles; in other languages, the two lexical categories require an entirely different analysis.
Given this introduction, Bantu tone is therefore a topic that is rich in data that bear not only on phonological theory but also on theories of grammar and component interaction. I hope that students and faculty with different theoretical interests will participate in this seminar. We will read detailed studies of the tonal systems of a number of Bantu languages and discuss implications for phonological theory and theories of phonology-morphology and phonology-syntax interaction. Students will choose a language to investigate, giving a presentation of the data and developing a re-analysis of (a portion of) the data, which will be written up as a term paper. Perhaps from this seminar we too can author a paper or two that pushes linguistic theory.
LING 254A SEM 2
"Locatives", Spring 2008
Marcus Kracht
Time and Place: Tuesday & Thursday 14-15:50, Rolfe 3123
Course ID: 65340602
Description:
Locative expressions are abundant in language, yet they have not been as much studied as, for example, tense. Recently, one notices a growing interest in space and spatial expressions (there is a new series "Explorations in Language and Space" by Oxford University Press and various books among other by Robert Levinson). The proseminar is devoted to the study of locatives. I propose to cover (depending on time) the following topics:
☛ Morphological systems (locative cases in various languages, especially Uralic and Caucasian languages, locative PPs in Indo-European)
☛ Semantics of spatial expressions: Jackendoff has proposed a layered structure, one element defining the spatial region, the other a path. We first study the subtleties of specifying the region.
+ axes and frames (Levinson): some
special properties of Australian indigeneous languages, Inuit, and Oceanic
languages.
+ factors determining the use of Ps
(Tversky, Landau)
+ modes and directionality (special
attention to Uralic languages, in particular Finnish (V. Fong))
+ orientation (Nam, Kracht)
☛ Interactions between syntax and semantics:
+are local cases structural or
semantic? (Vainikka vs Niikanne)
+case selection and the bimorphemic
analysis (Kracht)
☛ Metaphorical uses of spatial expressions
☛ fictitious motion, ception (Talmy, Givon)
☛ aspects of cognitive grammar (Langacker)
☛ Historical development: how spatial expressions "radiate" into other domains (tense, possession, etc)
Prerequisites:
No special knowledge besides basic linguistic training is needed.The mathematics and formal semantics which is involved is fairly basic.
LING 254A
Topics in Linguistics
ACQUISITION OF A-CHAINS AND
UNACCUSATIVITY
Nina Hyams
Children generally appear not to produce or understand verbal passives and raising under verbs like seem and appear until ages 5-7 (though there are claims that this is not so for passives in languages such as Sesotho, Quiché, and Inuktitut). There are various accounts of this late development, the most prominent of which is the Maturation of A-Chain Hypothesis (Borer & Wexler 1987, 1992). Other proposals include the External Argument Requirement Hypothesis (Babyonyshev et al. 2001); the Universal Phase Requirement (Wexler 2004); the Universal Freezing Hypothesis (Hyams and Snyder 2005); the Canonical Alignment Hypothesis (Hyams et al. 2004); and the Theta Transmission Hypothesis (Fox & Grodzinsky 1998; Okabe & Sano (2002). On the other hand, there is evidence from Romance auxiliary selection and other sources that children have unaccusative structures, requiring A-movement, by age 2-3 (Snyder et al. 1995; Hyams & Snyder 2005, Lorusso 2004, Friedmann 2007, a.o.).
The goals of this class are three-fold:
(i) to review the early and recent literature in this area (and related syntax),
(ii) to try and resolve the apparent conflicting findings, and
(iii) to investigate other sources of data to evaluate the various approaches.
My hope is that we will be able to uncover other A-chain or unaccusative phenomena (especially in languages other than those most typically looked at) which can be examined in acquisition.
Course requirements include class presentations of readings and a term project.
Note: The class is currently scheduled for Mon. Wed. 11-1, but this conflicts with Syntax 3. I would like to find a 3-hour slot once a week that is convenient for everyone. If you are interested in attending, please let me know what your availability is.
Linguistics 252, Peter Hallman
Topic: Aspect and (In)definiteness at the Syntax-Semantics Interface
A predicate's aspectual type commonly affects the definiteness or specificity of NPs related to that predicate, as in 'definiteness effect' contexts or interactions of the Finnish kind where Case is involved (for one verb class, object Case marks definiteness, for another it marks predicate aspect). The proper analysis of these phenomena is bound up with the question of whether definiteness and specificity, and for that matter Case, are 'scope' phenomena, i.e. correlated with syntactic configurations, for this tells us how syntactic the interactions with aspect are. Further, the question of just exactly what we mean by 'definiteness' and 'specificity' and whether these terms can be defined language-independently is an important preliminary question. In this seminar we investigate what definiteness and specificity are, whether they can be appropriately characterized as scope phenomena, and then attempt to address the question of what explains their mysterious attraction to aspectual alternations. Requirements: readings and a term paper.
We will have
an organizational meeting as scheduled on Monday, Jan. 7 at 2:00 p.m. in Rolfe
3114. The course may need to be
subsequently rescheduled so if you're interested in attending but have a
schedule conflict let me know and we'll take you into account.
Linguistics 251: Metrics
Bruce Hayes
Metrics studies how conventionalized rhythmic patterns are manifested in phonological material. Metrical systems can vary greatly across languages, including systems based on stress, on syllable weight, on the distribution of word boundaries, on tone, and on various mixtures thereof. They can apply to song, chant, and spoken poetry.
Metrics has long served as a kind of laboratory for phonology: by studying the application of phonological structure to rhythm, we obtain independent evidence for the nature of phonological structure. In addition, the fairly limited, controlled character of a metrical system makes it a useful testing ground for proposals about linguistic theory in general, notably the character of constraints and their interaction.
This course will include a mixture of:
-- Empirical
study (a survey of metrical systems of different types, with particular
emphasis on English)
-- Theory of metrics (metrical patterns, metrical constraints, how constraints
interact, grammatical components)
-- Learnability theory (application of recent tools, including stochastic OT
and Maximum Entropy). The goal here is to employ simple, general theories of
constraints and constraint schemata, and let hardwork inductive learning
algorithms deploy them to account for the richness of the data patterns.
Anticipated assignments for 4-unit students: readings, a few exercises, and an paper
Linguistics 251: Development of speech perception
Megha Sundara
In any organism’s interaction with its environment, the ability to sort input into categories or kinds, is fundamental to survival. Without the ability to categorize, every instance is unique, and consequently, unrelated to the organism’s previous experience and knowledge. In this proseminar, we will try to understand how learners of different ages and language experience acquire phonetic categories that are fundamental to language. We will be discussing the perception of phonetic categories – methods for testing, development trajectories, mechanisms, and its role in typical and atypical language acquisition.
In addition to participating in class discussions and attending presentations (by other students and by me), students will be expected to read primary sources. In the reading list (a) classic articles are interspersed with recent research (b) using several different methodologies – perception testing, training studies of adults, testing of infants (c) with different kinds of learners – monolinguals, sequential bilinguals, simultaneous bilinguals, international adoptees, language impaired children and monkeys.
The proseminar will be held Mondays and Wednesdays, 9:00am-10:50am in Rolfe 3127.
Linguistics 218, Mathematical Linguistics II
Ed Keenan
Time: TTh 2:00 B 4:00
Ling 218 is a variable content course; it presupposes the content of Ling 180/208.
The focus of this offering is the mathematical characterization of grammatical categories (gc=s) in natural language. The first four weeks treat algebraic semantic properties of gc=s.
Weeks 1-3 Expressions in Argument Categories denote in freely generated boolean lattices (algebras). The free generators are the individuals. This works well for DP. What about CP? What about infinitival arguments (e.g. To talk with your mouth full is impolite)? Modifier categories (modifying adjectives, intensifiers, manner adverbs) denote in factor algebras (defined in class). Quantifiers denote in algebras of conservative functions. Indexing categories, such as locative PPs (John is smiling in Ben=s picture) and temporal deictics (John only smoked once yesterday) lack an algebraic characterization at the moment.
Week 4
Generalized quantifiers: Why do they arise in natural language? Why do
they exhibit scope ambiguities, B
an emergent property, contrary to the usual pattern (Borer In Name
Only, 2005
Readings for weeks 1-4 class handouts, an overview paper on Quantifiers for the Handbook of Semantics, and a version of Beyond the Frege Boundary (by yours truly).
Weeks 4- 8: abstract syntax, specifically syntactic properties of gc's. We use Bare Grammar (Keenan & Stabler, 2003) as a partial text, plus class handouts and selections from Baker's Lexical Categories 2003 CUP. We consider several candidates for universal properties of categories B Syntactic operations may change category, but must respect sameness of category, AIs recursion a primitive (as many linguists from Chomsky on down claim), or is it a consequence of quantity? What basis is there for saying that expressions in different languages have the same category? Is the set of possible human grammars closed under duals? How can we generalize across languages which are not isomorphic? Our answer to these questions builds on the notion of linguistic invariant. They are the expressions, properties of expressions, relations between expressions,... which are invariant under the structure preserving maps (automorphisms) of the language. Function words are the lexical expressions mapped to themselves by the automorphisms. An invariant property might be has category C for C some gc in the grammar. Some relations, such as c-command are universally invariant from their definition; others, such as the Anaphor-Antecedent relation, appear to be invariant empirically even though we can make up formal grammars in which they are not. A basic question here is Which semantic relations are syntactically invariant (coded in the grammar) and which ones aren't? Arguably the Anaphor-Antecedent relation is invariant, the entailment relation is not. A more general algebraic way to put some of these questions is: How much can we tell about a language from the automorphism group of its grammar? (So we cram in a little group theory in this second half).
The last two weeks are open, to allow us time to pursue matters of interest that arise.
Work required:
2 units: do the erratically assigned homework problems.
4 units: homework plus a paper.
Sun-Ah Jun: Fall Proseminar (251A): Topics in Phonetics & Phonology
Topic: Focus Prosody
Organizational meeting: Thursday 2pm at Haines A20
The course will cover readings related to the prosodic realizations of focus, emphasizing cross-linguistic data. Students will acquire the methodological and analytical skills to read the literature and to conduct original research on this topic.
Linguistics 252, Philippe Schlenker
Topic: Presupposition
In the 1980's, the analysis of presupposition contributed to a 'dynamic turn' in formal semantics: the concept of 'meaning as truth conditions' was replaced with a more powerful notion, whereby the meaning of a sentence was identified with its potential to modify the beliefs of the speech act participants. This seminar has two goals:
(i) We will introduce the motivations, formal tools and empirical achievements of the dynamic analysis of presuppositions.
(ii) We will discuss critiques of the dynamic approach and ongoing attempts to offer a systematic alternative to it.
(Time permitting, neighboring topics such as anaphora and implicatures might be touched upon.)
Requirements
(i) Do the
readings and participate actively in class discussions.
(ii) Give one class presentation and/or do some exercises.
(iii) Write one term paper.
Preliminary Website
http://www.linguistics.ucla.edu/people/schlenker/presupposition-ucla-07.html
Linguistics 254,
Fall Syntax Proseminar
Topic: Adverbial PPs and Light Prepositions
Tentative Time & Place: Thursdays 11-2 in Rolfe 3106
(We will meet at that time in that place this Thursday Sept 27 to discuss organizational matters and people's specific interests regarding the content of the prosem. If you would like to attend the course but cannot make it to that meeting please email me beforehand.)
Content:
While the fine semantics and syntax of adverbs has received considerable attention recently, thanks in part to the work of Cinque (1999), adverbial prepositional phrases have inspired less work. We will focus on a subset of these that raise particularly interesting questions for interpretation and syntactic licensing, namely those introduced by the light prepositions {in, on, at, to, for}. These have in common the fact that they can be missing/unpronounced in circumstances when they seem to be semantically present, e.g.,
He stayed
here (for) three hours.
I've lived (in) all the places she's lived (in).
Nothing special happened (on) that day.
I was told they could arrive (at) any moment.
He drove us (to) someplace where we had never been before.
In some cases, particularly with temporals, an overt light preposition may be impossible:
She will arrive (*in/*on/*at) next week.
An optional light preposition may become obligatory if the PP is moved:
John has been
a Republican (for) many years now.
John has *(for) many years now been a Republican.
Such facts can lead to a number of lines of inquiry, among which we will choose based on the interests of the attendees (and their linguistic competence--surprisingly, at least among Familiar European Languages, patterns like this seem to be widespread). Issues could include:
- Can we resolve the debate between McCawley/Emonds vs. Larson on whether the absence of an overt preposition indicates an inherently case-marked NP versus a PP with a silent head? Should the same answer be given for English and for a rich case language like Finnish?
- If we are dealing with silent prepositions, can their distribution be related to other phenomena of "silence" for which theoretical apparatus has been developed, e.g. null complementizers, gapping, empty categories,...?
- How can we characterize the set of nouns that can be relativized without an overt preposition (or wh-phrase that arguably contains one), cf.
the time
(when) he came to see me vs.
the occasion *(when/on which) he came to see me
- How can we explain why certain kinds of D(P)s license "omission" of prepositions more readily than others, cf.
I'll meet my love {(on) some sunny day/ *(on) a sunny day}.
ETC.
Requirements: [modified slightly from Philippe's]
(i) Do the
readings and participate actively in class discussions.
(ii) Give one class presentation.
(iii) Write one term paper.
Linguistics 218: Mathematical Linguistics II
Spring 2007, TR 9 - 10:50, Rolfe 3123 (tentative)
Marcus Kracht
I shall devote the course to the study of compositionality. All linguists agree that a compositional theory is better than a noncompositional one. But which theory can legitimately claim to be compositional and why? I guess that few people can really answer this question satisfactorily. Moreover, the formal literature is sometimes at war with ordinary intuitions. But I think substantial progress has been made, and I want to talk about that.
There is a manuscript of a lecture I held last year, which you can find at
http://kracht.humnet.ucla.edu/marcus/html/cmp.pdf
Linguistics 225: Writing a Teaching Grammar
Spring 2007, MW 9-11
Pam Munro
This course will be a practically oriented course dedicated to producing a draft of a teaching/reference grammar aimed at engaged high school level non-linguists while beginning development of other enrichment and teaching materials.
The language we'll work with is Pima, a Uto-Aztecan language spoken on two
reservations in central
It's an endangered language: although there are still some fairly young speakers, we don't know of any children acquiring the language.
The goal of this course is to produce a basic teaching grammar that can be
used as a resource for students in UCLA Linguistics 114 and by heritage
learners of Pima in
Hopefully we'll also have the assistance of Marcus Smith, C.Phil. and Pimologist extraordinaire.
If you take this course, you'll
• acquire firsthand experience in producing accessible documentation for an endangered language • learn a great deal about a fabulous language • have the satisfaction of helping to create a product that will be used seriously both by heritage learners and by UCLA undergraduates
No prior field experience is required. Please email any questions to munro@ucla.edu.
Locatives
Marcus Kracht
Topics in Syntax and Semantics (Linguistics 252B)
Tuesdays and Thursday, 2 - 4 pm, Public Policy 2319
This seminar will be dedicated to the study of locatives. We shall look at
(a) the semantics of local expressions: I will both discuss model-theoretic semantics (due to Zwarts and myself) as cognitive semantics (Talmy)
(b) the syntax and morphology (cases, PP structure)
(c) the historical development: where do local expressions come from and into what do they develop
I have given this course before and there are course notes on my website (see http://kracht.humnet.ucla.edu/marcus/courses/locatives/locatives.html). However, the course notes do not reflect all the content and are very rough. I hope that I can take the time and rewrite at least parts of them. I also point out that Peter Svenonius will be giving a minicourse on adpositions (mainly devoted to spatial adpositions), which will nicely complement the seminar.
Proseminar (Linguistics 252): Three Minicourses
There will be a series of mini courses in syntax this quarter regrouped under a 252 proseminar. This proseminar will be listed under Dominique's name on the registrar's site.
It will meet according to the following schedule:
Mondays 5-7pm starting Monday the 15th of January until the end of February
Fridays 2-4pm in February
Another two hour slot weekly in February to be determined later.
Doubling in the Left Periphery in West Germanic Dialects
Martin Prinzhorn
From 1/15 to 2/ 2 (4meetings)
In this mini-course, we will at two doubling phenomena, doubly-filled Comp and complementizer agreement. The two phenomena are usually described as being optional in the variants they occur in. I will concentrate on this question of optionality.
In the history of generative grammar, many optional phenomena turned out not to be optional after all. By taking a closer look at doubly-filled Comp and complementizer agreement in various Germanic dialects, it will be shown that there are indeed external licencing conditions for both constructions having to do with different types of relative clauses, scope of pronouns etc.
Another question which we may touch upon, and provide a partial answer to,
is why both phenomena seem to be restricted to verb second languages (verb
final in embedded clauses).
David Adger
Language Variability and Linguistic Theory
From 2/5 to 2/16 (5 meetings)
These lectures look at the question of whether there is variability (a range of forms with the same meaning which provide a choice of expressions to a speaker) in I-Language, and if so how it is best modeled, and what challenges modeling it raises for Linguistic Theory. It takes a number of case studies, mainly from one dialect of Scottish English, and uses these to explore the theoretical question and to provide some tentative answers.
Good background readings can be found in Cornips and Corrigan, 2005, Syntax
and Variation. I'll spend some of the time going through Adger, 2006,
'Combinatorial Variation' Journal of Linguistics. For background on Labovian
approaches, Tagliamonte 2006, Analysing Sociolinguistic Variation, CUP is good,
and Alison Henry's paper in the Blackwells Handbook of Language Variation and
Change is also worth reading.
Adpositions
Between 2/19 and 3/2 (5 meetings)
The course is on the structure of adpositions cross-linguistically, and the relationship of that structure to the compositional meaning of expressions of location and directed motion. The focus will be on spatial adpositions and equivalent expressions, such as local cases in languages like Finnish and Lezgian. The results have important implications for the nature of categories in syntax and for our understanding of the syntax-semantics interface in general.
The course draws on the results so far of the Moving Right Along project, a five-year project (2005-2009) investigating these issues. See http://www.hum.uit.no/mra/ for more information.
Suggested background readings:
Proseminar: The Prosodic Word (Ling. 251A/B, Special Topics in Phonetics & Phonology)
TR 11:00-12:50
Kie Zuraw
INTRODUCTION
When we describe prosody, phonological alternations, or phonotactic restrictions, we must specify the domain of application of the rules or constraints involved. For example, saying that a nasal assimilates in place to a following obstruent is insufficient: do the nasal and obstruent have to be in the same word? if they’re in adjacent words, does the syntactic relation between the two matter? does it matter whether a pause intervenes?
This question of domains has been approached in a variety of ways. Most common these days is to use a prosodic hierarchy (Selkirk 1980, ...). The grammar assigns to an utterance a prosodic tree, with nodes such as intonational phrase, phonological phrase, p-word, foot, and syllable; rules or constraints are sensitive to this prosodic structure.
Current literature tends to draw freely on prosodic structure, without explicit comparison to other approaches. The purpose of this proseminar is to examine the evidence for one level of prosodic structure, the p-word (short for “prosodic word” or “phonological word”).
WHY THE P-WORD?
The p-word is roughly a syntactic word, but with various language-specific modifications: a function word may be combined with an adjacent content word, prefixes (and, less often, suffixes) may be excluded from their stem’s p-word, and a compound may include more than one p-word.
The p-word has been proposed to do a variety of jobs:
• minimal domain for stress assignment
• domain for certain segmental rules/constraints
• (less commonly) domain for phonotactic restrictions
• unit of prosodic morphology
I’ve chosen to examine the p-word, rather than some other level of structure, for several reasons:
• Unlike the syllable and the foot, the p-word hasn’t been
subjected to much critical examination.
• More than the syllable or the foot, the p-word is dependent on
morphological/syntactic structure.
• Unlike utterances, intonational phrases, or phonological phrases,
p-words lack an intonational signature.
• More than utterances, intonational phrases, or phonological
phrases,p-words are likely to be precompiled (because they are small enough
that many are frequent), which raises some psycholinguistic questions.
• I’m interested in what determines whether a morphologically
complex word is treated as complex or simple by the phonology. Explanations
based on assignment of p-word boundaries seem to cover some of the same ground
as processing explanations (e.g., decomposed vs. direct lexical access). Do we
really need both?
• Cross-linguistic asymmetries concerning left vs. right edge of stems,
or prefixes vs. suffixes, are psycholinguistically tantalizing.
• Because a p-word is roughly the amount of material that is often
supposed to be generated in the lexicon, we can compare the prosodic approach
to competing views of the relationship between phonology and morphology
(lexical phonology, intra-paradigm correspondence).
COURSE PLAN
In the proseminar, we will examine...
• reasons researchers have espoused the p-word (domain for certain
segmental rules, domain for stress assignment, unit in prosodic morphology,
cliticization facts)
• how well the different purposes of the p-word line up: within a
language, can a single algorithm for p-word construction account for the
domains of multiple segmental and prosodic rules?
• competing explanations: boundary types, erasure of morphological
boundaries (whether by rule or for psycholinguistic reasons), interleaving of
phonology and morphology (i.e., Lexical Phonology), paradigm uniformity
• the typology of p-words (what tends to constitute a p-word, and what
phenomena tend to take the p-word as their domain): what is the typology, and
is it better explained by the prosodic approach or by competing approaches?
After a few sessions of lecture-and-discussion to set the stage, we’ll move to student-led discussions of readings, with occasional lectures. I want to try something new (to me): two students will be responsible for each paper, with one student presenting the paper’s analysis (say, a p-word analysis of various segmental phenomena in Irish), and the other devising and presenting an alternative (say, an analysis using boundary symbols). Note “presenting”, not “arguing for”: unlike in debate club, you don’t have to pretend to agree with the side you’re responsible for. Adjustments will be made to this format depending on the type of paper.
COURSE REQUIREMENTS
For 2 units (251B), participate in the discussions, including taking turns to present.
For 4 units (251A), do the above plus write a final paper related to the course topic.
Suggested paper topics—you can propose others:
• Survey a set of roughly word-level phenomena within a language. Can
a consistent p-word structure be proposed for all of them?
• Take on a case (perhaps one encountered in class) where different
diagnostics of the p-word are known to disagree. What account can you propose?
• Compare a p-word-based explanation to a p-word-free explanation of some
phenomenon (perhaps one that’s already been analyzed one way in the
literature).
Proseminar: The Syntax/Semantics Interface (Ling 252A/B)
I'll be teaching a semantics proseminar 252A/B this quarter. The class will be about various issues at the syntax/semantics interface.
We'll be reading and discussing recent published and unpublished workon various issues, with the aim of understanding the proposals, sorting out the data landscape, and identifying areas where those proposals leave room for improvement.
The list of topics is open to student input, but tentatively includes at this point:
- copular constructions (predicational, equative,...), including clefts
- comparatives and superlatives
The first meeting for this class is scheduled for this coming Monday,Oct
2nd, from 2-3:50pm in
Proseminar: Focus & Intonation (Ling 251A/B)
Spring 2006, M/W 2--3:50pm, Public Policy 1256
Have you noticed how problems in intonation and prosody have been in hiding lately?
That's because they know we're out to get them!
We (Sun-Ah Jun & Daniel Buring) will co-teach a proseminar entitled
Focus & Intonation (Ling 251A/B)
This is a class on the relation between syntax, focus, and prosody.
Our aim is to bring together current research on intonational phonology, prosodic phonology, and information structure, working towards a theory that integrates the results of these different sub-fields, with special emphasis on ways to use experimental methods to address theoretical questions. We intend this class to be a starting point for original research in the area of focus and intonation.
The full course description is now available at
http://www.humnet.ucla.edu/humnet/linguistics/people/buring/locker/focus.and.intonation.pdf.
Note that Ling 111 - while useful - is *not* a prerequisite for this class. Neither is knowledge of formal semantics, which is not at all central for our purposes (that's right, folks, no lambdas!).
Hope to see many of you there!
Daniel & Sun-Ah
Linguistics 205: Morphology
Bruce Hayes
Tues., Thurs 9-11, Public Policy 2292
Topics for this year include morphological analysis
(problem sets in
Linguistics 254: Language Processing in Children
Nina Hyams and Carson Schuetze will be doing a proseminar (254) on “Language Processing in Children”, currently scheduled for Mondays 11-2. Details to follow.
Winter 2006
See announcement for Spring 2006, above (Buring/Jun).
Linguistics 251A/B: Speech Perception
This pro-seminar will meet MW 9-11 in Rolfe 3106, together with Colin Wilson’s 217 (Experimental Phonology). 251 lectures are on Mondays, 217 lectures on Wednesdays. Experimental Phonology also has a section meeting some weeks, Mondays 1-2, recommended for 251 students. Some of this quarter’s 422 meetings (Wednesdays 1-2) will also be relevant for this course. The prerequisite for this course is Ling. 104/204 or equivalent.
Topics covered on Mondays include acoustic cues; cue weighting; categorical perception; category structure; gestural perception; role of experience; infant perception; tone perception; statistical methods. The schedule of topics for both 217 and 251 is here.
Linguistics 254: Probability and Statistics in Linguistics
Time: Tuesdays/Thursdays 11-12.50
Where: Bunche 3157
This course is about probability and statistics in linguistics.
The web site is http://www.linguistics.ucla.edu/people/Kracht/courses/mathling2/statistics.html
The previous course on this topic, on which the notes on that web site are based, was quite theoretical and not intended for use in this course. This time I will give a rather more practical course, for which a more appropriate background site is this one:
http://corpus.linguistics.berkeley.edu/~kjohnson/quantitative
The course shall achieve two goals. It will make you familiar with the basics of probability theory, statistics, and R, a software package that allows to do quite complex statistical analysis. It will also tell you what all this has to do with linguistics, and so we shall look at some data and analyse it. The precise content will be tailored according to need. In addition to what the notes by Keith Johnson provide, I may talk about word frequency, probabilistic grammars and/or data oriented parsing.
Linguistics 207: Semantics II
This quarter's Ling 207 (Semantics II) will meet on Mondays and Wednesdays, 9-10:50am, Rolfe 3123.
This course follows 'Semantics I' (Ling 200C), and is highly recommended for students with an interest in meaning. We will discuss the following topics:
Exercises and readings from the literature will be assigned on a weekly basis.
Pre-requisites: C180 or C208, 200C [please contact me if you do not have one the pre-requisites]
For any questions, do not hesitate to contact me.
Linguistics 251
Proseminar: Phonotactics
Meets MW 11-1 in Rolfe 3127.
Phonotactics is the system of principles
that determines the phonological well-formedness of words and other sequences.
In Chomsky and
This proseminar will consider phonotactics from three points of view:
Prerequisite: at least one graduate course in phonology
Course requirements are two exercises and a term paper project.
The Web site for this course is located here.
Linguistics 218
Marcus Kracht and Edward Keenan
The topic for this quarter is "Counting Techniques and Statistics". For details on the course see
http://www.linguistics.ucla.edu/people/Kracht/courses/mathling2/math-ling2.html
Linguistics 254: Comparative Processing of Language and Music
The goal of this proseminar is to compare the processing of language with the processing of music in order to see how research in each of these domains can inform the other.
Parallels between language and music have been noted since ancient times, and concepts from generative grammar have been famously applied to music in a book by Lerdahl and Jackendoff (1983). That work, however, is mostly concerned with characterizing a "competence grammar" of human knowledge of (Western tonal) music, whereas in this proseminar our focus will be on the distinct (though obviously not entirely independent) issue of the processing of language vis-à-vis music, that is, how the two are produced or comprehended in real time, including some discussion of the development of these processing systems in children. There are many parallels in these two processing domains (see below), but differences in the way that music and language are used, particularly in Western culture, make certain activities easier to study in one domain than another, e.g., people tend to listen to the same fairly small set of musical works many times, whereas this hardly ever happens with texts, consequently it is easier for music listeners to focus on interpretive properties to the exclusion of "content". This provides opportunities for complementary research strategies.
I don't have a complete syllabus at this stage because I am still getting to know this literature myself and I want to tailor it to the interests of the participants, but here are some themes that I know there's work on:
Likewise, I won't be presupposing any knowledge of language processing--you needn't have taken 232 or 213C.
My intention, rather unconventionally for a proseminar, is to make this a broad survey, picking and choosing topics that are of most interest to theparticipants. I hope you'll join us!
Linguistics 251: Prosody in Disordered Speech
Place and time: Rolfe 3131, MW 11-1pm. Prerequisite: Ling 211 (Intonation) or knowledge of English ToBI.
Linguistics 252: Case
Hilda Koopman and Anoop Mahajan
The organizational meeting will be held on Tuesday October 5 at 12 noon in Linguistics Conference Room. We plan to meet for about an hour on Tuesday in order to discuss possible schedule options and the structure/coverage of the course.
The broad topic remains the phenomena of case including morphological, quirky and structural case/Case.
If interested, please do come for the organizational meeting. If you would like to attend the seminar but can't make it to the organizational meeting, please send an email to Hilda and me outlining what time slots work the best for you.
The course will meet once a week for 3 hours.
Linguistics 252: Current Issues in Semantics
I'll be teaching a semantics seminar (252a/b-sem2) this Fall, which is scheduled to meet T/Th 2-4pm at a yet-to-be-disclosed location.
The highly original title is `Current Issues in Semantics' -- if you want it even snazzier you can add an umlaut to the `e' in semantics.
Things I thought about including are
Email me with any other suggestions. Though 207 is recommended as a prerequisite, I believe you will be able to follow (most of) this seminar without it.
Linguistics 252: Medieval Logic
This Fall I'll offer a Graduate Seminar on Medieval Logic. The course will be based on a book I am writing on the topic. The aim is to give an account of a kind of notation, and semantics for that notation, that is presupposed by medieval authors on logic. Various medieval doctrines will be formulated and criticized.
Familiarity with the Predicate Calculus with relation symbols and identity will be presupposed. Additional advanced work is logic is desirable.
Students will have an option of writing a research paper or doing weekly problem sets.
The course is currently scheduled to meet on Thursdays 3-6 in Dodd 399.
The website for the course is at http://www.humnet.ucla.edu/humnet/phil/faculty/tparsons/Medieval Logic Fall 2004/
The seminar first meets on September 30.
Tthis quarter I'm giving a reading seminar in Malagasy. We will go through a first grade reader, one collection of (very) short stories, and a selection of largely glossed newspaper articles. Noro Ramahtafandry, a native speaker, will attend the seminar meetings. We meet every Wednesday afternoon 2:00 - 5:00 in the conference room.
Spring 2004
Linguistics 205: Morphology
T/Th. 10-12, 3131 Rolfe Hall
I plan to cover the following topics in this year's offering:
1) Descriptive morphology: survey of phenomena of inflectional and derivational morphology. About three problem sets, mostly solved using the "Extended Word and Paradigm" theory of Stephen Anderson.
2) Morphology and syntax: the Mirror Principle, incorporation, polysynthetic languages
3) Level ordering and bracketing paradoxes
4) The questions of bases: can we locate a member of an inflectional paradigm that serves as the base for formation of the other members? Assessment of the "double-base" phenomenon, of which various examples have been put forth by Steriade and others.
5) Productivity: Wug testing, theories of morphological productivity, how productivity is learned, paradigm gaps, the relation of productivity to level ordering
I will also consider covering particular topics of special interest to the participants.
Work for the course (tentative) is planned to consist of three problem sets and a term paper.
[top]
Linguistics 252
Proseminar: Remnant Movement and Remnant Deletion
Remnant movement has come to occupy a central role in syntactic theory over the last few years. The purpose of this seminar would be to discuss some of the central ideas and analyses that use/promote/require remnant movement. This will include discussion of recent work by Richie Kayne, Gereon Mueller, Koopman and Szabolcsi and others. In the later part of the course, we will discuss remnant deletions deletion of remnant constituents. In particular, the focus will be on the treatment of constructions like pseudo-gapping and gapping (and possibly VP deletions and sluicing). Here we cover some relevant work by Jayaseelan, Lasnik, Johnson, Merchant and others.
The course will meet on Wednesday 2-5 p.m. (If you have are interested in attending the course and have a conflict, please let me know. I know that there is a one hour overlap with TA training course that may affect some of the first years. I will try to see if we can work some compromise there.)
Linguistics 252
Proseminar: Event Semantics
(meets simultaneously with Philosophy 287)
Tuesdays 4-6:50 in Dodd 399
No meeting on April 27. Possible extra meeting (optional) on June 15.
We will cover various issues in Event Semantics. This is an approach to the semantics of natural langauge that posits underlying quantificaiton over events and states in ordinary sentences, such as 'Maria left' and 'Gerald knows Cynthia'. Topics will likely include the logic of modifiers, implicit and explicit talk about events, perception sentences (small clauses), events versus facts, events versus states, unaccusative verbs, causative verbs, thematic relations, states underlying nouns and adjectives. We will spend some time on recent work by Tanya Reinhart that supplements and also challenges much of the theory already discussed.
Each enrolled student will give a 20 minute presentation. Students enrolled for a letter grade will submit a course paper.
For course readings, follow this link.
Linguistics 252
Proseminar: Locatives
Locative expressions are abundant in language, yet they have not been as much studied as, for example, tense. Recently, one notices a growing interest in space and spatial expressions (there is a new series "Explorations in Language and Space" by Oxford University Press and various books among other by Robert Levinson). The proseminar is devoted to the study of locatives. I propose to cover (depending on time) the following topics:
-> Morphological systems (locative cases in various languages, especially Uralic and Caucasian languages, locative PPs in Indo-European)
-> Semantics of spatial expressions: Jackendoff has proposed a layered structure, one element defining the spatial region, the other a path. We first study the subtleties of specifying the region + axes and frames (Levinson): some special properties of Australian indigeneous languages, Inuit, and Oceanic languages.
factors determining the use of Ps (Tversky, Landau)
modes and directionality (special attention to Uralic
languages, in particular Finnish (V. Fong))
orientation (
-> Interactions between syntax and semantics:
are local cases structural or semantic? (Vainikka vs
Niikanne)
case selection and the bimorphemic analysis (Kracht)
-> Metaphorical uses of spatial expressions
fictitious motion, ception (Talmy, Givon)
aspects of cognitive grammar (Langacker)
-> Historical development: how spatial expressions "radiate" into other domains (tense, possession, etc)
Prerequisites: no special knowledge besides basic linguistic training is needed. The mathematics and formal semantics which is involved is fairly basic.
[top]
Linguistics 254
Proseminar: Aspectual Matters in Acquisition
In this proseminar I will focus on the acquisition of Aspect. There are three distinct areas of research into children's aspect development. The first dates back to the early 70's, when various studies seemed to show that children's acquisition of verbal inflections (for tense and grammatical aspect) were heavily influenced by the aspectual properties of particular verbs (actionsarten). The Aspect-before-Tense hypothesis, which claimed that Aspect was an ontologically prior category, developed out of this line of research. This early research was based on children's spontaneous productions. More recently, there has there have been a number of experimental studies in several different languages (e.g. English, Russian, Polish, Italian, etc.) probing children's interpretation of aspectual and tense markers. Although results are not consistent, there seems to be some renewed support for the Aspect-before-Tense hypothesis (now called the Aspect First hypothesis). A third line of research into Aspect development comes from the study of the interpretation of children's nonfinite verbs in root contexts, e.g. root infinitives and bare participles. In my own work, I have argued that Aspect plays an important role in the licensing of non-finite root clauses. Although these three areas of research have not overlapped, I believe there is a unified account to be found.
I would like to accomplish two things in this seminar. First, read the relevant literature in these three areas and try to connect the dots. Second, do hands-on investigation of the aspectual systems in whatever child languages we (collectively) know, using (ideally) the CHILDES database as a first step. This of course requires that we first understand something about how the adult language works in this regard.
The proseminar will be seminar-style, which is to say, everyone will be responsible leading discussions of particular reading and for presenting empirical results on the different child (and adult) languages.
[top]
Linguistics 209A, Introduction to Computational Linguistics
Marcus Kracht
Tues/Thurs.
10-12,
The course in Winter 04 will as usual be taught to both graduate and undergraduate students, and for computer science and linguistics students alike. Computer science students will get a chance to see the linguistic side of it, while linguistics students will get a chance to do some language oriented programming. The language we shall use is OCaML. Since it is not widely known, I will make sure that everybody has access to the program as well as documentation. I will teach how to use it as we go along. The first two week will get people started on the features of OCaML, especially its typing. This is a good excuse to teach something about typing in general. After that we shall look at finite state automata and transducers. Finite state technology has estalished itself quite firmly in in practical applications of phonology and morphology, and there are nowadays quite sophisticated tools for linguistic analysis. We will look at the theory and try to implement some techniques. After that we shall turn to context free grammars. We shall again first deal with the theory, for example parsing techniques and normal forms. Then we shall implement some algorithms.
Prerequisites: Knowledge of either programming (in whatever language) or MathLing 1.
Materials will be posted on
http://www.linguistics.ucla.edu/people/Kracht/courses/compling1/comp-ling-1.html
[top]
Linguistics 221: Computational Semantics
Marcus Kracht
Time and locations to be announced
Bureaucratic note: the Linguistics Department has approved this new course, but it will take a while until it gets onto the registrar's list. If people need to be officially enrolled before January please tell me so and we shall find a solution, perhaps by using a course number from the 25x series.
The course will be flexible. I have several scenarios in mind, which can be invoked and mixed depending on interest:
a) We can study semantical mechanisms in language, such as presupposition projection, anaphor resolution, tracking reference, time and space. Typically, the discussion will be centered around ways in which languages function and what the algorithms look like. (This is what I wrote as the original course definition. However, there is nothing sacrosanct about it.)
b) We can study the syntax-to-semantical representation mapping. There exist precursors to this, for example the book by Kamp and Reyle. But we may look beyond that, looking at the way syntax and morphology go together to define the representation.
c) The experimental variant: define an artificial world, define the semantics of expressions and check out the truth in a model. Have it be a tiny world à la SHRDLU, where the machine can be asked to change things and then ask it questions. Or one could implement a fragment in the Montagovian tradition, now with "real" models in place of formulae.
Prerequisites: Depending on the scenario, more or less computing skill or more or less knowledge of semantics/math.ling./syntax will be necessary.
[top