Archive for May 9th, 2011
Speaker: Robert Henderson (UCSC)
Title: A corpus-based approach to quantifier scope disambiguation
Time: Monday, May 9, 11:30am
(Joint work with Scott AnderBois and Adrian Brasoveanu)
Most investigations of quantifier scope are concerned with the range of possible scopes for sentences with multiple quantifiers. Instead, this study examines the actual scopes, i.e., the pragmatics of quantifier scope disambiguation. Although actual usage facts fall outside of semantics proper, we are interested in them because they too are facts about natural language interpretation. Moreover, they are facts which provide indirect evidence about the semantics of quantifiers, e.g. if “some” has a stronger preference for narrow scope than “a”, then their semantic representations should arguably be different. To address these questions, we assembled a corpus of sentences from LSAT logic puzzles and tagged it for quantifier scope. In this talk I’ll present our corpus and compare several models of quantifier scope disambiguation.
Speaker: Melissa Kline
Title: Children’s comprehension and production of transitive sentences is sensitive to the causal structure of events
Time: Tuesday 5/10, 46-3189, 12-1pm
Young children learn about causal relationships not only from observation, but also from the language they hear. Two novel-verb studies show that preschoolers recognize that transitive sentences like `Sarah broke the lamp’ express a relationship between cause and effect. Previous work has explored only differences between prototypical events of different categories, systematically conflating causation with a number of other semantic features. The present studies show that children attend specifically to causal structure, as indicated by spatiotemporal contiguity between action and outcome subevents. Children were more likely to give transitive descriptions (‘She wugged it’) to causal than to noncausal versions of an event. Children also reliably selected a causal event when asked to ‘find where she wugged [it].’ Transitive sentences thus provide children with a rich source of evidence about the verbs they learn and the events they encounter in the world.
“Looking beyond English: MIT professor uses linguistics in an ESL classroom to teach scientific principles, empower a new generation of critical thinkers” is a nice article about Wayne O’Neil and Maya Honda’s work with linguistics in high schools.
The last regularly scheduled Phonology Circle for the semester will feature practice talks for RUMMIT (which will take place at Rutgers on Mon 5/16):
Speaker: Coppe van Urk
Title: On the distribution of clashes
Speaker: Suyeon Yun
Title: String-based domains of duration preservation
Time: Tuesday 5/10, 5-6pm, 32-D831
RUMMIT, the semesterly meeting of phonologists from Rutgers, UMass, and MIT will be held at Rutgers on May 16, 2011, from 11am - 5pm.
A schedule and directions can be found here.
There will be two meetings of the LFRG next week.
WHO: Ciro Greco
WHAT: Double Access Reading
WHEN: May 11, 5:00PM-6:30PM (note the unusual time)
In many languages, when a present tense stative predicate is embedded under a past tense attitude verb we observe a particular temporal interpretation of the sentence know as Double Access Reading (DAR). In particular, the embedded eventuality must be interpreted as temporally overlapping both the attitude’s time and the utterance time:
(1) Ezra ha detto che Frida è incinta.
Ezra said that Frida is pregnant
The present tense in (1) seems to be sensitive both to the matrix tense and to the utterance time. Many scholars claimed that DAR is a consequence of a core property of temporal dependencies, namely that tenses in embedded attitude contexts cannot take their denotation only with respect to the utterance time (Enç 1987, Higginbotham 1995, Giorgi 2010). Therefore, tenses in embedded attitude contexts cannot be interpreted as mere indexicals.
However, in the following example, the present tense need not to be interpreted as to the matrix time; the crucial point is the overlapping with the utterance time. Prima facie, (2) represent a problem for the generalization above:
(2) Ezra ha detto che Frida è a casa oggi.
Ezra said that Frida is at home today.
Although (2) seem to show that there are cases in which an embedded tense can be interpreted as a mere indexical, we will argue that this kind of examples do not represent a real counterexample to a theory of DAR. Our aim is to show that these cases can be accounted for using a semantics of futurates along the lines of Copley (2009).
WHAT: SALT practice talks
WHEN: Friday 5/13, 2pm (the usual time)
This Friday through Sunday, May 13-15, MIT Linguistics will host the 20th anniversary meeting of Formal Approaches to Slavic Linguistics (FASL). During the three days of the conference, almost 30 talks (including several plenary lectures) will be presented by Slavic linguists from a dozen different countries. In the two decades since the very first FASL was organized by Jindrich Toman at the University of Michigan twenty years ago, the conference has become one of the major venues for reporting research on the structure of all the Slavic languages. FASL has also served as an inspiration for similarly-themed conferences for other languages areas such as AFLA for Austronesian and FAMLi for Mayan. The second meeting of FASL (which cemented its status as an annual event) was held here at MIT, where it was co-organized by Sergey Avrutin and by Masha Babyonyshev, to whose memory this twentieth-anniversary conference will be dedicated. The FASL talks will span all areas of Slavic formal linguistics: syntax, semantics, morphology and phonology.
So welcome to, ????? ?????????? ?, vitajte na, ????????????? ??, vitejte na, dobrodošli na, witamy na, ??????? ??????? ??, ????? ????? ?, & witaj?e k…FASL! See you there!
Speaker: Bronwyn Bjorkman
Title: The crosslinguistic defaultness of be
Time: Thursday, May 12, 12:30-1:45pm
In this talk I propose that auxiliaries such as be are a morphological strategy for realizing verbal inflection that has been “stranded” by the syntactic derivation. This is an alternative to the view that auxiliaries head dedicated projections idiosyncratically subcategorized for in particular inflectional contexts. I argue that viewing auxiliary verbs as morphological defaults allows us to unify two superficially different patterns of auxiliary use: a pattern like the one we find in English, where auxiliary-taking categories always occur with an auxiliary (i.e. the passive and progressive), and a pattern that we find in languages like Latin and Kinande (Bantu) (among other languages), where auxiliaries only arise in certain combinations of categories. Kinande, for example, has synthetic past tense and progressive forms, but requires an auxiliary to express both at once (the past progressive). In addition to developing a model of verbal inflection that can account for both these patterns, I discuss some implications of the approach for inflectional marking in counterfactuals.