All talks are from 3:30 to 5:00 p.m. in room 32-141.
David Beaver (University of Texas, Austin)
IT-constructions
Within the context of a broader project concerning what speakers take to be at-issue, I will present an analysis of the semantics and discourse function of a large range of constructions that I refer to as Inquiry Terminating (IT) constructions. In English, these include it-clefts and exclusives such as "only", "just", and "mere(ly)". I claim that despite their many differences, such constructions cross-linguistically have much in common: they are always focus-sensitive or focus markers, they have a uniform semantics (modulo differences related to syntactic category), and as a discourse function they are always used to mark that a proposition provides a complete answer to what the speaker takes to be the current question.
The similarity of a cleft sentence of the form "It's X that Ys" to a sentence "only X Ys", with the exclusive "only" is easily seen: in both cases X is realized with an intonational focus, and both constructions are used when X Ys and nothing else does. Yet cleft sentences and exclusive sentences have many differences. Some of these are shown in the following minimal pairs, each of which shows a contrast in acceptability when an exclusive is replaced by a cleft.
1a) Not only did Kai laugh, but Irene laughed too.
1b) # It wasn't Kai who laughed, but Irene laughed too.
2a) Mary ate pizza and she only ate pizza.
2b) # Mary ate pizza and it was pizza she ate. (cf. Horn 1981)
3a) Kai: Only David laughed.
Irene: # Yes, and Brady laughed too.
3b) Kai: It was David who laughed.
Irene: Yes, and Brady laughed too.
I aim to account for the main similarities and differences between the meanings of clefts and exclusives, the interaction both types of construction have with focus, and a puzzling range of further data, including evidence drawn from experiments that I've been conducting with colleagues on IT-constructions in English, French, German, and Hungarian. I'll then compare with other accounts, showing e.g. that a recent proposal for the meaning of clefts due to Buring can be subsumed under the current account in all the cases where it works, but not where it doesn't.
Yosef Grodzinsky (McGill)
The Analysis of Negative Quantifiers: Multi-modal Evidence
Recent discussion of negative quantifiers (see Penka, 2011 and references therein) focuses on two main questions: do these quantifiers decompose, and if so, what are the mechanisms for decomposition? In this talk, I will describe a series of neuro- and psycholinguistic experiments my colleagues and I have conducted with healthy and brain-damaged participants that aim to provide relevant evidence. These experiments recorded responses from several modalities, as participants were analyzing sentences with positive and negative proportional and degree quantifiers in German and English (e.g., mehr/weniger-als-die-Hälfte die Kreise sind Gelb, many/few of the circles are blue).
All experiments used a Parametric Proportion Paradigm (PPP): participants were exposed to sentence-scenario pairs, and were requested to make truth value judgments. Sentences contained a quantifier in subject position, and scenarios depicted a proportion between 2 types of objects. Proportion was a parameter, systematically varied across images that were presented with each sentence type.
Our first experiment used functional MR imaging to extract a signal that represents localized brain activity. It aimed to identify brain loci that evince an intensity differential between the contrasting stimuli. Signal intensity for sentences with negative quantifiers was higher than that for their positive counterparts only in Broca’s region. Importantly, no other localizable intensity contrasts were found.
A second experiment (currently only a pilot) confronted English speaking, focally brain damaged, Broca’s aphasic patients with the same task. However here, the dependent measure was error rate. The results suggest a remarkably selective deficit: While patients performed near-normally on the positive quantifiers, their scores were drastically reduced when the stimuli contained negative quantifiers.
A third experiment attempted to take a deeper look at the behavioral signature of quantifier analysis through a study of complex RT functions obtained from healthy participants. Here, too, we observed that the signature of negative quantifiers is quite distinct from that of their positive counterparts.
In this talk, I will try to connect these results, obtained through different modalities from different populations, to previous ones that come from parametric studies of overt syntactic movement with healthy participants in fMRI, and with Broca’s aphasic patients. I will propose that a generalization over the experimental results supports an analysis of sentences with negative quantifiers that assumes covert movement. I will then try to situate these results in the broader context of a research agenda that tries to create a brain map of syntactic and semantic knowledge.
Benjamin Spector (Institut Jean-Nicod)
Generalized Scope Economy (Joint work with Clemens Mayr, ZAS Berlin)
It is a well-known fact that the relative scopes of several operators in a sentence do not always correspond to their relative surface positions. In order to account for such cases, covert scope shifting operations (CSSOs) such as QR and reconstruction have been assumed to apply, resulting in hierarchical structures that deliver the correct interpretation. Such operations, however, are not completely free to apply. One type of restriction on the application of CSSOs has been extensively studied, namely, locality constraints that prevent a CSSO from covertly moving an operator out of a so-called scope-island. There is, however, a second type of restriction that has not been studied to the same extent. In many cases, a CSSO seems to be able to shift the scope of a certain expression x in a sentence S but cannot affect the scope of some other expression – call it y – in a structurally parallel sentence where x has been replaced with y.
Consider for instance the following pair:
(1) Every student did not attend the talk
> Inverse-scope possible: Not every student attended the talk
(2) More than two students did not attend the talk
> Inverse-scope marginal: ??? No more than two students attended the talk.
In other words, the possibility of applying a CSSO seems to depend in part on the nature of the expressions that it targets. As far as we know, the only attempt to account for these types of restrictions in a general way is due to Beghelli & Stowell (1997). These authors propose to account for all the observed restrictions in terms of a cartographic analysis. They assume that CSSOs such as QR or reconstruction target different landing sites depending on the surface position and the nature of the item that undergoes the operation. Although this proposal has broad empirical coverage, it is not really explanatory, since it does not provide a principled account of why the hierarchy of landing sites is the way it is. We develop an alternative approach whose goal is to account for the observed restrictions in a unified way. We propose a new licensing constraint on CSSOs, which is itself a generalization of Fox’s 2000 Scope Economy. In short, we argue that a CSSO can only apply if the resulting interpretation is not logically stronger than or equivalent to the surface-scope interpretation. We will show that such a principle makes correct predictions for a broad range of structures, discuss how exactly it should be implemented, and address apparent counterexamples.
Meghan Sumner (Stanford)
Effects of indexical variation on the perception and recognition of spoken words within and across accents
As listeners, we face a speech signal that is riddled with variation. We are exposed to words, but a single word is produced differently each time it is uttered. These words stream by listeners at a rate of about 5 – 7 syllables per second, further complicating the listeners’ task. How listeners map a speech signal onto meaning despite massive variation is an issue central to linguistic theory. One problem we currently face is that the vast majority of these different realizations of words are understood equally well by listeners.
While speech is variable, it is also informative. In any window of speech, listeners are presented with cues to sounds, sound patterns, words, speakers and their intentions, emotions, accents and other social characteristics. In this talk, I suggest that in order to understand how listeners take all the individual parts of a word that often vary drastically and understand them as quickly and adeptly as they do, we must understand the influence of these different types of information in speech perception.
In this talk, I begin with an assumption that listeners, by default, use these ever-present cues together. I examine the perception of phonological variants sensitive to typically co-present phonetic and social cues. First, I present data from phoneme categorization tasks designed to examine the effects of different phonological variants across different modes of speech (careful vs. casual). Rather than comparing the responses to words with a frequent phonological variant (e.g., tap) to those with a less frequent member of a variant pair (e.g., [t]) embedded in controlled word-frame, I examine the perception of these variants in phonetic and social contexts in which they are typically heard by listeners. Second, I present data from priming tasks designed to examine the recognition of words with different phonological variants across accents. In this case, I show that an out-of-accent variant results in different behavioral responses dependent on the accent of the speaker. From these data, I argue that many effects attributed to phonological variants are illusory and that the emphasis on the variants hides important patterns linking acoustic variation and social representations.
I suggest that the data from these experiments support a view in which cued social attributes influence perception and recognition at a low-level. This work helps explain effects of phonological variants that are oftentimes conflicting, highlights the effects of social factors independent of the rich lexicon, and has implications for how linguistic units are stored and recalled by listeners.
Uli Sauerland (ZAS Berlin)
Attitudes and Embedding
The question I investigate is whether sentence embedding is universally used for the linguistic expression of propositional attitudes (cf. Cristofaro 2003, Oxford UP). Recently, it’s been claimed that in some languages (at least Old Babylonian, Teiwa, Pirahã, Matses, Kobon) propositional attitudes are never expressed by embedding but only by other means: quotation, structures akin to coordination or even independent sentences. I present results mostly from fieldwork investigations of three of these languages (Matses, Pirahã, and Teiwa). My results provide evidence for syntactic embedding in all three languages. They also show some novel variation concerning embedding, namely total indexical shift in Matses and clause-like complementizers in Teiwa.
Hagit Borer (USC)
The Domain of Content
An investigation of the properties of derived nominals reveals a number of rather surprising facts. First, derived nominals with non-compositional Content (or Sense) cannot be Argument Structure nominals (Complex Event Nominals in the sense of Grimshaw, 1990), contrasting, as such, with identical morpho-phonological forms which do happen to have compositional Content (e.g. transformation in its technical linguistic sense vs. transformation as transparently composed from transform). Second, Argument Structure nominals, but not necessarily others, must embed a constituent that is otherwise a possible independent verb, thereby making e.g. aviation, fiction andpetulance perfectly licit derived nominals, but not with an embedded event structure. The contrasts, as it turns out, cannot be accounted for by a lexicalist theory of word formation, nor can they be explained by appealing to any model in which roots are allowed to select arguments. The contrasts, however, can be derived within a wholly syntactic approach to argument structure and to the formation of complex words, in which the domain of non-compositional Content (atomic Content) is defined on the basis of structurally delimited, phonologically realized, syntactic constituents, and is crucially accessible by phase. The argumentation and the conclusions will thus point towards a system of complex word construction which must be syntactic. It will further points towards the need to revise at least some aspects of our understanding concerning the interaction between sound, and specifically phonological realization, and meaning, the latter specifically as in Content.
