Puzzles in Natural Language
25 novembre 2013
Institut Jean-Nicod, ENS, Pavillon Jardin, 29, rue d'Ulm, 75005 Paris.
Organisateur: Cathal O'Madagin
10.45-11.45: Cathal O'Madagain (IJN),
"Pointing and Indexicals"
Distinctively among semantic types, the reference of a pointing gesture is determined in part by the physical location of the token gesture itself. If I point my finger intentionally at a banana, I refer to the banana; if I reorient my arm in space so that my finger is aligned with the direction of the fridge, I refer to the fridge. Not so, it would appear, for other referring terms: no matter where you put a copy of Hamlet, the meaning in its sentences stays the same. Arguably, the pointing device is the most basic referential device in language-acquisition, providing the foundation for all subsequent referential acts (Tomasello, Carpenter, and Lizkowski 2007). But does the spatially sensitive semantics of pointing carry over into the semantics of other referential devices? Here I argue that another fundamentally important set of referential devices derive their semantics from that of pointing. These are the indexicals – words like ‘I’, ‘now’, ‘here’ and ‘this’. The semantics of these terms has been a topic of ongoing controversy, but I argue that if we pay attention to the spatio-temporal location of token indexicals in their various uses, we can identify a single basic rule that unites and explains their use, and one that suggests that indexicals are in fact modified pointing signs.
11.45-12.45: Jonas Akerman (IJN),
"Subjectivist metasemantics for indexicals"
In a Kaplanian semantics for a language containing indexical expressions, the semantically relevant features of an utterance situation are represented by context-indices, i.e. n-tuples of parameters for speaker (or agent), place, time, etc. Once we try to apply such a semantics to actual utterances, the question arises as to what, more precisely, the semantically relevant features of an utterance situation are. In this talk, I explore a subjectivist approach, according to which context-indices are determined by subjective states of the language users. Contrary to a popular idea, I argue that these subjective states are not to be identified with intentions. Moreover, I show how the proposed subjectivist approach allows the semantic theory to play a central role in an account of communication.
2.15-3.15: Erin Zaroukian (IJN),
"From evaluation to desire: capturing the effect of 'would'"
Across many languages, there is a systematic alternation in which evaluative predicates become desiderative predicates in the presence of would.
(1) a. Sandra likes cookies.
b. Sandra would like cookies.
(2) a. Cookies are great.
b. Cookies would be great.
I (and my colleague Charley Beller) develop a compositional analysis of evaluative and desiderative predicates in which the subjunctive modal would introduces a comparison between the current situation and alternatives situations, whereas bare evaluatives provide a judgment about the current situation. Our analysis captures a range of contrasts between evaluatives and desideratives, including the ability to appear with generic singular indefinites (4) and the availability of faultless disagreement (5)-(6).
(4) a. #Sandra likes a cookie.
b. Sandra would like a cookie.
(5) A: Cookies are great. / Cookies would be great.
B: No, cookies are not great. / No, cookies would not be great.
(6) A: I like cookies. / I would like cookies.
B: #No, I do not like cookies. / #No, I would not like cookies.
3.15-4.15: David Rey (IJN)
"How is Time Represented in English? A Temporalist Approach"
Prior’s seminal work on temporal logic led early formal semanticists to analyze English tenses as intensional sentential operators. Since the early seventies, though, linguists have progressively gathered a powerful body of evidence against the Priorean account of tenses, which has come to be regarded by most theorists working in the field as empirically inadequate. One important source of evidence concerns the ways in which tenses interact with other temporal expressions that appear in the same sentence, including other tenses, time adverbs, and nominal phrases. In this talk I will rely on this data in order to argue for an intensional account where tenses are formalized as non-sentential operators. On the view that I will propose, temporal operators act on a class of pre-predicative syntactic constituents (which I call radicals) rather than acting on sentential formulas. I will argue that this account predicts the right truth-conditions for some kinds of English sentences that were problematic for the standard Priorean account.