Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
A Modal Semantics For Copy Raising Diane Doran, McMaster University This work provides a novel semantic analysis of copy raising verbs (CR), i.e. a construction which contains a base generated subject, a verb of perception (looks/seems/sounds), and a like-complement which usually contains a copy pronoun (Kim 2014), e.g. (1) John seems like he is cooking. I argue that CR verbs come with an obligatory and covert “perceiver” argument. I provide evidence that this argument has a temporal dimension, and propose a modal semantic analysis in the style of Kratzer (1981, 1989). Interestingly, sentences like (1) are felicitous only if the speaker can directly perceive John, whereas “expletive” and canonical raising sentences (It seems like John is cooking, and John seems to be cooking) can be uttered when one simply sees pots boiling on John’s stove (the Puzzle of the Absent Cook; Asudeh & Toivonen, 2012). Asudeh and Toivonen account for the pattern by proposing the notion of “perceptual source” (Psource), which in CR constructions is obligatorily the matrix subject, and “perceptual goal” (Pgoal), which is often unexpressed, but represents the perceiver of the event. I present a new observation that the temporal interpretation of the CR verb is necessarily linked to this perceiver, as demonstrated in example (2): (2) Ken’s wife Mary went on a hike without him, and when she came home she was in great spirits. Ken later says to his friend: a. Mary seemed like she enjoyed the hike. b. #Mary seemed to enjoy the hike. I follow Wurmbrand (2014), who argues that when seem takes an infinitive complement, the reference time of the infinitive is given by the reference time of seem. I argue that the contrast in (2) follows if the perceiver/Pgoal is an obligatory argument of the CR verb: in a situation where Ken was not present at the time of the enjoy event (cotemporaneous with the seem event), (2b) is infelicitous, revealing that the Pgoal role must be tied to the time of the matrix seem event. If we accept that the Pgoal is an obligatory argument, the question to be answered is how exactly it is supplied when there is no overt PP. The most obvious assumption would be that the speaker acts as the Pgoal, yet there are cases where it seems that the Pgoal must be someone other than the speaker. As the examples in (3) show, some copy raising verbs do not allow the speaker to express a proposition which they know to be false: (3) You are watching your friend Tom cutting onions, and tears are streaming down his face. a. Haha, Tom looks like he is really upset. b. ?#Haha, Tom seems like he’s really upset. c. #Haha, Tom seems to be really upset. I argue for a modal analysis of the contrast seen in (3). I model the Pgoal as the individual whose impression/knowledge is relevant for providing the worlds to be quantified over, analogous to the role of the judge in predicates of personal taste (Lasersohn 2005, Stephenson 2007). In the absence of overt material stating otherwise, I argue that it is in fact the speaker that fills this role, but the type of restrictors in the lexical entries of look and sound allows the speaker to make statements that are incompatible with their own knowledge. Look and sound restrict the set of possible worlds to only those that are compatible with the sensory evidence that is available to the speaker at utterance time. Seem however restricts the worlds to those compatible with the speaker’s epistemic alternatives, which is a larger set and is determined by the knowledge that the speaker holds at utterance time. Furthermore, I argue that the worlds provided by the restrictor are then ranked according to a stereotypical/normative ordering source to provide the set of BEST worlds (Kratzer 1981, Portner, 1998). This is what makes (3b) and (3c) infelicitous in the above context; it is part of the speaker’s knowledge that Tom is in fact cutting onions, so the proposition that he is upset will not be among the BEST worlds. 1 A Modal Semantics For Copy Raising Diane Doran, McMaster University Bibliography Asudeh, A. (2002). Richard III. CLS, 38, 31-46. Asudeh, A., & Toivonen, I. (2012). Copy raising and perception. Natural Language & Linguistic Theory, 30(2), 321-380. Giannakidou, A., & Mari, A. (2014). Future and universal epistemic modals: reasoning with nonveridicality and partial knowledge. Kim, J. B. (2014). English copy raising constructions: Argument realization and characterization condition. Linguistics, 52(1), 167-203. Kratzer, A. (1981). The notional category of modality. Words, worlds, and contexts, 38-74. Kratzer, A. (1989). An investigation of the lumps of thought. Linguistics and philosophy, 12(5), 607-653. Landau, I. (2011). Predication vs. aboutness in copy raising. Natural Language & Linguistic Theory, 29(3), 779-813. Lasersohn, P. (2005). Context Dependence, Disagreement, and Predicates of Personal Taste*. Linguistics and philosophy, 28(6), 643-686. Portner, P. (1998). The progressive in modal semantics. Language, 760-787. Rett, J., & Hyams, N. (2014). The acquisition of syntactically encoded evidentiality. Language Acquisition, 21(2), 173-198. Stephenson, T. (2007). Judge dependence, epistemic modals, and predicates of personal taste. Linguistics and Philosophy, 30(4), 487-525. Von Fintel, K., & Heim, I. (2002). Lecture notes on intensional semantics. Massachusetts Institute of Technology. Wurmbrand, S. (2014). Tense and aspect in English infinitives. Linguistic Inquiry, 45(3), 403447. 2