Bickel, Balthasar. 1997. Aspectual scope and the difference between logical and semantic representation. Lingua 102, 115-131.11/28/05’s broadcast of Bickel Time is brought to you by Roy Warnock and the 7800 TnA Foundation. Model: - This is a selection based model. - There are phases -) and transitional boundaries (-). - A phase is described variously as a “nonpunctual stretch of time” or “an extended stretch of activity” or “duration” or “temporal extension” (118). - Boundaries are punctual transitions between phases.- Imperfective selects phases. Perfective selects transitions. (124)Main argument: - Semantics representation and logical representation are not the same. - Aspectual operators can bully semantic representations including those built up bypragmatic inference. - BUT logical representations are out of bounds.1. Introduction (115-117)We know that grammatical/viewpoint aspect interacts with Aktionsart. (Here grammaticalaspect refers to the ways of viewing situation types in terms of their internal structure. Aktionsart is the temporal characteristics of predicates or whole propositions.)This interaction is of primary interest in selection theories. The basic idea is that “aspect and Aktionsart are in an operator-operandum relation. Aspect operators can then be said to ‘select’ matching items in the Aktionsart” (116).Example: the French imparfaitGiven the French in (1a) you can conclude (1a’). Given (1b), however, you cannot conclude (1b’):(1a) Il se baignait. (1a’) Il s’est baigné. “He was bathing.” “He bathed.”(1b) Il se noyait. (1b’) Il s’est noyé. “He was drowning.” “He drowned”How does a selection-based model deal with this? The Aktionsart of “to bathe” in (1a) is [φ]. Being telic, the Aktionsart of “to drown” in (1b) is [φ τ]. The imperfective in (1a) and (1b) selects a phase. By selecting the phase in “to bathe” [φ], the whole event takes place. In contrast, in selecting the phase in “to drown” [φ τ] we miss the final transition. So “the subject referent was in a stage that COULD have culminated in being drowned but did not necessarily do so” (116).The telic/atelic difference is a semantic difference, meaning a difference based on language-specific lexical properties. Different languages may have different Aktionsart specifications. Since we know that aspect selects parts of language-specific Aktionsart, “it seems that aspect markers clearly operate on semantic representations” (116).The question to be addressed in this paper is whether aspect markers can operate on otherrepresentations besides semantic ones. Here Bickel calls the domain of operation of aspect markers “cognitive scope” (117).1Bickel’s conclusion is that aspect can operate on representations built by pragmatic information but not logical information. This means that we cannot equate semantic and logical form.2. What kind of information can aspect operate on? (117-120)“Aspect can operate on conversational implicata” (117). Example: The English progressive with punctual predicates. For (3) to be felicitous, it must have a repetitive reading:(3) The light was flashing.The punctual change of state predicate flash only has a transition in the Aktionsart [τ]. Bickel’s selection-based account says that the progressive is an imperfective that selects for phases. If there is no phase for the imperfective to select, then the sentence is ill-formed. Bickel says that (3) is saved from ill-formedness by Gricean implicature. The hearer assumes that the speaker means repeated flashings and this gives a phase for selection. The pragmatically enriched representation is given in (4b):(4a) work (4b) flash | | work’(x) PL(flash’(x)) | | | [φ] φ ([τ]) (The basic lexical Aktionsart is in the brackets.)Bickel argues that a Gricean solution prevents us from having to analyze the imperfectiveas polysemous (with a normal and an iterative reading). If the imperfective is polysemous, how do we avoid getting an iterative reading from a non-punctual predicate like work [φ] in (4a)? Bickel’s analysis allows the imperfective to select the phase in work. Pragmatic enrichment only comes into play if there is no phase to select. Note that Bickel doesn’t address what’s happening with states. If we assume that a state has a phase in the Aktionsart (since it has temporal extension), why are some states allowed in progressive but not others? And why do the allowed states seem to be the exception?(“I’m loving it.” but “*I’m knowing the answer.”)In phase selection, it doesn’t matter whether the phase is part of a verbal predicate (as in 4a work) or linked to an operator (as in 4b and 6). Furthermore, it doesn’t matter how theoperator is derived. Example: The pragmatically derived phase in (4b) flash is just as selectable as the semantic encoded phase from the plural in (6):(5) Chefs, soldats, tous mouriaient “Chiefs, soldiers, all died” | |(6) PL(mourir’(x)) | | φ ([φ τ]) This uniformity is captured in (7).(7) The Aspectual Uniformity Hypothesis: Aspect and Aktionsart representations have the same format and this format is the same on all levels of meaning composition (lexical semantics, morphological derivation, sentential semantics, and pragmatic enhancement).23. Constraining the scope of aspectual operators (120-124)In this section Bickel defends his separation of semantic and logical form and argues that logical form is not within the scope of aspectual operators.Example: English enterGiven that the logical form for(8a) John entered the room.is a change from ~p to p (with p meaning “John is inside the room”), Bickel expects thereto be a transition followed by a phase of being inside, as in (9):(9) BECOME(inside’(x,y)) | | [τ φ]The problem is that the progressive in (10) does not refer to a phase of being inside (This is what Bickel means by “applying the (imperfective) operator to to enter does not give a state reading” (121)):(10) John was entering the room (when Sue called him back.)Bickel says that instead, we get a telic processual