Citation Context

...hen participants are simultaneously presented with spoken language whilst viewing a visual scene, their eye movements are very closely synchronised to a range of different linguistic events in the speech stream. The linguistic sensitivity of this technique has been validated recently in studies by Allopenna, Magnuson & Tanenhaus (1998), Eberhard et al. (1995) and Sedivy et al. (1999). For example, in addition to the demonstration that eye movements are closely synchronised to the referential processing of the concurrent linguistic input (Eberhard et al., 1995; Sedivy et al., 1999; see above), Allopenna et al. (1998) demonstrated that the probability of ®xating one object amongst several when hearing that object's name is closely related, and the ®xations closely timelocked, to phenomena associated with auditory word recognition. This temporal sensitivity to both word identi®cation and subsequent referential processing makes the methodology ideally suited to our investigation of verb-mediated referential processing. 2. Experiment 1 2.1. Participants Twenty-four participants from the University of York student community took G.T.M. Altmann, Y. Kamide / Cognition 73 (1999) 247±264 249 part in this study. Th...

Citation Context

...ies denoted by the subsequent noun vase, and as each adjective is encountered, so the constraints it conveys are used to further re®ne the set of referents which satisfy the accumulating constraints (=-=Altmann & Steedman, 1988-=-; Sedivy, Tanenhaus, Chambers & Carlson, 1999). On hearing an expression such as `pick up the ornate red vase', participants looking at a visual scene containing an ornate red vase will initiate eye m...

Citation Context

...edivy et al. (1999). For example, in addition to the demonstration that eye movements are closely synchronised to the referential processing of the concurrent linguistic input (Eberhard et al., 1995; =-=Sedivy et al., 1999-=-; see above), Allopenna et al. (1998) demonstrated that the probability of ®xating one object amongst several when hearing that object's name is closely related, and the ®xations closely timelocked, t...

Citation Context

...ard et al. (1995) and Sedivy et al. (1999). For example, in addition to the demonstration that eye movements are closely synchronised to the referential processing of the concurrent linguistic input (=-=Eberhard et al., 1995-=-; Sedivy et al., 1999; see above), Allopenna et al. (1998) demonstrated that the probability of ®xating one object amongst several when hearing that object's name is closely related, and the ®xations ...

Citation Context

... establish anaphoric dependencies between that projected expression and the context, and that it can attempt to do so on the basis of the thematic ®t between the entities in the context and the verb (=-=Altmann, 1999-=-). Central to this claim is that what is projected is not structure per se, but interpreted structure. One issue which is not resolved by these data concerns the consequence of verbmediated reference....

Citation Context

...xperience of the entities that play a role in the event to which a verb refers. Thus, thematic roles re¯ect world knowledge that changes dynamically during language learning as experience is amassed (=-=Altmann, 1997-=-, 1999). McRae et al, (1997) argue that when the verb is encountered, verb-speci®c knowledge about typical agents and patients (and whatever other roles tend to be associated with the verb) is activat...

Citation Context

...propriate object in visual context (as determined by that semantic information) even before the semantic properties of the direct object become available. Unlike the word-by-word stop-making-sense judgement task employed by Altmann (1999), the experiments we report below did not require any arti®cial segmentation of the linguistic input, and in the case of Experiment 2, did not require participants to do anything other than look-and-listen. The methodology we adopted for this study, and which was used also by Eberhard et al. (1995) and Sedivy et al. (1999), is based on an early observation by Cooper (1974), who pointed out that when participants are simultaneously presented with spoken language whilst viewing a visual scene, their eye movements are very closely synchronised to a range of different linguistic events in the speech stream. The linguistic sensitivity of this technique has been validated recently in studies by Allopenna, Magnuson & Tanenhaus (1998), Eberhard et al. (1995) and Sedivy et al. (1999). For example, in addition to the demonstration that eye movements are closely synchronised to the referential processing of the concurrent linguistic input (Eberhard et al., 1995; Sedivy et...

Citation Context

...¯ect dependencies between verbs and their objects irrespective of their subjects, whereas plausibility effects re¯ect contingencies between verbs, their objects, and their subjects (so selectional restrictions can be lexicalised in a way that plausibility can not). There do exist data which speak to this latter issue, albeit indirectly. Chambers, Tanenhaus, Eberhard, Carlson & Filip (1998) investigated the processing of prepositional phrases such as `inside the can' in the context of instructions such as `Pick up the cube. Now put it inside the can'. In a visual scene containing just one can, Chambers et al. (1998) reported earlier saccadic movements to the can when the preposition was inside than when it was below (which, given the nature of the visual arrays, was treated as `in front of'), and there was some evidence that eye movements to the can were initiated during the preposition itself. In some respects, this case is equivalent to our own manipulation, using prepositions which either do (inside) or do not (below) select amongst the different entities, instead of, as in our case, verbs which either do (eat) or do not (move) select amongst different entities. Chambers et al. (1998) also report a ca...

Citation Context

...at we believe was extracted at the verbs in our studies, and which guided eye movements to the thematically appropriate visual world object, appears to be of the same kind that McRae, Ferretti and Amyote (1997) referred to during their discussion of verb-speci®c knowledge and thematic role assignment. They view thematic roles as verb-speci®c concepts that are formed through experience of the entities that play a role in the event to which a verb refers. Thus, thematic roles re¯ect world knowledge that changes dynamically during language learning as experience is amassed (Altmann, 1997, 1999). McRae et al, (1997) argue that when the verb is encountered, verb-speci®c knowledge about typical agents and patients (and whatever other roles tend to be associated with the verb) is activated and compared against candidate noun ®llers. Thus, role concepts activated at the verb are compared against one or more lexical concepts. The data we have reported here are entirely compatible with the view espoused by McRae et al. (1997), and although our studies were not designed to test directly the details of the proposal of McRae et al., they do take the theoretical claims further. Our data demonstrate that verbG.T.M....