Presentations

Abstract

c

Abstract

c

Abstract

Recent empirical evidence suggests that language-mediated eye gaze around the visual world varies across individuals and is partly determined by their level of formal literacy training. Huettig, Singh & Mishra (2011) showed that unlike high-literate individuals, whose eye gaze was closely time locked to phonological overlap between a spoken target word and items presented in a visual display, low-literate individuals eye gaze was not tightly locked to phonological overlap in the speech signal but instead strongly influenced by semantic relationships between items. Our present study tests the hypothesis that this behaviour is an emergent property of an increased ability to extract phonological structure from the speech signal, as in the case of high-literates, with low-literates more reliant on syllabic structure. This hypothesis was tested using an emergent connectionist model, based on the Hub-and-spoke models of semantic processing (Dilkina et al, 2008), that integrates linguistic information extracted from the speech signal with visual and semantic information within a central resource. We demonstrate that contrasts in fixation behaviour similar to those observed between high and low literates emerge when the model is trained on either a speech signal segmented by phoneme (i.e. high-literates) or by syllable (i.e. low-literates).

Abstract

In sentence comprehension, readers and listeners often anticipate upcoming information (e.g., Altmann & Kamide, 1999). We investigated two aspects of this process, namely 1) what is pre-activated when anticipating an upcoming word (the contents of predictions), and 2) which cognitive mechanisms are involved. The contents of predictions at the level of meaning could be restricted to functional semantic attributes (e.g., edibility; Altmann & Kamide, 1999). However, when words are processed other types of information can also be activated, such as object shape representations. It is unknown whether this type of information is already activated when upcoming words are predicted. Forty-five adult participants listened to predictable words in sentence contexts (e.g., "In 1969 Neil Armstrong was the first man to set foot on the moon.") while looking at visual displays of four objects. Their eye movements were recorded. There were three conditions: target present (e.g., a moon and three distractor objects that were unrelated to the predictable word in terms of semantics, shape, and phonology), shape competitor (e.g., a tomato and three unrelated distractors), and distractors only (e.g., rice and three other unrelated objects). Across lists, the same pictures and sentences were used in the different conditions. We found that participants already showed a significant bias for the target object (moon) over unrelated distractors several seconds before the target was mentioned, demonstrating that they were predicting. Importantly, there was also a smaller but significant shape competitor (tomato) preference starting at about a second before critical word onset, consistent with predictions involving the referent’s shape. The mechanisms of predictions could be specific to language tasks, or language could use processing principles that are also used in other domains of cognition. We investigated whether performance in non-linguistic prediction is related to prediction in language processing, taking an individual differences approach. In addition to the language processing task, the participants performed a simple cueing task (after Posner, Nissen, & Ogden, 1978). They pressed one of two buttons (left/right) to indicate the location of an X symbol on the screen. On half of the trials, the X was preceded by a neutral cue (+). On the other half, an arrow cue pointing left (<) or right (>) indicated the upcoming X's location with 80% validity (i.e., the arrow cue was correct 80% of the time). The SOA between cue and target was 500 ms. Prediction was quantified as the mean response latency difference between the neutral and valid condition. This measure correlated positively with individual participants' anticipatory target and shape competitor preference (r = .27; r = .45), and was a significant predictor of anticipatory looks in linear mixed-effects regression models of the data. Participants who showed more facilitation from the arrow cues predicted to a higher degree in the linguistic task. This suggests that prediction in language processing may use mechanisms that are also used in other domains of cognition. References Altmann, G. T. M., & Kamide, Y. (1999). Incremental interpretation at verbs: Restricting the domain of subsequent reference. Cognition, 73(3), 247-264. Posner, M. I., Nissen, M. J., & Ogden, W. C. (1978). Attended and unattended processing modes: The role of set for spatial location. In: H.L. Pick, & I.J. Saltzman (Eds.), Modes of perceiving and processing information. Hillsdale, N.J.: Lawrence Erlbaum Associates.

Abstract

Hub-and-spoke models of semantic processing which integrate modality specific information within a central resource have proven successful in capturing a range of neuropsychological phenomena (Rogers et al, 2004; Dilkina et al, 2008). Within our study we investigate whether the scope of the Hub-and-spoke architectural framework can be extended to capture behavioural phenomena in other areas of cognition. The visual world paradigm (VWP) has contributed significantly to our understanding of the information and processes involved in spoken word recognition. In particular it has highlighted the importance of non-linguistic influences during language processing, indicating that combined information from vision, phonology, and semantics is evident in performance on such tasks (see Huettig, Rommers & Meyer, 2011). Huettig & McQueen (2007) demonstrated that participants’ fixations to objects presented within a single visual display varied systematically according to their phonological, semantic and visual relationship to a spoken target word. The authors argue that only an explanation allowing for influence from all three knowledge types is capable of accounting for the observed behaviour. To date computational models of the VWP (Allopenna et al, 1998; Mayberry et al, 2009; Kukona et al, 2011) have focused largely on linguistic aspects of the task and have therefore been unable to offer explanations for the growing body of experimental evidence emphasising the influence of non-linguistic information on spoken word recognition. Our study demonstrates that an emergent connectionist model, based on the Hub-and-spoke models of semantic processing, which integrates visual, phonological and functional information within a central resource, is able to capture the intricate time course dynamics of eye fixation behaviour reported in Huettig & McQueen (2007). Our findings indicate that such language mediated visual attention phenomena can emerge largely due to the statistics of the problem domain and may not require additional domain specific processing constraints.