Linguistic Influences in Categorization Are Phonological Priming Effects

Linguistic Influences in Categorization Are Phonological Priming Effects

Linguistic Influences in Categorization Emmanuel M Pothos1 Nick Chater2 Andrew J Stewart3 1 Department of Psychology, University of Edinburgh 2 Department of Psychology, University of Warwick 3 School of Psychology, Bangor University [email protected] There is an abundance of experimental evidence for how language affects similarity and catego- rization. But no account has been forthcoming as to how these results potentially affect formal models of categorization, that are typically insensitive to language information. In this paper we explore this issue via simple categorization tasks under different linguistic manipulation con- ditions. More specifically, we found that participants’ expectations about category structure is indeed affected by the linguistic label given to the category exemplars presented, and that such effects appear to be beyond the scope of current formal models of categorization. Are Phonological Priming Effects Fragile? The Interaction of Phono- logical and Orthographic Forms in Masked Priming Ram Frost Riki Gotesman Sarah Tayeb Merav Ahissar Department of Psychology, The Hebrew University, Jerusalem, Israel [email protected] Most demonstrations of prelexical coding of phonology involve the backward masking paradigm. In contrast, forward masking is considered to be sensitive to orthographic form rather than to phonological structure. Using forward masked priming, we examined how phonological and or- thographic similarity interact in the recognition of target Hebrew words or nonwords. The deep Hebrew orthography allows simple manipulations in which the orthographic structure is kept constant while phonological dissimilarity gradually increases, and vice-versa. We measured the impact of phonological and orthographic similarity in various contrast/luminance conditions. Our results showed an identical linear effect of letter or phoneme alterations. Thus, we obtained clear evidence for both orthographic and phonological priming. Moreover, the effects for words and nonwords were identical. These results demonstrate that even subtle phonological manip- ulation affects masked priming, and that this effect reflects prelexical coding rather than post- lexical search. Examining the Contact Account of Own-Race Bias in Face Pro- cessing Using a Flicker Paradigm: Support for the Development of Differential Perceptual Skills Ben C Jones School of Psychology, University of St. Andrews, Scotland, UK [email protected] Using Asian and Caucasian face images presented within a flicker paradigm (Rensink, O’Regan and Clark, 1997), a statistically significant own-race bias in face-change detection rates is demon- 1 strated for both Asian (N = 32) and Caucasian (N = 32) participants. The adoption of the flicker paradigm allows the role of low-level perceptual skills to be disambiguated from higher-level, semantic-driven processes as change detection represents a perceptual discrimination task em- ploying sub-optimal (i.e., very brief) stimulus presentation. The observed own-race bias in change detection rates can thus be attributed to the development of differential (i.e., race specific) low- level perceptual skills and is consistent with the contact account of own-race bias in face process- ing (in which differential contact with own- and other-race face exemplars generates the devel- opment of differential perceptual skills—Valentine and Endo, 1992). The Representation and Processing of Integers Joseph Tzelgov Ben Gurion University of the Negev, Beer Sheva, Israel [email protected] Subjects were presented with pairs of numbers for magnitude comparison. The set of numbers contained both positive and negative integers. Performance in three tasks were evaluated: (a) numerical comparisons of numbers differing only in their numerical size, (b) numerical com- parisons of the absolute magnitude of numbers differing only in their numerical size, and (c) physical comparisons of numbers differing in physical and numerical size. The results obtained were consistent with the assumption that integers are represented in terms of two dimensions, magnitude and polarity, and only the processing of magnitude is automatic. Individual Differences in Strategies to Deal with Working Mem- ory Tasks Edward N˛ecka Jarosław Orzechowski Institute of Psychology, Jagiellonian University, Poland [email protected] In spite of Saul Sternberg’s conviction, STM scanning is not always serial and exhaustive. Whether it is serial or parallel, and exhaustive or self-terminating, depends on task’s conditions and in- dividual preferences. In a series of five experiments it has been found that slow presentation of stimuli (1000 ms. per item) usually produces serial and exhaustive effects, as first observed by Sternberg. However, quick presentation of stimuli (300 ms. per item) often makes scanning pro- cesses more parallel and self-terminating. Moderate speed of presentation (750 ms. per item) is the one which allows individual preferences and strategies to appear. Some people behave in the way which suggests parallel and self-terminating scanning, others stick to serial and exhaustive strategy of search. These differences partly depend on individual storage capacity of one’s WM. People wit capacious WM are more often inclined to adopt serial strategy, whereas those with less capacious WM prefer parallel search. Focusing in Complex Sentences Jamie Pearson1 Rosemary J Stevenson1 Massimo Poesio2 1 HCRC, University of Durham, UK 2 ICCS and HCRC, University of Edinburgh, Scotland UK [email protected] 2 According to centring theory (Grosz et al, 1995), the utterance subject is the highest ranked en- tity and hence the most accessible to a subsequent pronoun. However, “utterance” (the focusing unit) is not defined. If the unit is the sentence, then ranking is difficult to determine in complex sentences in which each clause has a separate subject (Kameyama, 1998; Suri & McCoy, 1994). Alternatively, the unit may be the clause, with the ranking in the final clause determining acces- sibility for a subsequent pronoun. Two self-paced reading-time experiments investigated this issue, using complex sentences containing that-complements. Target sentences contained a subject pronoun referring either to the main clause subject of the preceding (complex) sentence or to the complement clause subject. Reading times were faster when the pronoun referred to the main clause subject, suggesting that the complement clause is embedded and does not update the focus. We conclude that the focusing unit is the sentence. Modelling Language Processing Asymmetries in the Cerebral Hemi- spheres Padraic Monaghan1 Michal Lavidor2 Richard Shillcock1;3 Andrew Ellis2 1 Institute for Adaptive and Neural Computation, Informatics, Edinburgh Univ. 2 Department of Psychology, University of York 3 Department of Psychology, University of Edinburgh [email protected] Speculations concerning language processing in the cerebral hemispheres have recently benefit- ted from connectionist models that instantiate the separate processors of the two hemispheres (e.g., Reggia et al., 1998). We present a connectionist model of lexical decision that implements the anatomy of the visual pathways such that the right visual field (RVF) projects initially to the left hemisphere (LH), and the LVF to the RH (Shillcock, Ellison & Monaghan, 2000). Lavidor and Ellis (submitted) reported that orthographic neighbourhood size influences lexical decision when words are presented to the RH but not to the LH. After training on a realistic corpus of words, the model closely reflects the asymmetric neighbourhood effect. We discuss the implications of this asymmetry as an emergent effect of the anatomical structure of the model interacting with the characteristics of the lexicon. Verbalisation may Interfere With or Facilitate Multiple Face Recog- nition Charity Brown Toby J Lloyd-Jones Department of Psychology, University of Kent at Canterbury, UK [email protected] A substantial line of research shows that participants describing a previously seen face (compared to those providing no description) are less accurate in later identifying that face from a line-up of faces (e.g., Schooler & Engstler-Schooler, 1990). However, two experiments reported here re- veal both interference and facilitative effects of verbalisation upon the speed and accuracy with which multiple to-be-remembered unfamiliar faces are later recognised. Compared with provid- ing no description, describing a single face following the presentation of multiple non-verbalised faces elicited longer response latencies for later recognition of those faces (Experiment 1). In con- trast, describing each face presented at study benefited later recognition accuracy (Experiment 2). 3 Although verbalising a single face after presentation of multiple faces may ‘overshadow’ the ap- plication of perceptual processes suited to face recognition, verbalising each individual face after presentation provides protection from verbal overshadowing. Number Processing Modulates the Speed and Accuracy of Spatial Behaviour Martin H Fischer Psychology Department, University of Dundee, Scotland [email protected] Parity judgments for small numbers (1, 2) are faster with the left hand, while parity judgments for larger numbers (8, 9) are faster with the right hand. This response speed effect suggests a Spa- tial Numerical Association of Response Codes (SNARC effect; Dehaene et al. [1993], JEP-General, 122, 371–396). Bisecting long digit strings with a pencil showed a similar

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    239 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us