1. Hemispheric differences in word meaning processing: Alternative interpretations of current evidence Wiltrud Fassbinder Connie Tompkins University of Pittsburgh Note For ease of exposition, this poster uses language that appears to describe observed functional differences in divided visual field or neurolinguic studies as “functions” of a particular hemisphere. However, while it is assumed that each hemisphere supports processes that are crucial in making meanings available in different ways for further processing, this does not mean that all processes underlying these functions are localized in a single hemisphere.

2. The “Standard” Model Left Hemisphere (LH)  maintains activation for strongly related meanings  Inhibition for weakly related meanings  Lean, compatible & coherent representations, efficient processing Right Hemisphere (RH)  maintains activation for weakly related meanings  No inhibition  Metaphors, jokes, inferences, insight problems Several models of hemispheric differences in word semantic processing converge on the this proposal (Beeman, 1998, Beeman & Bowden, 2000; Burgess & Lund, 1998; Chiarello, 1998; Koivisto & Laine, 2000), here referred to as the “standard” model.

3. The Standard Model  has become widely accepted  is used to guide research into and account for language deficits after left- and right-hemisphere brain damage, e.g., Impairments comprehending inferences or non-literal meanings after right-hemisphere brain damage (Beeman, 1993; Brownell, 2000) Impairments in inhibiting subordinate meanings of ambiguous words in patients with aphasia (Copland et al., 2002)

4. The Standard model  However, Several caveats about methods and interpretation of evidence usually cited in support of the standard model need to be considered, especially with respect to 1. Prime visual field 2. Choice of semantic priming measure 3. Level of processing 4. Targeted aspect of meaning alternative interpretations of evidence possible

5. General Methods  Investigating word meaning activation: semantic priming comparison of related & unrelated prime-target pairs (see below) In DVF presentation, targets always presented in the lateral visual field Primes presented either in lateral or central visual field  Divided visual field (DVF) presentation Words presented to side of visual field, visual information reaches contralateral hemisphere first Performance advantages for one visual field interpreted as respective processing advantages of the target hemisphere Correlations with known functional anatomy, e.g.,  right-visual field/left hemisphere (rvf-LH) advantage for words  left-visual field/right hemisphere (lvf-RH) advantage for global visual processing

6. 1. Prime Visual Field Caveat 1: Central & lateral prime presentations lead to conflicting results  E.g., hemispheric differences for strongly and weakly related word pairs tend to emerge only with lateral primes (e.g., Chiarello et al., 1990).  Several authors have argued that lateral primes are a better measure of hemispheric capabilities because they maximize lateralized processing (e.g., Shears et al., 2003).  Problem: evidence from studies with centralized primes accepted if it supports model, (e.g., Nakagawa, 1991; Burgess & Simpson, 1988), but rejected if it does not support model.  Evidence is not evaluated based on coherent theoretical account of prime presentation differences

7. 1. Prime Visual Field Caveat 2: Lateralized primes might be less ecologically valid  Even if lateralized primes provide better measure of hemispheric capabilities, the extent to which these play a role in normal processing is unclear (Joanette & Goulet, 1998)  However, the standard model addresses hemispheric differences under normal comprehension conditions.  For reading, content words are typically read centrally.  Because DVF studies measure meaning activation of prime meanings, central prime presentation should be more relevant to the standard model.  If central primes are ecologically more valid, only studies with central primes should be considered

8. 2. Choice of Priming Measure  Priming is usually measured for accuracy (ACC) and reaction time (RT)  Two different priming measures are used: Priming (unrelated – related) Most basic priming measure Comparison of  Facilitation (RT/ACC faster/higher than neutral condition)  Inhibition (RT/ACC faster/higher than neutral condition) More detailed: does priming derive from an increase/ decrease in activation of related and unrelated meanings, relative to a “neutral” level ? Examples of word pair types: Unrelated: bread – nurse Related:doctor – nurse Neutral:blank – nurse Priming InhibitionFacilitation Example of priming measures (RT)

9. 2. Choice of Priming Measure  Validity: frequently repeated/ semantically empty primes might be processed differently from other word primes (e.g., Brown, Hagoort, & Chwilla, 2000; Jonides & Mack, 1984).  Uninterpretable Results in DVF studies (e.g., Anaki, Faust, & Kravetz; Burgess & Simpson; Shears & Chiarello, 2003).  Example: reaction times for neutral prime-target pairs from Anaki et al., 1998 The crucial finding – facilitation for metaphorically related meanings for lvf-RH targets - derived from an unexplainable increase in reaction time for neutral primes in the lvf-RH metaphor condition metaphorliteralunrelated rvf- LH 861 ms857 ms852 ms lvf- RH 901 ms861 ms856 ms Caveat 1: Use of neutral primes problematic Reaction times to neutral prime-target pairs in Anaki et al., 1998

10. 2. Choice of Priming Measure Caveat 2: The two measures can lead to conflicting results  Inhibition results consistent with the standard model: weakly related meanings inhibited.  Priming results inconsistent with standard model: weakly related meanings are primed. Example: Priming data from Nakagawa, 1991  If priming is the more reliable measure, studies should be mmcompared based on this measure Priming Inhibition

11. 3. Level of Processing: lexical or semantic?  The standard model addresses semantic processing  Caveat: Almost all relevant studies use associated prime- target pairs Association has been argued to be lexical rather than semantic (e.g., Fodor, 1983) Studies might be measuring lexical rather than semantic effects.  Further studies need to use non-associated stimuli when investigating strongly and weakly related meanings

12. 4. Targeted aspect of meaning 1. Strength of semantic relatedness  Degree of relatedness between prime and target meanings.  Priming reflects network connectivity of meanings 2. Degree of meaning dominance  Degree of strength of each meaning of an ambiguous word.  Priming reflects mapping from word form onto their respective meaning representations Bank institution> Studies cited in support of the standard model rely on two different semantic relationships:

13. 4. Targeted aspect of meaning Caveat 1: The standard model theoretically confounds the two relationships  Lack of theoretical precision in the model  The standard model needs to be incorporate the distinction between these two factors Caveat 2: Dominance studies experimentally confound the two relationships  Targets related to dominant meanings are highly associated, targets related to subordinate meaning are weakly associated  observed hemispheric differences could be due to strength of semantic relatedness, meaning dominance, or an interaction of both  One study that attempted to distinguish the two effects did not include the critical condition of a target strongly related to the subordinate meaning of the prime (Atchley et al., 1999)  Further research needs to differentiate between the two effects

14. Priming results in studies with central prime presentation: semantic similarity  Comparison of results is not straightforward, because they have used different relatedness proportions, and not all studies provide significance data for priming results. But results so far suggest: LH results inconsistent with standard model  Weakly related meanings not inhibited in LH  Activation reflects degree of meaning relatedness Operationalization Strong (s) vs. weak (w) LHRHsignificant “weak” activation in LH? RP Chiarello et al., 1990Associated category coordinates (CCs) vs. non-associated CCs s=w yes.29 Chiarello et al., 1992Associated CCs vs. non-associated CCs s>ws=wyes.50 Nakagawa, 1991 1 Opposites vs. remote associatess>w (57 ms)s>w (37 ms)Probably (28 ms).70 1 No significance data availabe for priming, only numerical results are used. At 750 ms SOA: rvf-LH strong 613 ms, weak, 670 ms, unrelated698 ms; lvf-RH strong 661 ms, weak 698 ms, unrelated 701 ms

15. Priming results in studies with central prime presentation: meaning dominance  Comparing priming results from five studies that used central primes Four studies consistent with standard model (Burgess & Simpson, 1988; Atchley, et al., 1996; Atchley et al, 1999, Atchley et al., 2001) Two studies inconsistent with standard model. However, probably due to  Problematic dominance criterion (Hasbrooke & Chiarello, 1998)  Inexplicable longer RT for neutral primes in a single condition (Anaki et al, 1998)  Results consistent with standard model: only dominant meanings maintained in LH RH also maintains subordinate meanings

16. An alternative interpretation of current evidence  Results suggest different interpretation of priming effects for targets presented to the rvf-LH  LH maintains meanings or features dependent on degree of relatedness  Inhibition of incompatible meanings  No inhibition of weakly related meanings  Priming pattern consistent with results from studies with central primes/ central targets  In these studies, LH drives responses  Because all relevant studies used associated targets, replication necessary  Non-associated targets  For studies of meaning dominance: controlled for degree of semantic relatedness

17. References Anaki, D., Faust, M., & Kravetz, S. (1998). Cerebral hemisphere asymmetries in processing lexical metaphors. Neuropsychologia, 36, 691-700. Atchley, R. A., Burgess, C., Audet, C., & Arambel, S. (1996). Timecourse, context effects, and the processing of lexical ambiguity in the cerebral hemispheres. Brain and Cognition, 30, 277-280. Atchley, R. A., Burgess, C., & Keeney, M. (1999). The effect of time course and context on the facilitation of semantic features in the cerebral hemispheres. Neuropsychology, 13, 389-403. Atchley, R. A., Story, J., & Buchanan, L. (2001). Exploring the contribution of the cerebral hemispheres to language comprehension deficits in adults with developmental language disorder. Brain and Cognition, 46, 16-20. Beeman, M. (1993). Semantic processing in the right hemisphere may contribute to drawing inferences from discourse. Brain and Language, 44, 80 -120. Beeman, M. (1998). Coarse semantic coding and discourse comprehension. In M. Beeman & C. Chiarello (Eds.), Right hemisphere language comprehension: Perspectives from cognitive neuroscience (pp. 255- 284). Mahwah, NJ: Lawrence Erlbaum Associates. Beeman, M.J. & Bowden, E. (2000). The right hemisphere maintains solution- related activation for yet-to-be solved insight problems. Memory & Cognition, 28, 1231-1241. Burgess, C. & Lund, K. (1998). Modeling cerebral asymmetries in high- dimensional space. In M. Beeman & C. Chiarello (Eds.), Right hemisphere language comprehension: Perspectives from cognitive neuroscience (pp. 215-244). Mahwah, NJ: Lawrence Erlbaum Associates. Burgess, C. & Simpson, G. B. (1988). Cerebral hemispheric mechanisms in the retrieval of ambiguous word meanings. Brain and Language, 33, 86- 103. Brown, C. M., Hagoort, P., & Chwilla, D. J. (2000). An event-related brain potential analysis of visual word priming effects. Brain and Language, 72, 158-190 Brownell, H. (2000). Right hemisphere contributions to understanding lexical connotation and metaphor. In Y.Grodzinsky & L. Shapiro (Eds.), Language and the brain: Representation and processing (pp. 185-201). San Diego, CA: Academic Press. Chiarello, C. (1998). On codes of meaning and the meaning of codes: Semantic access and retrieval within and between hemispheres. In M. Beeman & C. Chiarello (Eds.), Right hemisphere language comprehension: Perspectives from cognitive neuroscience (pp. 141-160). Mahwah, NJ: Lawrence Erlbaum Associates. Chiarello, C., Burgess, C., Richards, L., & Pollock, A. (1990). Semantic and associative priming in the cerebral hemispheres: Some words do, some words don't...sometimes, some places. Brain and Language, 38, 75- 104. Chiarello, C., Richards, L., & Pollock, A. (1992). Semantic additivity and semantic inhibition: Dissociable processes in the cerebral hemispheres? Brain and Language, 42, 52-76. Copland, D. A., Chenery, H. J., & Murdoch, B. E. (2002). Hemispheric contributions to lexical ambiguity resolution: Evidence from individuals with complex language impairment following left-hemisphere lesions. Brain and Language, 81, 131-143. Fodor, J. A. (1983). The modularity of mind. Cambridge, MA: MIT Press. Hasbrooke, R.E. & Chiarello, C (1998). Bihemispheric processing of redundant bilateral lexical information. Neuropsychology, 12, 78-94 Joanette, Y. & Goulet, P (1998). Right hemisphere and the semantic processing of words: Is the contribution specific or not? In E. Visch-Brink & R. Bastiaanse (Eds.), Linguistic levels in aphasiology (pp. 10-24). San Diego: Singular Jonides, J. & Mack, R. (1984). On the cost and benefit of cost and benefit. Psychological Bulletin, 96, 29- 44. Koivisto, M. & Laine, M. (2000). Hemispheric asymmetries in activation and integration of categorical information. Laterality, 5, 1-21. Nakagawa, A. (1991). Role of anterior and posterior attention networks in hemispheric asymmetries during lexical decisions. Journal of Cognitive Neuroscience, 3, 313-321. Shears, C. & Chiarello, C. (2003). No go on neutrals? An interhemispheric account of semantic category priming. Laterality, 8, 1-23.