Functional asymmetry in sheep temporal cortex JonathanW.Peirce 1,2,CA and Keith M. Kendrick 1 1 LaboratoryofCognitiveandDevelopmentalNeuroscience,BabrahamInstitute,Babraham,CambridgeCB24AT,UK 2 PresentAddress:NewYorkUniversity,4WashingtonPlace809,NewYork,NY10003,USA CA CorrespondingAuthor:jwp@cns.nyu.edu Received13August2002; accepted13September2002 DOI:10.1097/01.wnr.0000048023.74602.60 Sheep, like humans, show a bias in favour of the left visual ¢eld whendiscriminating familiar faces.This,inhumans,is thoughtto be causedby a righthemisphere dominance for processing faces involving the temporal cortex. We have directly investigated inter-hemispheric di¡erences in face-processing in sheep by re- cordingthefrequenciesandresponsepro¢lesofsinglecellsinthe temporalcortex whichrespondselectively to faces.While there was no evidence for increased frequencies of face-sensitive neu- roneswithintherighttemporalcortex,ortheirrelativeselectivity for individual faces, there was a pronounced response latency di¡erence between the two hemispheres. The cells in the left hemisphererespondingselectivelytoparticularfacesdidsoupto 400mslaterthanthoseintherightandwithagreaterdegreeof temporalvariabilitybetweencells.Nohemisphericlatencydi¡er- ences were found, however, in other cells responding to general visualstimuliortomanyfaces.Thedatasuggestthat,whilespecia- lised face-sensitive neural circuits in the right hemisphere may playakeyroleintherapid(o 200ms)identi¢cationanddiscrimi- nation of facial identity, those on the left may be involved more with slower processes associated with integrating the emo- tional or mnemonic consequences of recognition. NeuroReport 13:2395^2399  c 2002LippincottWilliams&Wilkins. Key words:Face;Faceperception;Hemi¢eldbias;Lateralisation;Recognition;STS;Temporalcortex INTRODUCTION One of the most robust findings of face recognition studies in humans has been that of hemifield bias. It was originally noted by Wolff [1] that the left side of a face looks more like the owner than the right, when subjects looked in free viewing conditions at stimuli where one side had been mirrored onto the other. Subsequent studies have shown that faces presented wholly in one visual hemifield are recognised faster [2,3] and more accurately [4,5] on the left. Also when chimeric faces (where the left half of one face is combined with the right half of another) are presented [6–9], the left hemifield is more influential in the subjects’ perceptions of faces. These effects are all thought to be caused by the greater use of the right hemisphere (RH) temporal cortex for face perception tasks [10,11]. Sheep are able to recognise [12] and remember [13] both sheep and human faces. Like humans they can recognise familiar faces using internal configural cues [14] and show impaired recognition of inverted faces but not other objects [12]. They also show a robust hemifield bias when perceiving faces [14], recognising faces more accurately from the left than the right side. Furthermore, studies using expression of the immediate early gene c-fos as a marker for neural activation have found the right temporal cortex to be more strongly activated during a discrimination task between pairs of faces than the left [15]. Sheep may, therefore, be a useful model with which to study the precise nature of differential encoding of faces by neural circuits in the right and left temporal cortices using single-cell electrophysiological techniques. Several candidate (non-mutually exclusive) mechanisms may give rise to perceptual asymmetry in face recognition. More cells in one hemisphere could be sensitive to faces, leading to greater processing of the contralateral visual hemifield. Alternatively, the hemispheres may respond with different latencies. Third, the cells in each hemisphere might respond to different characteristics of the faces (e.g. the configuration of features vs the qualities of each feature), such that one hemisphere is more or less specialised in its recognition of faces. The current study aimed to distinguish between these hypotheses by comparing the sensitivity, selectivity and latency of face-sensitive cells in the left and right temporal cortices. MATERIALS AND METHODS Subjects: Two Horned Dorset ewes were used for RH unilateral electrophysiology. A further four sheep were implanted with bilateral recording wells. There was no evidence, from response times or proportions of cell types, that the RH recordings in the unilaterally and bilaterally recorded animals differed (nor was there any reason to 0959-4965  c LippincottWilliams&Wilkins Vol13 No18 20December2002 2395 VISION,CENTRAL NEUROREPORT Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.