Hemispheric Differences in Object Identification

Brain and Cognition 45, 119–128 (2001)doi:10.1006/brcg.2000.1260, available online at http://www.idealibrary.com on

Hemispheric Differences in Object Identification

Sean P. McAuliffe and Barbara J. Knowlton

University of California, Los Angeles

Published online January 17, 2001

Although the right hemisphere is thought to be preferentially involved in visuospatial pro-cessing, the specialization of the two hemispheres with respect to object identification is un-clear. The present study investigated the effects of hemifield presentation on object and wordidentification by presenting objects (Experiment 1) and words (Experiment 2) in a rapid visualstream of distracters. In Experiment 1, object images presented in the left visual field (i.e., tothe right hemisphere) were identified with shorter display times. In addition, the left visualfield advantage was greater for inverted objects. In Experiment 2, words presented in the rightvisual field (i.e., to the left hemisphere) under similar conditions were identified with shorterdisplay times. These results support the idea that the right hemisphere is specialized withregard to object identification. 2001 Academic Press

Key Words: left hemisphere; right hemisphere; object; recognition; laterality; orientation.

Although the two cerebral hemispheres are preferentially involved in processingdifferent kinds of information, the specialization of the two hemispheres with respectto object identification remains unclear. Hemispheric specialization for informationprocessing does not extend to the early visual areas (Hubel & Wiesel, 1959, 1962)—both hemispheres process early visual information in a similar fashion (i.e., bar detec-tion or local Fourier analysis), but each hemisphere processes input from a differenthalf of visual space. Faster processing in a particular hemifield is used as evidencefor specialization in the contralateral hemisphere. While this logic does rest upon anumber of assumptions (e.g., that processing of laterally presented stimuli is notfundamentally different than processing of stimuli presented to the fovea), it has beensuccessfully used to demonstrate some important findings regarding hemispheric spe-cialization (Biederman & Cooper, 1991; Hellige & Cowin, 1996; Hellige, Cowin, &Eng, 1995; Hellige & Scott, 1997; Leehey & Cahn, 1979; Levine & Banich, 1982;Marsolek, 1999; Polich, 1978; Sergent & Hellige, 1986).

With regard to the processing of complex shape information used to identify ob-jects, previous studies using lateralized presentation have often yielded mixed andcontradictory results. A few studies (McKeever & Jackson, 1979; Young & Bion,1981) have reported a right visual field (RVF) advantage for a naming judgment.Unfortunately, as Biederman and Cooper point out in their 1991 study (Biederman &Cooper, 1991), these studies had some significant methodological problems—non-random presentation of images to the two hemifields, repeated presentation, presenta-tion of images in such a way that favored one hemifield, and use of a limited stimulus

Address correspondence and reprint requests to Barbara Knowlton, Department of Psychology, FranzHall, University of California, Los Angeles, Los Angeles, CA 90095. E-mail: [email protected].

1190278-2626/00 $35.00

Copyright 2001 by Academic PressAll rights of reproduction in any form reserved.

120 MCAULIFFE AND KNOWLTON

set. Moreover, other studies failed to find significant differences between hemifieldprocessing of object images (Levine & Banich, 1982; Young & Bion, 1981).

In one of the most carefully controlled studies to date, Biederman and Cooper(Biederman & Cooper, 1991) found no significant effects of hemifield presentationon recognition and priming. While no significant effects were found, Biederman andCooper (Biederman & Cooper, 1991) did report a trend (p , .10) for significantlyfaster naming times for left visual field (LVF) presentation of unprimed objects inExperiment 2. Biederman and Cooper suggest that it seems intuitive that an object’slocation should not have any significant effect on our ability to recognize it (Bieder-man & Cooper, 1991). From an ecological standpoint, this intuition makes perfectsense—because objects appear as often to our left as to our right, there is no adaptivereason for our visual system to preferentially process images appearing to the leftor to the right. However, the visual system uses a neural implementation (i.e., thehemispheres are specialized to some degree) that might produce differences in pro-cessing efficiency for different retinal (e.g., hemifield) locations. While Biedermanand Cooper (1991) suggest that any effects of hemifield presentation are likely to beslight and relatively inconsequential, their use of naming response times as a depen-dent variable might have prevented them from observing significant effects. Becausenaming latencies are heavily influenced by variance in the production of the namingresponse, we used a new paradigm that provided a sensitive measure of visual pro-cessing.

While determining a visual field advantage for processing objects in general mightbe informative, interactions between object stimuli and visual field could provideclearer insight into the functional organization of visual processing because theseinteractions would suggest that the two hemispheres are more efficient at processingdifferent kinds of visual information. For example, evidence from neuropsychologicalstudies suggest that the RH might be better at identifying objects presented in non-canonical orientations (canonical orientations are the typical views of an object; non-canonical orientations represent nontypical views such as rotations in depth and/orrotations in the picture plane; Layman & Greene, 1988; Warrington & Taylor, 1973).Presumably, processing of objects in noncanonical orientations might then be pro-cessed faster in the RH/LVF (relative to the LH/RVF).

Patients with left-hemisphere damage have difficulty identifying objects because ofassociations between object image descriptions and semantic knowledge (Feinberg,Rothi, & Heilman, 1986). If the locus of semantic knowledge about the object is themost important determinant of recognition time, then objects might be identifiedfaster in the LH/RVF, especially if they are presented in canonical orientations. Alter-natively, since the right hemisphere is specialized for visuospatial processing, objectsmight be identified faster in the RH/LVF because the right hemisphere can constructa perceptual representation faster than the LH/RVF. For noncanonical (i.e., inverted)object images, we expected to find a RH advantage based upon the fact that patientswith RH lesions often have difficulty identifying images presented in noncanonicalorientations (Layman & Greene, 1988; Warrington & Taylor, 1973).

In Experiment 1, we used a new visual identification paradigm designed to providea more accurate measure of the visual processing necessary to recognize objects. Ineach trial, participants named a familiar object appearing in one of two rapid serialvisual streams of nonobject distracters (Fig. 1). In the first block of trials, imageswere displayed with very brief (i.e., 60 ms) stimulus onset asynchronies (SOAs). Atthis SOA, object images were very difficult to name. All objects not named in thisfirst block were presented again in a second block of trials where the SOAs of imagesin the visual streams were increased slightly (i.e., 30 ms). This staircase procedurecontinued until an SOA was reached (i.e., 390 ms) where virtually all the objects

HEMISPHERIC PROCESSING 121

FIG. 1. Trial presentation sequence (see text for details).

could be named. The SOA at which an object was first named served as the dependentvariable—this dependent variable provided an estimate of the processing time neces-sary to name each object. This paradigm has been used to detect processing differ-ences as small as 5 ms (McAuliffe & Knowlton, 2000). Participants identified imagesdepicting familiar objects that varied in visual field position (left vs right) and incanonical orientation (canonical vs inverted).

Experiment 2 examined the effects of presentation hemifield on word recognitionin order to determine whether the laterality effects in Experiment 1 were specific tothe identification of objects. Experiment 2 used a paradigm that was similar to Experi-ment 1 with the exception that words were presented in a stream of nonword distract-ers. In addition, the target words were presented in capital letters while the distracterswere presented in small letters.

EXPERIMENT 1

Experiment 1 investigated the effect of visual hemifield of presentation (left vsright) and object orientation (upright vs inverted) on object identification times. Thisexperiment used a new paradigm designed to detect subtle but potentially importantdifferences in hemispheric processing of objects in canonical and noncanonicalviews.

Method

Participants. Twenty-four (16 males and 8 females) University of California at Los Angeles studentsparticipated to fulfill a course requirement. Participants had normal or corrected-to-normal vision. Allsubjects were self-reported as right handed.


Design. This study used a two-factor design with Hemifield (Left vs Right) and Object Orientation(Upright vs Inverted) as within-participants variables.

Materials. Stimuli consisted of black line drawings against a white background. Images subtendedapproximately 2.5° of visual angle. The center of the image was presented 4° from fixation. Stimuliwere constructed using the Canvas graphic program. The stimuli were presented on a Macintosh colormonitor. The entire experimental session was controlled by a program written in the MacProbe (Hunt,1993) experiment programming language.

Procedure. Subjects read aloud all the names of object images at the beginning of the experiment.The experimenter then read instructions to the participant indicating the sequence of a trial. Participantswere instructed that a rapid stream of images would be displayed on both sides of the screen and thattheir task was to identify an object appearing among nonobject distracters in one of the visual streams.Participants were instructed to divide their attention equally to the two streams. Participants viewed thedisplay binocularly from a distance of approximately 70 cm. Figure 1 depicts the sequence of imagepresentations for a trial. Each trial began with a fixation oval displayed for 600 ms. After a blank screenappeared for 600 ms, 17 nonobject distractor images (randomly chosen from a set of 24) and the probeimage depicting an object were displayed in a rapid visual stream with the probe image appearing in aserial position randomly chosen between 7 and 13 (inclusive). In the first block of the probe phase, allstimuli (distracters and probe object) in the visual stream were presented for 30 ms. Presentations werealternated between the left and right visual fields (e.g., left visual field for 30 ms, then right visual fieldfor 30 ms, then left visual field for 30 ms etc.). Therefore, the stimulus onset asynchrony for each imagein a particular hemifield was 60 ms. On half the trials, the first image appeared in the right visual field,on the other half of the trials the first image appeared in the left visual field. After the visual streamshad been presented, the participant was asked to identify the nameable object appearing in one of thetwo visual streams. The first three letters of the participant’s response were recorded by the experimenterand matched against the name of the object to verify the correctness of the response. Any objects notidentified in the first block were presented in the second block, in which objects were presented witha slightly longer SOA (i.e., 90 ms). This procedure was repeated, adding 30 ms to the SOA for eachblock of trials until all objects were identified or a SOA of 390 ms was reached. Objects that could notbe identified with a 390 ms SOA were treated as errors. The participant was given six practice trialsusing objects not appearing in the prime or probe phase.

Each participant identified 48 probe images. Objects were counterbalanced across subjects so thateach object image was equally likely to appear in each hemifield (left or right) and in each orientation(upright or inverted). Across all subjects, both an object image and its left–right reflection were equallylikely to appear. This manipulation insured that no object image could be processed more rapidly in aparticular hemifield by the use of diagnostic features that appeared closer to the center of gaze in thathemifield. For example, a horse facing to the left might be identified more rapidly in the RVF becausethe horse’s head would be closer to the center of gaze and attention.

Equal numbers of objects were presented in each block of trials (i.e., at each SOA). In some cases,objects previously identified were presented again to maintain a balance of objects appearing on eitherside for each block of probe trials with the same SOA. Without this balance, subjects might be biasedto attend to one of the two streams. For example, if processing were superior in the LVF, then moreobjects would appear in the RVF during later probe blocks (because the objects in LVF would alreadyhave been identified). With more objects appearing in the RVF, subjects could attend preferentially tothe RVF. This uneven division of attention might produce a biased measure of the processing capabilitiesof the two hemispheres. The re-presentation of some previously identified objects to maintain positionaluncertainty of targets avoids this problem.

Results

Figure 2 shows the mean display times necessary for object identification timesfor each condition. A 2(upright vs inverted) 3 2(LVF vs RVF) repeated-measureanalysis of variance (ANOVA) revealed a significant effect of visual field [F(1, 47) 518.34, p , .0001] and orientation [F(1, 47) 5 123.70, p , .0001]. The interactionbetween visual field and orientation was also significant [F(1, 47) 5 4.64, p , .05].The ANOVA used subjects as a repeated measure within objects because a previousstudy had shown the variance of objects to be greater than the variance of subjectsusing this paradigm (McAuliffe & Hummel, 1997). For upright presentations, a re-peated-measure t test revealed a significant LVF advantage [t(47) 5 2.88, p , .01]of 7.8 ms (103.8 ms for LVF and 111.7 ms for RVF). For inverted presentations, a


FIG. 2. Experiment 1: Mean SOA necessary for object recognition as a function of PresentationHemifield (Left vs Right) and Object Orientation (Upright vs Inverted).

repeated-measure t test revealed a significant LVF advantage [t(47) 5 3.69, p ,.001] of 19.7 ms (156.2 ms for LVF and 175.9 ms for RVF).

Objects not identified with display times of 390 ms were recorded as errors. Errorrates were low. For upright presentations, there were 1.7% errors in the LVF and0.3% in the RVF. For inverted presentations, there were 5.5% errors in the LVF and5.5% errors in the RVF. A 2(upright vs inverted) 3 2(LVF vs RVF) repeated-measure analysis of variance revealed a significant effect of orientation [F(1, 47) 512.74, p , .001] but no significant effect of visual field (F , 1) on error rates. Theinteraction between visual field and orientation did not reach significance (F , 1).

Discussion

Objects were identified more rapidly in the RH/LVF. In addition, the magnitude ofthis RH/LVF advantage was significantly higher for inverted images. These findingssuggest that the superior visual-spatial processing of the right hemisphere may medi-ate the LVF advantage for recognizing objects in both canonical and (to an even largerextent) noncanonical orientations. It appears that the RH might construct perceptualrepresentations more efficiently—and this superior perceptual processing mediatesthe RH/LVF advantage.

The results reported here are partially consistent with the results of previous stud-ies. Biederman and Cooper (Biederman & Cooper, 1991) did observe a marginallynonsignificant RH/LVF advantage in their Experiment 2 (a 29-ms advantage usingnaming response times), although they only observed about a 4-ms advantage in theirExperiment 1. Taken together, their two experiments showed an LVF advantage thataveraged 16 ms. Although the LVF advantage in the current study was 8 ms forupright objects, this advantage was statistically significant because of lower variance.In addition, this 8 ms advantage represents about 7% of the overall display timenecessary to recognize the object. Moreover, for inverted objects, the LVF advantagewas about 20 ms, which represents about 13% of the display time necessary to recog-nize inverted images. While display time cannot be completely equated with visualprocessing time, the magnitude of the LVF advantage reported here appears to repre-sent a significant portion of the overall visual processing time.


EXPERIMENT 2

Experiment 1 demonstrated an RH/LVF advantage for identifying objects pre-sented in a rapid visual stream. While these results are consistent with the idea thatthe RH is preferentially involved with object processing, it is possible that this resultmay have been due to the fact that the RH is superior at overcoming the effects ofheavy visual masking (Polich, 1978). In Experiment 1, objects were presented in arapid visual stream of images in which each successive image masks the previous one.Experiment 2 used a similar masking paradigm, but used words instead of pictures astargets. Like objects, words can be identified by a unique configuration of compo-nents. However, objects and words appear to rely upon functionally dissociable brainsystems. If masking produced the RH/LVF advantage observed in Experiment 1,then we should observe a similar RH/LVF advantage in Experiment 2. On the otherhand, if the RH advantage reflects superior processing that is particular to objects,then we should observe a LH/RVF advantage in Experiment 2 when words are pre-sented instead of objects.

With regard to the processing of verbal information, a number of previous studiesusing lateralized presentation have showed an LH/RVF advantage (Hellige & Scott,1997; Hellige et al., 1995; Levine & Banich, 1982). However, some of this advantagemay be due to the fact that for languages that read left-to-right, the beginning of theword is closer to the center of attention when presented in the RVF. Following Helligeet al. (Hellige & Scott, 1997), we presented words vertically to avoid the problemof presenting the beginnings of words nearer to the center of gaze (and attention).

Experiment 2 used the paradigm employed in Experiment 1 with words as targetsand consonant string distracters. To facilitate processing, target words were capital-ized and distracters were presented in lowercase letters. The SOA at which a wordwas first identified served as the dependent variable. Equal numbers of three-, four-,and five-letter one-syllable words were used as targets.

Method

Participants. Ten (six males and four females) University of California at Los Angeles studentsparticipated to fulfill a course requirement. Participants had normal or corrected-to-normal vision. Allsubjects were self-reported right handed and were native English speakers.

Design. This study used a one-factor design with Hemifield (Left vs Right) as a within-participantsvariable.

Materials. Stimuli consisted of capitalized three-, four-, and five-letter one-syllable words presentedin the helvetica font. Distracters were consonant word strings consisting of randomly chosen lowercaseconsonants. Both probe words and distracter consonant strings were presented vertically. The stimuliwere presented 2° from fixation and subtended from 2° (three-letter stimuli) to 4° (five-letter stimuli)of visual angle vertically. The letters comprising the stimuli subtended about .5° of visual angle. Thestimuli were presented on a Macintosh color monitor. The entire experimental session was controlledby a program written in the Macprobe (Hunt, 1993) experiment programming language. Participantsviewed the display binocularly from a distance of approximately 70 cm.

Procedure. The procedure in Experiment 2 was identical to that of Experiment 1 with the followingexceptions. Probes were three-, four-, and five-letter words. Distracters were always the same length asthe target word. Subjects were instructed to identify the word in capital letters that appeared in eitherthe left or the right visual stream. Words were counterbalanced across subjects so that each word wasequally likely to appear in each hemifield (left or right). As in Experiment 1, some previously namedwords were re-presented so that equal numbers of words appeared in each hemifield for each block oftrials with the same SOA.

Results

Figure 3 shows the mean display times necessary for word identification for eachcondition. A repeated-measure t test revealed a significant LH/RVF advantage


FIG. 3. Experiment 2: Mean SOA necessary for word recognition as a function of PresentationHemifield (Left vs Right).

[t(9) 5 2.49, p , .05] of 17 ms (151 ms for RH/LVF and 134 ms for LH/RVF).Error rates in the two hemifields were not significantly different [t(9) 5 1.67, p ..10], although there were numerically fewer errors in the LH/RVF (9%) than in theRH/LVF (20%).

Discussion

In contrast to the RH/LVF advantage observed in Experiment 1 for objects, a LH/RVF advantage was observed when words were targets. These results suggest thatthe heavy masking and rapid serial presentation used in this paradigm is not solelyresponsible for the effects observed in Experiment 1. Instead, these findings supportthe idea object shape processing is more efficient in the RH and that word processingis more efficient in the LH.

GENERAL DISCUSSION

Two experiments investigated the effects of hemifield of presentation on objectand word processing. In Experiment 1, objects were identified faster in the RH/LVF,and this RH/LVF advantage was even greater for objects appearing in a noncanonical(i.e., inverted) orientation. In Experiment 2, words presented using a similar proce-dure were identified faster in the LH/RVF.

Given the findings observed here, why is it that Biederman and Cooper (Bieder-man & Cooper, 1991) had difficulty in obtaining reliable effects of hemifield presen-tation? Although Biederman and Cooper (1991) failed to find significant effects ofhemifield presentation on naming times, they did observe a RH/LVF advantage (anaverage of about 16 ms in two studies for unprimed objects) that was numericallycomparable than the advantage observed in this study (about 8 ms for upright objectsand 19 ms for inverted objects). Importantly, the paradigm used in the current studyprovides an accurate estimate of the visual processes involved in recognizing an ob-ject—the nonvisual (e.g., semantic analysis and naming) processes have relativelylittle impact on the dependent variable (i.e., the SOA at which an object can first beidentified; McAuliffe & Hummel, 1997). Therefore, while it may have been difficultfor Biederman and Cooper (1991) to detect a RH/LVF advantage of about 16 mswith an approximately 700-ms dependent variable (i.e., naming time), it was rela-


tively easy for the current study to detect this RH/LVF advantage using the currentparadigm with an approximately 100-ms dependent variable (i.e., SOA necessary tofirst name the object).

While the results of the current study strongly suggest a RH/LVF advantage foridentifying objects (p , .0001), the theoretical implications of such a hemisphereadvantage are not entirely clear. A RH/LVF advantage is not likely the result of thelocus of long-term object shape representations because these representations areusually bilateral (or even favor the LH to some degree) (Feinberg et al., 1986; Mack &Boller, 1977; McCarthy & Warrington, 1986). It may be that the RH/LVF advantageis due to superior perceptual processing in the RH. This idea is further supported bythe fact that inverted objects showed a greater RH/LVF advantage than upright ob-jects. Presumably, the RH was better suited to handle the greater perceptual demandsof identifying inverted objects (compared to identifying upright objects). Results ofprevious studies support this view in that RH damage produces a deficit in the identi-fication of objects presented in noncanonical orientations (Layman & Greene, 1988;Warrington & Taylor, 1973).

Because the RH has been shown to overcome the effects of masking better thanthe LH (Polich, 1978), it was possible that the RH/LVF advantage in Experiment 1was due simply to the heavy masking of the object images in the rapid visual stream.However, the results of Experiment 2 suggest that this is not entirely the case becausewe observed an LH/RVF advantage for word identification even though the stimuluspresentation and the degree of masking was similar to that used in Experiment 1.Instead, the results of the two experiments suggest that there is preferential processingof different kinds of visual information in the two hemispheres (verbal in the LHand object shape in the RH). Although the task used in Experiment 2 is certainly notidentical to the object identification task used in Experiment 1, it does share a numberof important elements (e.g., rapid visual stream identification and distracters that arevisually similar to the target).

The results may be consistent with the idea that the hemispheres are preferentiallyinvolved in processing different spatial frequencies (Delis, Robertson, & Efron, 1986;Fink, Marshall, Halligan, & Dolan, 1999; Grabowska & Nowicka, 1996; Hellige &Cowin, 1996; Kitterle, Hellige, & Christman, 1992; Michimata, 1997). According tothis view, the LH is preferentially involved in processing high spatial frequencieswhile the RH is preferentially involved in processing low spatial frequencies. Presum-ably, if low-spatial-frequency information is registered before high spatial frequen-cies, then the RH might accumulate information about object identity more quicklythan the LH does. However, it is not clear how this account would explain the greaterRH/LVF advantage for inverted objects.

The LH advantage for identifying words is consistent with the results from a num-ber of previous studies (Eng & Hellige, 1994; Hellige et al., 1995; Leehey & Cahn,1979). In the current study, as in a study by Hellige and Scott (1997), this LH/RVFadvantage was observed even though words were presented vertically. Therefore,these results support the idea that the LH is preferentially involved in the processingof verbal information, even when that information is not in a typical word form (e.g.,left-to-right for English; Hellige & Scott, 1997).

In summary, the findings reported here show that the hemispheric specializationinfluences object identification in normal subjects—objects can be identified fasterwhen presented to the RH/LVF. It is likely that this advantage is due to superiorperceptual processing in the RH. In addition, the RH/LVF advantage in object identi-fication was greater for objects presented in a noncanonical (e.g., inverted) view.This result demonstrates that in normal subjects, the right hemisphere is especiallyimportant for processing objects presented in noncanonical views, as has been sug-


gested by the results of neuropsychological studies. For processing words, an LH/RVF advantage was observed, offering further evidence for superior verbal pro-cessing in the LH.

REFERENCES

Biederman, I., & Cooper, E. E. (1991). Object recognition and laterality: Null effects. Neuropsychologia,29(7), 685–694.

Delis, D. C., Robertson, L. C., & Efron, R. (1986). Hemispheric specialization of memory for visualhierarchical stimuli. Neuropsychologia, 24(2), 205–214.

Eng, T. L., & Hellige, J. B. (1994). Hemispheric asymmetry for processing unpronounceable and pro-nounceable letter trigrams. Brain and Language, 46(4), 517–535.

Feinberg, T. E., Rothi, L. G., & Heilman, K. M. (1986). Multimodal agnosia after unilateral left hemi-sphere lesion. Neurology, 36(6), 864–867.

Fink, G. R., Marshall, J. C., Halligan, P. W., & Dolan, R. J. (1999). Hemispheric asymmetries in global/local processing are modulated by perceptual salience. Neuropsychologia, 37(1), 31–40.

Grabowska, A., & Nowicka, A. (1996). Visual-spatial-frequency model of cerebral asymmetry: A criticalsurvey of behavioral and electrophysiological studies. Psychological Bulletin, 120(3), 434–449.

Hellige, J. B., & Cowin, E. L. (1996). Effects of stimulus arrangement on hemispheric differences andinterhemispheric interaction for processing letter trigrams. Neuropsychology, 10(2), 247–253.

Hellige, J. B., Cowin, E. L., & Eng, T. L. (1995). Recognition of CVC syllables from LVF, RVF, andcentral locations: Hemispheric differences and interhemispheric interaction. Journal of CognitiveNeuroscience, 7(2), 258–266.

Hellige, J. B., & Scott, G. B. (1997). Effects of output order on hemispheric asymmetry for processingletter trigrams. Brain and Language, 59(3), 523–530.

Hubel, D. H., & Wiesel, T. N. (1959). Receptive fields of single neurones in the cat’s striate cortex.Journal of Physiology, 148, 574–591.

Hubel, D. H., & Wiesel, T. N. (1962). Receptive fields, binocular interaction and functional architecturein the cat’s visual cortex. Journal of Physiology (London), 160(1), 106–154.

Hunt. (1993). MacProbe. Aristometrics.

Kitterle, F. L., Hellige, J. B., & Christman, S. (1992). Visual hemispheric asymmetries depend on whichspatial frequencies are task relevant. Brain and Cognition, 20(2), 308–314.

Layman, S., & Greene, E. (1988). The effect of stroke on object recognition. Brain and Cognition, 7(1),87–114.

Leehey, S. C., & Cahn, A. (1979). Lateral asymmetries in the recognition of words, familiar faces andunfamiliar faces. Neuropsychologia, 17(6), 619–628.

Levine, S. C., & Banich, M. T. (1982). Lateral asymmetries in the naming of words and correspondingline drawings. Brain and Language, 17(1), 34–43.

Mack, J. L., & Boller, F. (1977). Associative visual agnosia and its related deficits: The role of theminor hemisphere in assigning meaning to visual perceptions. Neuropsychologia, 15(2), 345–349.

Marsolek, C. J. (1998). Object recognition in dissociable abstract and specific neural systems. 38th An-nual meeting of the Psychonomic Society, 72.

Marsolek, C. J. (1999). Dissociable neural subsystems underlie abstract and specific object recognition.Psychological Science, 10(2), 111–118.

Marzi, C. A., & Berlucchi, G. (1977). Right visual field superiority for accuracy of recognition of famousfaces in normals. Neuropsychologia, 15(6), 751–756.

McAuliffe, S., & Hummel, J. (1997). Judgments of visual similarity predicts visual priming for objects.Abstracts of the Psychonomic Society, 2, 7.

McAuliffe, S., & Knowlton, B. (2000). Long term retinotopic priming in object identification. Percep-tion & Psychophysics, 62(5), 953–959.

McCarthy, R. A., & Warrington, E. K. (1986). Visual associative agnosia: A clinico-anatomical studyof a single case. Journal of Neurology, Neurosurgery & Psychiatry, 49(11), 1233–1240.

McKeever, W. F., & Jackson, T. L. (1979). Cerebral dominance assessed by object- and color-naminglatencies: Sex and familial sinistrality effects. Brain and Language, 7(2), 175–190.


Michimata, C. (1997). Hemispheric processing of categorical and coordinate spatial relations in visionand visual imagery. Brain and Cognition, 33(3), 370–387.

Polich, J. M. (1978). Hemispheric differences in stimulus identification. Perception & Psychophysics,24(1), 49–57.

Sergent, J., & Hellige, J. B. (1986). Role of input factors in visual-field asymmetries. Brain and Cogni-tion, 5(2), 174–199.

Warrington, E. K., & Taylor, A. M. (1973). The contribution of the right parietal lobe to object recogni-tion. Cortex, 9(2), 152–164.

Young, A. W., & Bion, P. J. (1981). Identification and storage of line drawings presented to the leftand right cerebral hemispheres of adults and children. Cortex, 17(3), 459–463.

Documents

Hemispheric Differences in Object Identification