Perception & Psychophysics1974, Vol. 15, No.2, 383-390

Dichoptic and dichotic micropattern discrimination*E. WILLIAM YUND and ROBERT EFRON

Neurophysiology-Biophysics Research Laboratory, Veterans Administration Hospital150 Muir Road, Martinez, California 94553

Ss compared two rapidly successive, brief, discriminably different stimulus elements, called amicropattern, with a second micropattern composed of the same two stimulus elements presented inreverse temporal order. Discriminations could be made between two such micropatterns in the monaural(monocular) as well as in the dichotic (dichoptic) modes of presentation. Discrimination betweenmicropatterns was based on the perceptual dominance of the temporally trailing stimulus element inboth modalities and in both modes of presentation. While monaural (monocular) micropatterndiscrimination is significantly superior to dichotic (dichoptic) discrimination, the existence of dichotic(dichoptic) discrimination demonstrates that no essential peripheral process is required for micropatterndiscrimination.

Recent experiments in this laboratory in the visual,auditory, and vibratory modalities (Efron, 1973) havebeen concerned with the discrimination of brief,complex stimuli differing only in their fine temporalstructure. In these experiments, Ss compared twoconsecutive, brief, discriminably different stimuluselements, called a micropattern, with a secondmicropattern composed of the same two stimuluselements presented in the reverse temporal order. Ssreadily discriminated between micropattems havingreversed temporal order of stimulus elements in all threemodalities, even when the temporal asynchrony betweenthe stimulus elements was so far below the threshold formaking temporal order judgments that eachmicropattern was experienced as a unitary perceptualevent. The discriminations were based on the perceptualdominance of the second or trailing stimulus element.Additional studies revealed that the trailing stimuluselement phenomenally dominated the perception ofeach micropattem by virtue of a retroactive degradationof the perceptual experience of the leading element. Thisperceptual degradation of the leading stimulus elementby the trailing element did not have the properties ofretroactive masking.

In all the previous experiments, the two stimuluselements of each micropattem were presented to thesame eye, ear, or skin location-an experimentalparadigm which did not permit any analysis of theneuroanatomical locus of the retroactive interaction. Inthe experiments to be described here (in the auditoryand visual modalities), the same micropattems werepresented dichotically (dichoptically)-one stimuluselement of the micropattern to each ear (eye). If there isan essential peripheral process involved in micropattemdiscrimination in the basilar membrane or in the retina,the phenomenon of perceptual dominance of the trailingstimulus element would not be expected in such dichotic(dichoptic) presentations. The purpose of the presentexperiments was to determine (l) whether dichotic and

*From the VA Hospital and the Department of Neurology,University of California, School of Medicine, Davis, California.

383

dichoptic micropattern discriminations are possible,(2) whether the discriminations (if possible) are based onthe perceptual dominance of the trailing stimuluselement as was found in the monaural and monocularmodes, and (3) whether these discriminations in thedichotic and dichoptic modes are affected in the sameway by parametric manipulations as are those in themonaural and monocular modes.

METHODS

InstrumentalA complete description of the instrumentation and stimulus

parameters employed in these experiments has been previouslypresented (Efron, 1973) and will only be summarized here. Forthe auditory experiments, the outputs of two oscillators ofdifferent frequency were passed through attenuators, and thenthrough audio gates which generated brief tone bursts. Theduration of each tone burst and the sequence of the two tonebursts were controlled with a digital logic system (lconix6257-6010-6171) accurate to ±.01 msec. For the monauralexperiments, the output of the two audio gates was electricallyadded, passed through a bandpass filter, and delivered viacalibrated earphones to a single ear. For the dichoticexperiments, the output of each gate was passed through its ownbandpass filter and then delivered via a reversing switch to eitherthe right or left ear. For the visual experiments, the pulsed lightoutput of two glow modulators was passed through neutraldensity filters and a Wratten colored filter. The colored beamsfrom these two filters were optically combined using a V-shapedtwo-channel fiberoptic bundle-the end of which was defocusedto the plane of a circular aperture (5.5 deg), which was thenobserved in Maxwellian view. For the dichoptic experiments,two independent optical systems were used and the twoapertures were viewed binocularly (visually superimposed). TJietiming sequence and duration of each of the two colored lightflashes were controlled by the same digital logic system used forthe auditory experiments.

PsychophysicalIn the previous study of monaural and monocular

micropattern discrimination (Efron, 1973), a two-alternative"same-different" paradigm was employed. This same methodcould not be used for the present experimen ts as a consequenceof two confounding phenomenal cues in the dichotic mode-anear asymmetry cue and a sound localization cue. Theconfounding nature of these cues and the experimental designemployed to circumvent these difficulties is illustrated in Fig. 1.

384 YUND AND EFRON

hI MICROPAHERN 2nd MICROPAlTERN

A1 10001 11 JuooIR ~ L( L§

B1 \0001 11 ~R ~ u L§U

C 1 jJ5oo) Ji Jlsoo\R ~ Ll ~

0 1 jJsooJ ..J1 lo0oiR ~ Ll [§]

E1 jT5oo\ )1 ~R LOi§.U L[ ~

F1 InooI ...J1 IuooIR L::i§£] LL L:§U

G 1 I?ooI ...J1 InooIR ~ L( L:i§J

1.0 seeI~I

Fig. 1. Diagramatic representation of experimental paradigm.See text for details.

The two mjcropatterns illustrated in Fig. IA (where noelement asynchrony exists) are not discriminable when presentedmonaurally. They will be discriminable dichotically if anyasymmetry exists between the two ears. Two types of earasymmetry are known: (l) a difference in sensitivity of the twoears, and (2) a right or left ear "dominance." such as thatdescribed by Broadbent (954), Kimura (1961, 1967), Loweet al (1970), Efron and Yund (1974), and others in a widevariety of dichotic listening experiments. Both types ofasymmetry will give rise to the same phenomenal cue. Forexample, if the left ear is more sensitive (or more dominant)than the right, then the 1,500-Hz tone in the left ear in the firstmicropattern of Fig. lA will be louder than the 1,300-Hz toneburst in the right ear. The perceived dichotic chord will bedominated by the 1,500-Hz tone. For the same reason, the1,300-Hz tone will perceptually dominate the secondmicropattern. The S will consequently report that the twomicropatterns are discriminably different.

To circumvent any ear asymmetry cues, all dichoticmicropatterns were presented so that the same frequency wasdelivered to the same ear for both micropatterns of a single trial(Fig. IB). While this mode of presentation effectively eliminatesany cues resulting from an ear asymmetry, it introduces alocalization cue for all dichotic micropatterns with a nonzeroelement asynchrony. Thus, the first micropattern of Fig. IC willbe localized to the left of the midline, and the second will belocalized to the right of the midline-and this would be readilydetected by the S. However, when two identical micropatternsare presented, as in Fig. ID, they would have the sameintracranial localization. Even if no dichotic micropatterndiscrimination were possible, the use of a same-different methodcould still yield a 100% discrimination score merely on the factthat the micropatterns of Fig. ID have the same localization,while those of Fig. IC have a different localization. (The S couldreply "same" when the localization of the two micropatterns didnot. alter within a trial and could reply "different" when thelocus did alter within a triaL) In sum, same-different judgments

could not be employed since there was no way to presentasynchronous dichotic micropatterns with 'the same frequencyelements in the same ear for both micropatterns of a single trialwithout introducing the localization cue just described.

In the method employed for the present series of experiments,the S received only trials in which the first and secondmicropatterns were different with respect to the frequency ofthe trailing element (illustrated in Figs. lC, IE, IF, and IG). Hewas required to report if the first micropattern in each trial wasof higher or lower pitch than the second. The validity of thisform of report had already been established by the previousstudies (Efron, 1973), which unequivocally showed that thebasis for monaural micropattern discrimination was theperceptual dominance of the pitch of the trailing stimuluselement.

If only Configurations IC and IE were employed (where the1,500-Hz tone is always in the left ear in both micropatterns of asingle trial), then the micropatterns in which the 1,500-Hz tonewas the trailing element would always be localized to the right ofthe midline. Similarly, the micropattem in which the 1,300-Hztone was the trailing element would always be localized to theleft of the midline. If the S became aware of this cue, he mightthen base his reports on this localization difference. Tocircumvent the use of these localization cues, the Ss were alsopresented with trials having configura tions illustra ted in Figs. IFand IG. Configuration F is like ConfIguration C in that the firstmicropattern is heard on the left and the second is heard on theright-but, if there is a perceptual dominance of pitch of thetrailing element, then Configuration C will be reported as havinga low-high pitch sequence and F will be reported as having ahigh-low pitch sequence. Similarly, E and G have the samelocalizations, but would have opposite pitch sequence reports.An experimental session for a given stimulus element asynchronyconsisted of 25 trials of Configurations C, E, F, and G inpseudorandom order.

Since this psychophysical method was mandatory in thedichotic experiments, identical procedures were used in themonaural experiments so that the data of the two modes ofpresentation would be comparable.

Essentially identical methods were employed in the visualmodality. The Ss were presented with micropatterns consistingof brief flashes of light of differen t hue, i.e., red and greenstimulus elements. In this case, both micropatterns of a trial areperceived as flashes of yellow light. The Ss were required toreport if the first micropattern was more greenish yellow ormore reddish yellow than the second. In the dichopticexperiments, it was not necessary to take the elaborateprecautions to avoid localization-like phenomena, and thus thegreen flash was always delivered to the left eye and the red flashwas always delivered to the right eye.

All three Ss used for these experiments had had extensiveprevious experience in other visual and auditory psychophysicalstudies.

Nontemporal Stimulus ParametersFor the auditory experiments, the two stimulus elements were

tones 25 dB above the detection threshold determined for eachS. This low intensity level was used to avoid any effects of boneconduction across the head in the dichotic presentations. Thesame intensity levels were used in the monaural condition. ForExperiments I and 2 in the auditory modality, the frequenciesof the tones were 1,300 and 1,500 Hz. For Experiment 3, whichwas concerned with the effect of the difference in frequencybetween the two stimulus elements, a variety of frequencies wasemployed (vide infra).

For the visual modality, experiments were designed to testperformance with two opponent-color pairs. For the red-greenstimulus pair, Wratten No. 29 (red) and Wratten No. 58 (green)filters were used. For the monocular experiments, theluminances were approximately 110 and 300 fL, respectively.This value gave rise to a well-saturated yellow flash when the two

DICHOPTIC AND DICHOTIC DISCRIMINATION 385

stimulus elements were both 10 msec and simultaneous. For thedichoptic experiments, the red flash was 7.1 fL and the greenflash was 300 fL, This ratio yielded a yellow flash closelyresembling that used in the monocular mode.

For the blue-yellow stimulus pair, Wratten No.47 (blue) andWratten No. 15 (yellow) were used. In the monocularexperiments, the luminances were approximately 17 and 620 fL,respectively. This ratio of luminances gave rise to a white flashwhen the two stimulus elements were both 10 msec andsimultaneous. For the dichoptic experiments, the luminanceswere 27 and 478 fL, respectively-which resulted in a white flashof closely similar appearance.

[It will be observed that different ratios of luminances for thetwo colors had to be used in the dichoptic mode compared tothe monocular. These values were determined by adjustment ofthe dichoptic intensities to produce a hue similar to thecorresponding monocular stimulus. The requirement for morered, in the red-green case, and more blue, in the blue-yellow case,is in agreement with the work of Thomas, Dimmick, and Luria(1961) on binocular color mixing.)

Only one of the three Ss was able to achieve dichoptic colormixing-a finding not surprising in view of the well-knowndifficulties associated with binocular color mixing. The S whocould perform the dichop tic experiments was one of the authors(B.Y.), who had had previous experience with dichoptic visualstimuli.

Temporal ParametersThe temporal relationship between the stimulus elements of

the micropatterns was the independent variable in Experiments 1and 2. Figures 2A, 2B, and 2C illustrate the three types oftemporal relationships that were explored for both modalities.Figure 2A illustrates the dichotic (or dichoptic) condition inwhich both stimulus elements of each micropattern are of equal(10-msec) duration, and the onset and offset asynchronies alsohave equal durations. Figure 2A only illustrates theconfiguration seen in Fig. IC; configurations seen in Figs. IE,IF, and IG, while presented to the S, are not illustrated. For the

11II

/--1 ONSET ASYNCHRONY

}--: ELEMENT x DURATION

1--1 ElEMENT Y DURATION

1-; OFFSET ASYNCHRONY

100

90

fi5OOlL~_--~

ASYNCHKOHY' m.o<

Fig. 3. Comparison of dichotic and monaural micropatterndiscrimination as a function of element asynchrony. As indicatedby the inserted schematics, the onset and offset asynchronieswere always equal Their absolute values are indicated on theabscissa. The percent discriminated is plotted on the ordinate.

auditory modality, X was 1,500 Hz and Y was 1,300 Hz. For thevisual modality, X was red and Y was green in one experiment,While X was blue and Y was yellow in a second experiment.These same experiments were repeated in the monaural andmonocular modes in which the two stimulus elements of eachmicropattem were presented to the same ear (eye). The temporalrelationships, however, were identical.

Figure 2B illustrates the dichotic (or dichoptic) condition inwhich the stimulus elements of each micropattern haveasynchronous onsets but synchronous offsets. Trials illustratedin Figs. IC, IE, IF, and IG were presented, but onlyConfiguration IC is illustrated in Fig. 2B. Experiments with thistemporal configuration were performed monaurally(monocularly) as well as dichotically (dichoptically).

Figure 2C illustrates the dichotic (or dichoptic) condition inwhich the stimulus elements of each micropattern havesynchronous onsets but asynchronous offsets-once againillustrating only the configuration seen in Fig. lC. Theseexperiments were also performed monaurally and monocularly.

In all these experiments (Figs. 2A, 2B, and 2C), theindependent variable was the element asynchrony, the specificvalues of which will be described separately for each experiment.

Fig. 2. Diagramatic representation of the three types ofstimulus element asynchronies employed in these experiments.Only the dichotic (dichoptic) forms of these micropatterns areillustrated. In A, the onset and offset asynchronies are bothpresent and equal. In B, only onset asynchronies are present. InC, only offset asynchronies are present.

IONSET ASYNCHRONY =0

r--- ElEMENT x DURATION

I ELEMENT Y DURATION

r--' OFFSET ASYNCHRONY

1-' ONSET ASYNCHRONY

I I ElEMENT X DURATION

,---, ElEMENT Y DURATION

IOFFSET ASYNCHRONY· 0

RESULTS

Experiment lA-Monaural and Dichotic MicropattemDiscrimination

The temporal configuration of the auditorymicropattems employed in this experiment areillustrated in the schematic diagram of Fig. 'lA. Theduration of each stimulus element was 10 msec, with anexponential rise-decay time of 5.0 msec. The onset andoffset asynchronies were 8, 4, 2, 1, and 0 rnsec in fivedifferent experimental sessions-the last serving as acontrol determination where the performance would beexpected to be at the chance (50%) level. Thefrequencies of the two elements were 1,300 and1,500 Hz.

Both the monaural and dichotic results are presentedin Fig. 3, averaged across the three Ss, The elementasynchrony is plotted on the abscissa. The ordinaterepresents the percentage of trials in which the Ss gave

coY I

I x

LU

JlII

JlII

x ILCJ

4O...L--r---.---.----,--,--,--.....--,--,

Fig.A. A comparison of dichoptic and monocularmicropattern discrimination for red and green (Fig. 4A) and forblue and yellow (Fig. 4B). As indicated by the insertedschematics, the onset and offset asynchronies were always equalTheir absolute values are indicated on the abscissae.The percentdiscriminated is plotted on the ordinates.

the expected response, i.e., reporting that the pitch ofthe trailing element dominated the pitch of themicropattern. It can be seen in Fig. 3 that the Ssconsistently gave the expected response except wherethe asynchronies were small and consequently theperformance was near the chance level.

The most significant aspect of the results seen inFig. 3 is the demonstration that dichotic micropatterndiscrimination is possible. All of the dichotic nonzeroasynchrony points are significantly different from thezero asynchrony point (p < .05 for 1, 2, and 4 msec,p < .005 for 8 msec; paired values t, paired by Ss). As inthe monaural case, dichotic micropattern discriminationimproves as asynchrony increases. Presumably, largerasynchronies would lead to even better dichoticperformance. However, asynchronies exceeding 8 msecmight enable the S to judge the temporal order of thestimulus elements within the micropattern (Hirsh &Sherrick, 1961) and were not employed for this reason.

It is also clear from the results of Fig. 3 that monaural

micropattern discrimination is superior to the dichoticfor all nonzero asynchronies. This difference betweenthe monaural and dichotic values for the nonzeroasynchronies is significant at the .0005 level (pairedvalues t test with values paired for Ss and asynchronies).

In summary, dichotic micropattern discrimination ispossible throughout the range of asynchronies from I to8 msec. Throughout this range, dichotic performance isinferior to monaural performance.

Experiment IB-MonocuIar and DichopticMicropattem Discrimination

The temporal configuration of the chromaticmicropatterns employed in this experiment areillustrated in the schematic diagram of Fig. 2A. Theduration of each stimulus element was 10 msec, with arise-decay time of less than 0.4 msec. The asynchroniesbetween the two elements were identical to thoseemployed in Experiment lA.

Although all three Ss could perform monocularmicropattern discriminations with both the red-greenand blue-yellow color combinations, only one S couldachieve sufficiently stable binocular color mixing toperform these same discriminations in the dichopticmode. Therefore, only the data from this one S will bepresented here in detail. The results for this S for thered-green and blue-yellow color combinations are givenin Figs. 4A and 4B, respectively. The elementasynchrony is plotted on the abscissae. The ordinatesrepresent the percentage of trials in which the S gave theexpected response, indicating that the color of thetrailing element dominated the color of themicropattern. Again, as in the auditory case, the Salways gave the expected report except for the smallestasynchronies where he was near chance.

Dichoptic micropattern discrimination was obtainedfor both color combinations (note the open circles ofFigs. 4A and 4B). To test the significance of theseresults, the 100 trial sessions were divided into 4subsessions of 25 trials each; then t statistics werecomputed on these subsession results. For the case ofthe red and green elements, Fig.4A, the 8-, 4-, and2-msec asynchrony discriminations are significantlybetter than O-msec asynchrony (p < .0005, .0005, and.05, respectively). For the blue-yellow case, Fig. 4B, the8- and 2-msec asynchrony discriminations aresignificantly better than O-msec asynchrony (p < .005and .05) and the 4-msec values just miss significance atthe .05 level.

For both color combinations, it is apparent that themonocular performance is superior to the dichoptic. Ofall the nonzero asynchronies, only the results for 8-msecred-green and l-msec blue-yellow fail to yield asignificant monocular superiority. This statistical resultat these two points is to be expected, since bothmonocular and dichoptic performances are essentiallyperfect for the first and chance for the second.

A comparison of Figs. 4A and 4B also demonstrates

•6

ASYNCHRONY - M1H.

e

386 YUNDAND EFRON

100__X

»>: XX

90

+-- dichoptic

80

70 l'<--X ~ 11 @bO ~

l J!ed] 11 IIWdJR 8 [ 8x

so

400 4 6

DICHOPTIC AND DICHOTIC DISCRIMINATION 387

Fig. 5. (A) A comparison of monaural micropattemdiscriminations employing only onset or offset asynchronies.(B) A comparison of dichotic micropattern discriminationsemploying only onset or offset asynchronies. For both Figs. SAand S8, the abscissae represent the absolute values of the onsetor offset asynchronies. The ordinates represent the percentdiscriminated. The inserted schematics illustrate the temporalconfigurations.

6432o

so

70

80

100

~

,11)----------------~

~~

90 8 ~

~

~

~~

e:e ,:< 80 /.>;;!;:£it

70 IoooJ ESOOJl>(

'/ l

IIis , x-x

l§!I Il L:!iiiJ", R,

60 j/ loooI [i5OO]E>- --<il II

I.@ [l lii@so

60

100

understanding of the processes involved in micropatterndiscrimination.

What is important about these results, however, is thatthe two curves are different. Stimulus pairs withasynchronous element offsets were more readilydiscriminated than were pairs where the sameasynchrony occurred at the onset (p < .0005, pairedvalues t test with values paired for Ss and asynchroniesusing all nonzero asynchronies). Some factor other thanthe energy difference between the stimulus elements-adifference which is identical for each asynchrony forthese two curves-must account for this differencebetween these curves. The only possible factor is the finetemporal structure of the micropatterns which makesthe discrimination easier to perform in the secondexperiment, where the element offsets wereasynchronous, as was reported by Efron (1973). Ifdichotic micropattern discrimination is achieved bymeans of the same mechanisms as the monaural

that changing the colors of the stimulus elements fromred-green (Fig. 4A) to blue-yellow (Fig. 4B) has a similareffect upon the monocular and dichoptic discriminationfunctions. The monocular red-green monotonic functionrises more steeply than the monocular blue-yellow. Thissame difference between red-green and blue-yellowfunctions occurred for the other two Ss who wereunable to participate in the dichoptic experiments. Inlike manner, the dichoptic red-green curve rises moresteeply than the dichoptic blue-yellow. Thus, for both

.the monocular and dichoptic modes, the discriminationimproves more rapidly as asynchrony increases when redand green elements are used than when blue and yellowelements made up the micropatterns. Whatever may bethe cause of the superior red-green performancecompared to the blue-yellow, the fact that themonocular and dichoptic results both show this effect isan important finding.

Experiment 2A-Effects of On- and Off-Asynchronyin the Monaural and Dichotic Modes

The temporal configurations of the stimuli employedin this experiment are illustrated in Figs. 28 and 2C.There are two major parts to the experiment. In thefirst, the ability to discriminate monaural micropatternswith elements having asynchronous onsets (Fig. 2A) iscompared with that for micropatterns with elementshaving asynchronous offsets (Fig. 28). This part of theexperiment is a more extensive repetition (usingdifferent stimulus parameters and psychophysicalprocedures) of an experiment reported previously byEfron (1973). In the second part of the experiment, thesame stimuli were presented in the dichotic mode. Forall micropatterns in this experiment, the shorter stimuluselement was 10 msec, with the same 5.0-msecexponential rise and decay times as used previously. Thelonger element was 10, 11, 12, 14, or 18 msec (with thesame rise-decay times). These durations for the longerelement yielded asynchronies of 0, I, 2, 4, and 8 msec.The frequencies of the tone bursts were 1,300 and1,500 Hz.

The results for the monaural part of the experimentare presented in Fig. SA. The average percentdiscriminated for the three Ss is plotted as a function ofthe element asynchrony. The results for both types ofstimuli, those with asynchronous element onsets (crossesand solid line) and those with asynchronous elementoffsets (circles and dashed line), indicate thatdiscrimination performance was a monotonicallyincreasing function of the element asynchrony. Suchmonotonic functions are to be expected in thisexperiment, since one of the stimulus elements is longerand thus has greater energy in the first micropattern,while the other stimulus element is longer and has agreater energy in the second micropattern. As theasynchrony is increased, a monotonic increasingfunction is inevitable. Thus, either curve, in Fig. SA,taken by itself, is not of great interest with respect to an

40 .......-.-----...,..----,----r----...,

Experiment 2B-Effect of On and Off-Asynchrony inthe Monocular and Dichoptic Modes

The temporal configurations employed in the visual

experiments of this series were the same as those in theauditory experiments (2A) and are illustrated in Figs. 2Band 2e. The red-green monocular results for the S whowas able to perform dichoptically are shown in Fig. 6A.Since the visual system is apparently very sensitive torelatively small energy differences in the micropatterns,it was necessary to expand the scale on the abscissa andto include two values of asynchrony shorter than I msec(0.5 and 0.3 msec) to demonstrate the differencebetween on and off asynchronies. Micropatterns withoffset asynchronies were easier to discriminate than werethose with onset asynchronies. The difference indiscrimination performance reaches significance(p < .05) at 0.5-msec asynchrony.

The dichoptic results for these stimuli are given inFig. 6B. The dichoptic performance is degraded, butotherwise similar to the monocular. Off-asynchronyperformance was significantly (p < .05) better thanon-asynchrony performance at 1.0 msec.

In sum, in both the auditory and visual systems,discrimination performance is superior when an elementoffset asynchrony is employed rather than an onsetasynchrony. The same characteristic applies to thedichotic (dichoptic) as well as to the monaural(monocular) mode.

Experiment 3In Experiment IB, it was shown that changing the

colors of the stimulus elements from red and green toblue and yellow had the same effects on both monocularand dichoptic micropattern discrimination. An analagousexperiment was performed in audition to demonstratethat monaural and dichotic performance varies in thesame way with changes in tone frequencies.

It has been previously reported (Efron, 1973) that inthe monaural mode the perceptual dominance of thetrailing stimulus element increased as the difference infrequency (At) between the two elements of themicropattern increased. This experiment was repeated inthe monaural mode (using the present psychophysicalmethods and different stimulus parameters) and wasperformed in the dichotic mode for comparison. Thetemporal configuration seen in Fig. 2A was employedwith both stimulus elements having a duration of10 msec and an asynchrony of 8 msec. The frequenciesof Elements X and Y (see Fig.2A) were 1,600 and1,200 Hz, 1,500 and 1,300 Hz, 1,450 and 1,350 Hz,1,425 and 1,375 Hz, and 1,400 and 1,400 Hz in fivedifferent sessions. These values yield frequencydifferences of 400, 200, 100, 50, and 0 Hz,respectively-the last serving as a control-where chancelevels of discriminatory performance were expected if allother relevant cues had been eliminated.

The results of Experiment 3 are seen in Fig. 7, where,as in all previous experiments, the dichotic performanceis inferior to the monaural. However, the moresignificant aspect of these results is that both themonaural as well as the dichotic discrimination

II

3

ASYNCHRONY· m ...C

o

x~

~/ // ~---

~//

I/

/ l [bdJ lj [R:i"]I X-x • F.<MftI 1'1 tt;;..,;;t

If X '" ~ ~/

/f

ff

ff

fI•

so

60

70

90 B

100

388 YUNDANDEFRON

100

/1~=·~~90 A

eo

70 // x-x ~I x

E>--~~60 Iso

0

discrimination, the same discrepancy between the onsetand offset asynchrony performances would be expectedwhen such stimuli are delivered in the dichotic mode.

The results for the dichotic part of the experiment arepresented in Fig.5B and confirm this expectation.Again, the average percent discriminated is plotted as afunction of the element asynchrony. The similarity ofthese dichotic results to the monaural results of Fig. SAis obvious. Both onset and offset asynchronies producedmonotonically increasing functions, but, moreimportantly, the offset asynchronies lead to significantlyhigher average discrimination scores than were obtainedwith onset asynchronies (p < .0005, paired values t as inthe monaural case).

Fig. 6. (A) A comparison of monocular micropatterndiscriminations employing only onset or offset asynchronieswith red-green stimulus elements. (B) A comparison of dichopticmicropattern discriminations employi. only onset or offsetasynchronies with red-green stimulus elements. For both Figs.5A and 5B, the abscissae represent the absolute values of theonset or offset asynchronies. The ordinates represent the percentdiscriminated. The inserted schematics illustrate the temporalconfigurations.

DICHOPTIC AND DICHOTIC DISCRIMINATION 389

Fig. 7. A comparison of monaural and dichotic micropatterndiscriminations as a function of the difference in frequency ofthe two stimulus elements. The inserted schematics illustrate thetemporal configurations employed. The onset and offsetasynchronies were always equal (8 msec). The abscissa indicatesthe difference in frequency between f I and f2' The centerfrequency was 1,400 liz.

effect produced by the offset asynchrony operates onmonaural and dichotic micropatterns in the same way.This same parallelism is observed in the monocular anddichoptic presentations of Figs. 6A and 6B. Finally, theimprovement in discriminatory performance as Llf isincreased is observed for both monaural and dichoticmodes in Fig. 7. In sum, the fact that monaural(monocular) and dichotic (dichoptic) discriminationperformance varies in the same way with changes intemporal parameters (Experiments 2A and 2B) and withchanges in frequency-hue parameters (Experiments 3and I B) strongly indicates that the perceptualdominance effect seen in the two modes of presentationis essentially the same phenomenon.

If the two dominance effects are of the same type,then an explanation must be sought for the inferiorperformance levels in the dichotic (dichoptic] mode. Asmentioned earlier, this inferior performance could comeabout either because the dichotic (dichoptic) dominanceeffect is less powerful or because of some interferingfactor present in the dichotic (dichoptic) mode whichreduces efficiency of discrimination.

Perceptual interference between two different stimulidelivered to the two ears or eyes is already known undercircumstances similar to the one of the dichotic(dichoptic) micropattem experiments. In the auditorymodality, dichotic listening experiments (Broadbent,1954; Kimura, 1961, 1967; and Lowe et aI, 1970,among others) have demonstrated a significantinterference with digit, word, or nonsense syllablerecognition presented to the left ear (of right-handed Ss)when different words or nonsense syllables aresimultaneously presented to the right ear. More directlyrelated to the experiments reported here are the resultsof Efron and Yund (1974), showing an ear advantage forthe pitch of pure tone stimuli presented dichotically.(The three Ss employed for the present dichotic

performance increase as the frequency differencebetween the two elements of the micropattem increases.

DISCUSSION

Previous experiments using monaural and monocularpresentations have shown that micropatterns can bediscriminated even when the asynchrony between thetwo stimulus elements is as small as 2 msec. In all of themonaural and monocular experiments, thediscriminations were achieved as a consequence of theperceptual dominance of the trailing stimulus element i~each micropattern (Efron, 1973). In the presentmonocular and monaural experiments, performed withdifferent stimulus parameters and with a differentpsychophysical method, the same findings wereobtained. The results of the present experiments showthat a perceptual dominance of the trailing stimuluselement of the micropattern also occurs when theelements of the micropattern are presented to oppositeears or eyes in the dichotic or dichoptic modes.

Is the perceptual dominance effect observed in thedichotic (dichoptic) presentations the samephenomenonas that produced by monaural (monocular) presentationsor is it a consequence of two essentially differentprocesses having only a superficial resemblance?

Despite the fact that the Ss reported that the twodominance effects were indistinguishable, theirperformance in the dichotic mode was always inferior totheir performance monaurally. The one S who couldperform the dichoptic experiment also had an inferiorlevel of performance dichoptically compared to themonocular performance. This difference in performancecan be interpreted in either of two ways. On the onehand, it could be an indication that the two perceptualdominance effects are essentially different in the twomodes of presentation. On the other hand, theperformance differences could simply be an indicationthat the same phenomenon in the dichotic (dichoptic)mode is less powerful or that there is some interferingfactor which is present in the dichotic (dichoptic) modewhich reduces the apparent strength of the dominanceeffect.

A reexamination of the results of these experiments(Figs. 4A, 4B, SA, 5B, 6A, 6B, and 7) strongly supportsthe view that the two dominance effects are identical innature but different only in degree. In Figs. 4A and 4B,the monocular red-green monotonic function rises moresteeply than the monocular blue-yellow function. In thedichoptic experiments, a parallel result wasobtained-suggesting that the same factors which madethe red-green micropatterns easier to discriminatemonocularly were also operating dichoptically. InFig. SA, the monaural micropatterns were easier todiscriminate when the elements had asynchronousoffsets than when they had asynchronous onsets. In thedichotic experiments (Fig.5B), a parallel result wasobtained-indicating that the more powerful dominance

IIt

x

o

_____-~----x/

10 100 200

OIIHUHCI IN 'tlOIJINCY I"" I

390 YUND AND EFRON

experiments are now known to have a strong earadvantage. See Ss M.B., M.C., and B.Y. in Efron & Yund,1974.)

In the paradigm used for the present experiments, thesame frequency element was delivered to one ear in bothmicropatterns of a trial. This procedure, it will berecalled, was adopted so that any ear asymmetry did notmake the dominant pitch of the two micropatterns ofeach trial discriminably different when there was noelement asynchrony. It must be stressed that thisprocedure did not eliminate the ear asymmetry-itmerely ensured that any existing asymmetry affectedboth micropatterns of each trial equally. The earasymmetry, which nonetheless existed in each trial,would tend to degrade the salience of the perceptualdominance of the trailing stimulus element in alldichotic experiments. The reason for this can be seen byan inspection of Fig. lC. In a S having a left-earadvantage, the 1,500-Hz tone would sound louder thanthe 1,300-Hz tone in both micropatterns-thus makingthem more similar, i.e., harder to discriminate. Theperceptual dominance of the trailing stimulus element,however, has an opposite effect, making the twomicropatterns more dissimilar and more discriminable.The trailing 1,300-Hz tone in the first micropattern ofFig. IC will perceptually dominate the 1,500-Hz tone;the trailing 1,500-Hz tone in the second micropatternwill dominate the 1,300-Hz tone. Thus, any earadvantage will act to degrade the dichoptic micropatterndiscrimination performance and will make the differencebetween the monaural and dichotic performance greaterthan might be the case if no ear advantage were present.Thus, the fact that dichotic micropattern discriminationperformance is inferior to the monaural performance(Figs. 3, 5, and 7) might be merely a consequence of thestrong ear advantage effect which would tend to makemicropattern discrimination more difficult.

In the visual system, the best known example ofdichoptic competition is binocular rivalry. Also wellknown are the related difficulties associated withdichoptic color mixing; indeed, many Ss (the mostnotable being von Helmholtz) cannot perform this taskat all. Of the three Ss studied, only one could achieveadequate color mixing to discriminate dichopticmicropatterns. Given the strength of binocular rivalryand the difficulty of dichoptic color mixing, it is perhapssurprising that any dichoptic micropatterndiscrimination was possible. The fact that dichopticmicropattern discrimination performance is inferior to

the monocular performance (Figs. 4 and 6) might bemerely a consequence of binocular rivalry-aphenomenon which would strongly interfere withwhatever perceptual dominance was imparted by thetrailing stimulus element. Binocular rivalry is analogous,in this sense, to a changing (unstable) ear-advantageeffect, and would degrade dichoptic micropatterndiscrimination performance in the same way that an earadvantage would degrade dichotic performance.

In conclusion, the dominance of the trailing stimuluselement may actually be as powerful in the dichotic(dichoptic) mode as in the monaural (monocular)mode-the difference in performance characteristics inthe present experiments being merely due to theinterfering effects of ear advantage and binocular rivalry.

The phenomenon of perceptual dominance of thetrailing element of micropatterns and the associateddegradation of the perception of the first element (asdescribed in Efron, 1973) remains unexplained. Thepresent results, however, indicate that a peripheralinteraction between the two stimulus elements in theretina or basilar membrane is not essential to producethe perceptual dominance of the trailing element whichunderlies the capacity for micropattern discrimination.The ability of Ss to perform dichotic and dichopticmicropattern discrimination in these experimentsindicates the existence of an important central process inthis perceptual task.

REFERENCES,Broadbent, D. E. The role of auditory localization in attention

and memory span. Journal of Experimental Psychology, 1954,47,191-196.

Efron, R. Conservation of temporal information bY perceptualSystems. Perception & Psychophysics, 1973. 14. 518-530.

Efron, R., & Yund, E. W. Dichotic competition of simultaneoustone bUl'$ts of different frequencies: I. Dissociation of pitchfrom lateralization and loudness. Neuropsycbologia, 1974, 12,in prellS.

Hirsh, I. J., & Sherrick, C. E. Perceived order in different sensemodalities. Journal of Experimental Psychology, 1961, 62,423-432.

Kimura, D. Some effects of temporal lobe damage on auditoryperception. Canadian Journal of Psychology, 1961, 15,156-165.

Kimura, D. Functional asymmetry of the brain in dichoticlistening. Cortex, 1967,3,163-178.

Lowe, S. S., Cullen, J. K., Berlin, C. I., Thompson, C. L., &Willett, M. E. Perception of simultaneous dichotic andmonette monosyllables. Journal of Speech & HearingResearch, 1970, 13,812-822.

Thomas, F. H., Dimmick, F. L., & Luria, S. M. A studv ofbinocular color mixture. Vision Research, 1961, I, 108-120.

(Received for publication July 23. 1973;revision received December 2, 1973.)