Does so-called interhemispheric transfer time depend on attention? Does so-called interhemispheric transfer time depend on attention

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Does so-called interhemispheric transfer time depend on attention?Braun, Claude M J;Daigneault, Sylvie;Dufresne, Annie;Miljours, Sylvain;Collin, IsabelleThe American Journal of Psychology; Winter 1995; 108, 4; ProQuestpg. 527

Does so-called interhemispheric transfer time depend on attention?

C L AUDE M. J. BRAUN, SYLV IE DAIGNEAULT, ANNIE DUFRESNE, SYLVAIN MILJOCRS, and ISABELLE COL LIN Universite du Quebec a Montreal

To our knowledge none of many past attempts to experimentally modulate the crossed-uncrossed differential (CCD) or so-called "interhemispheric transfer time" derived from appropriate visual reaction time experiments has ever succeeded. The present 4 experiments were designed to establish that (a) under normal attentional constraints, significant CUDs would be obtained, but that (b) lateral mobilization of attention by probabilistic (valid and invalid) cuing as to target location would significantly and systematically alter the CUDs. In a baseline experiment, a significant CCD was obtained. In Experiments 2-4, CUDs were rendered nonsignificant by probabilistic cuing. Specific experimental conditions generally did not significantly influence the CUD in a systematic direction. In only one of many analyses of those results did an experimental effect on the CUD reach significance: In Experiment 3 intrasubject mean reaction times yielded a significant complex dissociation of CUDs as a function of type of cuing (valid or invalid) and stimulus onset asynchrony. It was concluded that the CUD can be significantly modulated by target location cuing, but only under very specific conditions. The direction of the trends in all the experiments, and in the significant finding, suggests that the CUD component that is most markedly influenced by such cues is a postcallosal, motor component.

Poffenberger's proposal ( 1912) that simple reaction time (SRT) in the uncrossed condition (a response by the hand ipsilateral to an eccentric visual stimulus) subtracted from the SRT in the crossed condition (a response by the hand contralateral to the stimulus) is an index of interhemispheric transfer time (IHTT) is generally accepted today (Bashore, 198 1; Milner, 1986). It is widely believed that the crosseduncrossed difference (CUD) principally represents the relay time difference between a "long" pathway requiring interhemispheric transfer (IHT) and a "short" pathway not requiring IHT (Jeeves, 1969). Rizzolatti (1979) expressed this point of view in its most radical form by stating that the CUD is nothing but a pathway-length effect. He predicted that experimental (cognitive) manipulation would never

AMERICAN JOURNAL OF PSYCHOLOGY Winter 1995, Vol. 108, l'>o. 4, pp. 52i-546 © l 995 by the Board of Trustees of the Univers1ty of Illinois


528 BRAUN ET AL.

significantly modulate the CUD. To our knowledge, all who have attempted to falsify this prediction using visual reaction time have so far failed. Three general approaches have been followed: (a) cognitive interference induced by a second task added to a simple detection paradigm (Milner, Jeeves, Ratcliff, & Cunnison, 1982; Rizzolatti, Bertoloni, & Buchtel, 1979; Rizzolatti, Bertoloni, & De Bastiani, 1982); (b) variation of stimulus complexity in go/no-go matching paradigms (Umilta, Frost, & Hyman, 1972; Umilta, Rizzolatti, Anzola, Luppino, & Porro, 1985); and (c) variation of stimulus-set size in go/no-go variants of the S. Sternberg task (Braun, Dumas, & Collin, 1994).

Despite these numerous findings supporting Rizzolatti's (1979) contention, we nevertheless believed the model implausible. Indeed, commissural fibers, numbering in the many millions (Aboitiz, Scheibel, Fisher, & Zaidel, 1992) are connected primarily to neurons in associative cortex known to play key roles in cognitive operations. For example, it is now known that parietal cortex play s a critical role in certain aspects of lateral mobilization of attention by probabilistic indicants serving as target-location cues (Baynes, Holtzman, & Volpe, 1986; Farah, Wong, Monheit, & Morrow, 1989; Posner, Walker, Friedrich, & Rafal, 1987). It appeared likely to us that powerful manipulation of attention by such probabilistic cues, affecting the activity of large numbers of cortical neurons, ought to influence input to, and output from, the commissure(s), particularly the corpus callosum. Specifically, we expected that invalid cuing ought to increase the CUD and that valid cuing ought to decrease the CUD. We expected this effect because we assumed that the key component in the effect of invalid cuing on the CUD would consist of a dampening (probably inhibitory ) influence on neurons serving as input to the corpus callosum. Along the same line of reasoning we assumed that the effect of valid cuing on CUD would consist of an enhancing (probably excitatory) influence on neurons serving as input to the corpus callosum. Supporting evidence that valid and invalid cuing exert excitatory and inhibitory influences, respectively, in the cued hemisphere has been obtained from single-unit recording experiments in primates (Bushnell, Goldberg, & Robinson, 1981; Goldberg & Bruce, 1985; Wurtz, Goldberg, & Robinson, 1980).

This series of experiments was therefore designed (a) generally, to falsify Rizzolatti's 1979 prediction, and (b) specifically, to demonstrate that valid cues reduce the CUD and that invalid cues increase the CUD. To this end, we first planned a baseline experiment designed to assure that a reliable CUD could be obtained in an uncued SRT paradigm. We then designed three experiments using the same par-


INTERHEMISPHERIC RELAY AND A TTEJ\;TION 529

adigm and equipment, modified only to test variations on the general theme of effects of valid and invalid cuing on the CUD.

EXPERIMENT 1

METHOD

Participants

Fifteen male and I5 female right-handed university students reporting no neurological, psychiatric, or substance abuse problems were paid to participate. Mean ages of the men and women were 25 (SD = 5.5) and 24 years (SD = 5.I), and mean education was I 6 (SD = 2.7) and I6 years (SD = 1.9), respectively. The sexes did not differ significantly with regard to age or education.

Procedure

A simple reaction time paradigm was implemented on an Apple-lie computer with an Applied Engineering timing card and an Amdek video monitor. Stimuli consisted of I-cm2 (2.5 arc degrees) white squares (636 cd/m2 luminance) on a black background (20 cd/m2). These measures were obtained with a Quantum Instruments Photo-Meter-I fitted with a PM-IO fiber-optic probe designed for directly measuring brightness of small areas of video screens. The experiment was conducted in a normally lit room (807 lx). This measure was obtained using another probe designed for capturing illuminance from all angles within a I80° hemisphere. All stimuli were presented to the right or left of center screen at II arc degrees (4.5 em) of eccentricity (defined in terms of center to inner edge of stimulus). Each participant's forehead rested against a stopper placed 50 em from the screen and eyes were at center screen level. A red light-emitting diode (LED) that was "on" throughout the entire experiment was placed at center screen and served as a fixation point.

The three conditions of stimulus duration were 20 ms, 80 ms, and I 40 ms ( ± a I 6-ms variation intrinsic to computer-generated video-display stimuli). The experimental routine proceeded as follows. After the experimenter provided standard instructions, the computer presented a series of 40 trials followed by a 2-min rest. The participant was required to immediately press the middle of the space bar of the keyboard with an index finger as soon as the stimulus was detected. The responding finger was 25 em in front of the participant's body midline, and the subject's arm and hand rested on a table. Each participant responded with the same hand throughout each set of 40 trials. Immediately after the response, the reaction time (RT) was displayed at the bottom of the screen (500 ms) or a message indicated that an excessive delay (>I ,000 ms) or an anticipation (<75 ms) had resulted in cancellation and a replacement, to come, of the trial. Each cancelled trial was replaced by another trial, resulting in 40 valid trials in each set. Feedback was presented to "shape out" response biases (anticipations, delays) and to


530 BRACN ET AL.

motivate the participants to maintain attention by attempting to improve their response speeds. After the feedback disappeared, an interval of 1,050 ms, 1, 150 ms, 1 ,300 ms, or 1 ,500 ms of black screen occurred. Each such interval was equiprobable and randomly distributed. Participants were instructed to look at the feedback and then immediately fixate the LED. Hemilocation of stimuli was equiprobable and pseudorandom. There was a maximum of three repetitions in a hemifield.

Seven women and 8 men started with the left hand, and 8 women and 7 men started with the right hand. On each subsequent set of 40 trials, the responding hand was alternated. Ten participants started with the 20-ms stimulus-duration condition, 10 with 80 ms, and 10 with 140 ms. Each participant completed four sets of 40 trials consecutively in each stimulusduration condition. Sequences of stimulus-duration conditions were counterbalanced. In all, each participant produced 480 "error-free" RTs (>75 ms, <1,000 ms).

The computer program tabulated R Ts into each cell of the design and computed medians. In addition, both types of errors (anticipations and delays) were tabulated by frequency in each cell of the design.

RESULTS

Errors

There were 960 anticipation errors ( <7 5 ms), amounting to 6.7% of correct responses . As expected, these were totally unrelated to hand, field, stimulus duration, or sex.

Though errors of excessive response delay (> 1 ,000 ms) were much less frequent (N = 82, i.e., 0.6% of correct responses), they manifested an interesting distribution and comprised several statistically significant effects as determined by chi-square analysis. The effects of stimulus duration on these delay errors consisted of fewer errors in the 80-ms stimulus-duration condition (N = 16) than in the 20-ms condition (N = 42), X2(1, N = 58) = 16, p < .001, and fewer errors in the 140-ms condition (N = 24) than in the 20-ms condition, x2(1, N = 66) = 4.9, p < .05. In addition, there were significantly fewer errors resulting from stimuli presented in the right field (N = 31) than in the left (N = 51), x2(1, N = 82) = 4.8, p < .05. Finally, a negative CUD observed at the left hand reached significance, x2(1, N = 48) = 3.9, p < .05. No other effect surpassed the alpha criterion of p = .05.

Reaction times

The following analyses are based exclusively on means of "correct" median RTs (>75 ms, <1,000 ms). A general 3 X 2 X 2 X 2 repeated measures analysis of variance (ANOVA) included the following within


INTERHEMISPHERIC RELAY Al\D A TTENTIO:'I/ 531

factors: stimulus duration, hand, and field. Sex was the sole between factor.

A main effect of duration, F(2, 28) = 6.6, p = .004, consisted of prolongation of RT as a function of decreased or increased stimulus duration from 80-ms targets. Repeated measures trend analysis revealed absence of a linear effect on the distribution of the RTs with increasing target duration (t = 1.4, p = .16) and presence of a quadratic distribution (t = 3.1, p = .004). This finding is in perfect agreement with the literature (Kaswan & Young, 1965). A main effect of field, F(l, 29) = 13.7, p = .0009, consisted of a right field advantage. The Hand X Field interaction, F(l, 29)= 4.7, p = .038, consisted of a significant positive overall CUD of 1.67 ms. A Hand X Sex interaction, F(l, 28) = 5.8, p = .023, consisted of a right-hand advantage for men and a left-hand advantage for women. The Field X Hand X Stimulus duration interaction did not approach significance, F(2 , 27) = 80, p = .64. The overall CUDs in the 20-, 80-, and 140-ms duration conditions were 0.8, 1.3, and 2.9 ms, and they all reached statistical significance individually. Repeated measures trend analysis revealed absence of a linear progression of CUDs as a function of increasing target duration, t(28) = .90, p = .38. There was also absence of a quadratic function, t(28) = .38, p = . 71. These negative findings highlight the fact that coefficients of variation of CUDs are much higher (9.84, 4.87, 3.01) than those of RTs (0.12, 0.12, 0.13), though this difference is exaggerated by the presence of negative CUDs.

DISCUSSION

Experiment I establishes that the CUD obtained with this design implemented on this equipment is reliable. It is also important, considering the experiments to be reported next, to underscore the fact that target exposure duration did not have a reliable impact on the CUD. However, because there appeared to be a continuously increasing prolongation of the CUD as a function of stimulus duration, we decided to include the stimulus-duration parameter in the next experiment which was designed to test the effects of valid and invalid target location cuing on the CUD.

EXPERIMENT 2

METHOD

Participants

Eleven male and l 0 female right-handed university students, none of whom had participated in Experiment l, were paid to participate. They reported


532 BRAUN ET AL.

having no neurological, psychiatric, or substance abuse problems. Mean ages of the men and women were 24 (SD = 5.7) and 23 years (SD = 2.0), and mean education was 16 (SD = 2.8) and 16 years (SD = 1. 1), respectively. The sexes did not differ significantly with regard to age or education.

Procedure

The equipment and experimental routine were identical to Experiment 1, except for the following: There was no LED; prior to each stimulus onset, a horizontal white arrow (2 em long X 1 em high), with an equiprobable pseudorandomly distributed arrow-onset to target stimulus-onset interval (stimulus onset asynchrony, SOA) of 1,050 ms, 1,150 ms, 1,300 ms, or 1,500 ms, was presented at center screen for 1 ,000 ms; the arrow served as a fixation point and as a probabilistic cue; the arrow pointed in the direction of the target on 80% of the trials and in the direction opposite to the target on 20% of the trials; the participant was told that "the arrow would usually point in the direction of the subsequent target"; there were two stimulusduration conditions (20 ms and 80 ms); the participant was told "not to try to anticipate the target but to respond as quickly as possible to the onset of the target"; 6 men and 5 women started with the left hand, and 5 men and 5 women started with the right hand; each participant completed 10 sets of 40 trials. The responding hand and stimulus-duration conditions were counterbalanced as in Experiment 1.

RESULTS

Errors

There were 97 antiCipation errors (RT <75 ms), amounting to 1. 15% of correct responses. As expected, these were totally unrelated to hand, field, stimulus duration, cue, or sex. However, SOA did have an effect on rate of anticipation errors. The longer the SOA, the more anticipation errors were made, x2(3, N = 97) = 13.2, p < .001. Errors of excessive delay (RT > 1 ,000 ms) were slightly less frequent (N = 89, i.e., 1% of correct responses). They manifested an interesting distribution and comprised several statistically significant effects as determined by chi-square analysis. The invalidly cued targets elicited a higher percentage of errors than the validly cued targets, x2( 1, N = 89) = 11.5, p < .00 1. The 80-ms stimulus-duration condition produced fewer errors than the 20-ms stimulus-duration condition, x2(1, N = 89) = 85, p < .0 1. Both validly and invalidly cued targets yielded a clear positive-error CUD for the right hand, and both resulted in a negative CUD for the left hand.

Reaction times

The following analyses are based exclusively on means of median RTs, excluding anticipatory (<75 ms) and delayed (>1,000 ms) re-


INTERHEMISPHERIC RELAY AND ATTENTION 533

sponses. A general 2 X 2 X 2 X 2 X 2 X 2 repeated measures ANOV A

included the following within factors: stimulus duration, hand, field, SOA, and cue. Sex was the sole between factor.

A powerful main effect of cue, F(1, 20) = 81.4, p = .0001, consisted of a shortening of R T as a function of valid cuing and a lengthening of RT as a function of invalid cuing. All subjects manifested this effect. RTs decreased continuously and linearly as a function of increasing SOA, F(3, 18) = 206.8, p = .0001 (Figure 1). Stimulus duration exerted a significant influence only in interaction with field, F(1, 20) = 5.4, p = .03. At 20-ms stimulus duration there was a right field advantage, whereas a left field advantage occurred at the 80-ms stimulus duration. Finally, the only other significant effect was a Cue X Duration X Field interaction, F(3, 18) = 8.6, p = .008. The Field X Hand interaction did not approach significance, F( 1, 20) = .23, p = .64. Effects of prime theoretical interest were far from significant. All of the following interactions were very weak: Field X Hand X

Cue, F( 1, 20) = .42, p = .52; Field X Hand x Duration, F( 1, 20) = .002, p = .960; Field x Hand X SOA, F(3, 18) = .37, p = .78; Field x

Hand x Cue x SOA, F(3, 18) = .34, p = .80; Field x Hand x Cue X Duration, F(1, 20) = .06, p = .8 1; Field X Hand X SOA X Duration, F(3, 18) = .58, p = .63; and Field X Hand X Cue x SOA X Duration, F(3, 18) = .79, p = .52. The overall CUD for Experiment 2 was 1.48 ms. The overall CUD in the valid-cue condition was 1.91 ms, and -0.22 ms in the invalid-cue condition. The CUD's progression from the first to the last SOA was unstable (12, 10, 27, -10 ms). Furthermore, the distribution of CUDs as a function of SOA and of cue type was particularly erratic (see Figure 2). The overall CUD was 1.53 ms in the 20-ms stimulus-duration condition, and was 1.44 ms in the 80-ms stimulus-duration condition. Figure 1 presents RTs and Figure 2 presents CUDs obtained at each SOA in valid and invalid cuing conditions.

DISCUSSION

The results of this experiment agree perfectly with Rizzolatti's 1979 prediction. None of the experimental manipulations influenced the CUD in a manner remotely reliable. Furthermore, the direction of the trend was in the opposite direction to that predicted, with a negative (therefore "briefer") CUD occurring in the invalid than in the valid condition. We believed, in hindsight, that this experiment may have included too few subjects and trials and too many parameters. Therefore, we planned a similar experiment with more subjects and more trials and decreased the stimulus-duration parameters to one.


534

310 -300 -

VJ 290 -.s 280 Cll

-§ 270 r-c

.Q 260 -ti <ll 250 -Cll

c::: 240 f-230 f-

0� I

BRAUN ET AL.

o ••.• •• •• INVALID ·o •• -....... ...

• ···

o .... " VALID ··············o

·--·--------·

I I I I I I I I I 1100 1200 1300

SOA (ms) 1400 1500

Figure 1. Means of (intrasubject) median reaction times in Experiment 2 at each stimulus onset asynchrony (SOA) in valid and invalid cuing conditions

40 r---

30 r---

20 r-

VJ 10r-

V7ALIO e�p\

0 ...... ·�

.

·· .

.

�

... / \

. 0 �----�\T .. ----�.�·-------------·�· �.--------� .s

Cl

3 -10 r--20 r-

-30 r-

. . . \ ./INVALID \

0 ....... . . . -4

0;( I I I I I I I .? oy,��--��--��--�

11 00 1200 1300 1400 1500 SOA (ms)

Figure 2. Crossed-uncrossed difference (CUD) in reaction time between means of (intrasubject) median reaction times in Experiment 2 at each stimulus onset asynchrony (SOA) in valid and invalid cuing conditions

EXPERIMENT 3

METHOD

Participants Fourteen male and 15 female right-handed university students, none of

whom had participated in previous experiments, were paid to participate.



They reported not having neurological, psychiatric, or substance abuse problems. Mean ages of the men and women were 26 (SD = 5.03) and 26 years (SD = 5.4 7), and mean education was 15 (SD = 1.85) and 15 years (SD =

2.08), respectively. The two sexes did not differ significantly with regard to age or education.

Procedure

The equipment and experimental routine were identical to Experiment 2, except for the following: There was only one stimulus-duration condition (I 00 ms) instead of two; the participants completed 12 sets of 40 trials instead of 10.

RESULTS

Errors

There were 20 omission errors (RT <75 ms) amounting to 0.001 o/c of correct responses, and 181 anticipation errors (RT > 1,000 ms) amounting to 0.013% of correct responses. Longer SOA again tended to favor increasing rates of anticipation errors, x2(3, N = 181) = 88.3, p < .000 1. There were no other main effects or interactions in the error patterns.

Reaction times

The same effects as obtained in Experiment 2 occurred here. The main effects of cue, F(1, 28) = 104.4, p = .0001, and SOA, F(3, 26) = 30.0, p = .0001, were again highly significant. The hand and field main effects and the interaction were all far from significance, F(1, 28) < 1.7, p > .21. Figure 3 presents RTs and Figure 4 presents CUDs obtained at each SOA in valid and invalid cuing conditions. The general CUD was 1.4 ms in the valid condition and -3.4 ms in the invalid condition.

DISCUSSION

Rizzolatti's 1979 prediction was again supported here. We were not able to significantly influence the CUD by experimental manipulation of attention. As in Experiment 2, invalid cuing produced an overall negative CUD and valid cuing an overall positive one, contrary to the initial hypothesis.

Because the Field X Hand X Cue X SOA interaction approached significance in this experiment, F(1, 28) = 2.0, p = .14, a last attempt to experimentally modulate the CUD with cuing was planned. In the two previous experiments the most dramatic (though nonsignificant)


536 BRAUI\ ET AL.

330 �--------------------------------�

(j)

320 f--310 f--

.s 300 f--<1> ,g 290 f--

g 280 fu � 270 fa:

260 f-

o •• ····o. · . e'\ ······.

�.�VALID •••••••••••. 0 . . .

• �ALID

···o·····

-·-----· 250

o��'--'��--�1 11 00 1 200 1 300 1400 1 500

SOA (ms)


40

30

20

10 (j) .s 0

g -1 0 (.)

-20

-30

-40

-

-

f-- e VALID .............. __ 1------.=._=----�(') ·-

...... �·. ·············o

.. o··· f--

1-

1-

1-

•• ··INVALID o·

0� I I 1 100

I I I I I 1 200 1 300 1 400

I 1 500

SCA (ms)


effect of cuing on the CUD occurred at the earlier SOA (1,050 ms). We therefore planned an experiment with only two SOAs, one even earlier at 1,000 ms and one later at 2,000 ms, hoping to reach a significant dissociation of CUD by Cue at the early SOA and a complete extinction of the effect at the late SOA. To further increase our chances of obtaining such an effect, we once more increased the


INTERHEMISPHERIC RELAY AND ATTEJ\TION 537

number of trials. Finally, we introduced an electro-oculographic procedure to monitor participants' ey e movements and further motivate them to observe the central fixation directive.

EXPERIMENT 4

METHOD

Participants

Eight male and 8 female right-handed university students, none of whom had participated in previous experiments, were tested. They reported having no neurological, psychiatric, or substance abuse problems. Mean ages of the men and women were 23 (SD = 1.41) and 23 years (SD = 3.20), and mean education was 15 (SD = 1.06) and 15 years (SD = 1.51 ), respectively. The sexes did not differ significantly with regard to age or education.

Procedure

The equipment and experimental routine were identical to Experiment 3, except for the following: There were now only two SOAs, 1 ,000 and 2,000 ms; the subjects completed 16 sets of 40 trials instead of 12.

Eye movements

Baseline procedures with a vertical and horizontal 2-pair electrode montage allowed us to determine each participant's characteristic expression of an eye blink, and of a lateral 11 arc degree eye deviation. We found that none of the subjects manifested these during experimentation.

RESULTS

Errors

There were 3 7 omission errors (RT > 1 ,000 ms) amounting to 0.36% of correct responses, and 280 anticipation errors (RT <75 ms) amounting to 2.7%. There was a highly significant effect of SOA on anticipation errors, with subjects making 78 errors at the 1 ,000-ms SOA and 202 errors at the 2,000-ms SOA, x2(1, N = 280) = 54.9, p < .0001. Men made more anticipation errors (170) than women ( 11 0), and this difference was statistically significant, X2( 1, N = 280) = 12.9, p < .001. There were no other main effects or interactions in the error patterns.

Reaction times

The same effects as obtained in the two previous experiments occurred here. The main effect of cue, F(1, 15)= 59.5, p = .0001, was


538 370

360 r-

350 r-� g 340 r-Q)

� 330 f-c

320 � 0

u <1l Q) 310 -a:

300 -

290 -

0�

BRAUN ET AL.

o ... ••••••••••• INVALID ··-- ....... .... . ........ •••• 0

·-

VALID -·

l I I I I I I I I 1000 1250 1500

SOA (ms) 1750 2000


8 .-----------------------------------,

6r-

41-

2r- .. .... ···· .. .. .

. o

� VALID •• •• g or--- e----------��-----------• ---... 0 :::::> -2 r(.)

-4 1-

�6 1-

-8 I-

.. .. ·· ..

0 ••• ·····;NVALID

:{ I I I I I I 0

��--��--L-� 1000 1250 1500 1750 2000

SOA (ms)


again highly significant. However, the main effect of SOA, though still significant , was markedly dampened, F(l, 15) = 5.93, p = .029. The main effects and interaction involving hand and field were nonsignificant, F(1, 15) < 3.17, p > .097. Figure 5 shows RTs and Figure 6 shows CUDs obtained at each SOA in valid and invalid cuing con-



ditions. The general CUD in the valid condition was 0.5 ms and 0.3 ms in the invalid condition.

DISCUSSION

We again failed to falsify Rizzolatti's 1979 prediction. We were not able to experimentally modulate CUDs beyond the alpha criterion by means of attentional manipulation. Once again nonsignificant trends were in a direction opposite to those initially hypothesized. Invalid cuing at the early SOA produced a negative CUD, and valid cuing at the early SOA a positive one, contrary to our initial expectation.

Furthermore, rather than extinguishing the dissociation of CUD by cue type, the late SOA reversed the direction of the effect, though again not in a reliable (significant) manner. We now realize, in hindsight, that subjects did not become indifferent to the cue even at such a late SOA as 2,000 ms. In any case, it seems that the subjects redistributed their attention within the time frame imposed upon them, and realized very strongly and consciously that whenever a target did not occur immediately at offset of the arrow cue, a 2-s (black screen) interval would necessarily occur. So they were probably in an intense anticipatory state at the late SOA instead of in a relaxed state as we had imagined.

GENERAL DISCUSSION

We prefer to use intrasubject medians, as was reported in all the results sections, as input for A!'I.'OVAs because these are less affected than means by extreme scores, which in simple RT experiments tend to consist of self-corrected omission errors. However, to the extent that (a) experimental effects may plausibly operate at the extremities of the RT distributions as in Experiments 2, 3, and 4 and (b) truly outlandish R Ts ( <7 5 or > I ,000 ms) are automatically excluded, as was the case here in all the experiments, it can be argued that intrasubject mean R Ts may legitimately serve as input to ANOV As. And this we did. All the effects found to be significant, using intrasubject medians, were also found significant using intrasubject means. Similarly, all the effects found to be nonsignificant, using intrasubjects medians, were nonsignificant using intrasubject means-with a single exception. In Experiment 3, the Field X Hand x Cue X SOA interaction reached significance, F(3, 26) = 3.67, p = .027: This is the single instance of this interaction reaching significance in the three last experiments, each involving several tests of attentional modulation


540 BRAUN ET AL.

of CUDs. Clearly, the norm is for the CUD to be unreliably influenced by cognitive variables in general, and probabilistic cuing of target location in particular. On the other hand, any attempt to experimentally manipulate a 1 to 3-ms effect in R T, the variance of which is at least 1 0 times greater than the effect manipulated, is bound to be extremely difficult to bring to significance. Therefore, one should not dismiss consideration of trends in such a context. Indeed, one trend in particular is remarkably resilient in the present series of experiments, and that is the production of a group-averaged negative CUD in all three relevant experiments in the invalid cue condition at brief SOA. This trend cannot be accounted for by the standard explanation of CUD as a "pathway-length effect." Obviously, the CUD may be more complicated than that.

We believed at the outset of this series of experiments that increased CUDs ought to result from invalid cuing. The logic of this was that attentional demobilization (mismatch) ought to deteriorate both general SRT and interhemispheric communication. Indeed, we thought the hemisphere receiving the invalidly cued target ought to be inhibited and ought therefore to send a degraded signal across the commissure(s) to the finger-response-emitting hemisphere (in the crossed condition). Prolonged RT is often associated with increased CUDs, though admittedly in choice RT (Bashore, 1981; Zaidel, 1 983). There is evidence that increasing stimulus eccentricity significantly increases the CUD in SRT (St. John, Shields, Krahn, & Timney, 1 987). Finally, manipulation of stimulus intensity significantly alters visual evoked potential CUDs (Lines, Rugg, & Milner, 1 984) . These latter three findings suggest that early visual processing is critical.

Consequently, we expected that invalid cuing ought to increase the CUD. However, the opposite pattern occurred in Experiments 2 , 3, and 4. A negative CUD always occurred in invalid trials, at least at brief SOA. How then might this apparent paradox be resolved? We propose that the interaction of cuing and SOA effects with the CUD critically depends upon processes occurring in the response-emitting hemisphere. More specifically, we propose that in a crossed situation, invalid cuing "primes," "activates," or "engages" the responding but not the receiving hemisphere in the short term, and that this activation is critical for unimanual SRT and therefore generates a negative CUD. According to the same logic, in the uncrossed condition, it is the receiving rather than the responding hemisphere that is primed, activated, or engaged by the invalid cue, resulting in a positive CUD. In the valid cue condition, the CUD would then logically be increased because the responding hemisphere is primed in the uncrossed condition and not in the crossed condition. Figures 2, 4, and 6 suggest


INTERHEMISPHERIC RELAY A:-.ID ATTE:-./TION 541

that this motor enhancement of the invalidly primed "postcommissural" hemisphere is of medium latency and of brief duration. It seems to build up to a maximum between the 1 ,050-ms and 1, 150-ms SOA (meaning, in this series of experiments, a critical interval of 50 to 150 ms), and it seems to decline to nothing by an SOA of 1,300 ms. It then is transformed into dysfacilitation (inhibition?) if the subject is enabled to be certain of a late and fixed SOA (2,000 ms) and therefore to anticipate the occurrence of the target after the cue. It also appears plausible to us that the use of a central arrow cue in the present series of experiments may have contributed more to this particular pattern of effects than peripheral cues would have. We will now review the independent evidence in support of this general proposal.

Evidence of a critical contribution of motor systems in interhemispheric communication has been convincingly presented for certain types of tasks. In particular, certain types of bimanual coordination have been clearly demonstrated to depend heavily upon motor processing in the supplementary motor area and to relay specifically through the anterior corpus callosum (Brinkman, 1984; Chan & Ross, 1988; McNabb, Carroll, & Mastaglia, 1988; Preilowski, 197 5 ). However, these tasks are very different from the one investigated in the present study. Some commentators who have considered the issue of whether the cuing effect critically depends on visual versus motor processing have opted for the latter (Jon ides, 1981; Remington & Pierce, 1984; Rizzolatti, Riggio, Dascola, & Umilta, 1987; Tassinari, Aglioti, Chalazzi, Marzi, & Berlucchi, 1987: but for counterevidence, see Klein, 1980; Mangun, Hansen, & Hillyard, 1987; Nagel-Leiby, Buchtel, & Welch, 1990). It is the oculomotor system which has been most clearly shown to be functionally linked to the cuing effect (Klein, 1978; Posner, 1980; Remington, 1980). However, motor systems controlling hand movements have also been shown to be linked to both the cuing effect and oculomotor systems (Buchtel & Butter, 1988; Mather & Fisk, 1985; May lor & Hockey, 1985; Posner & Cohen, 1984; Posner, Cohen, & Rafal, 1982; Possamai, 1991; Rafal, Posner, Friedman, Inhoff, & Bernstein, 1988; Verfaellie, Bowers, & Heilman, 1988a, 1988b), but this is more true of peripherally emplaced than centrally emplaced cues. Both location precues and neutral precues, and particularly variations of SOA associated with these, can in varying conditions, independently or in combination, strongly influence the speed of voluntary saccadic eye movements to lateral targets (Braun & Breitmeyer, 1988; Fischer & Boch, 1983; Fischer & Ramsperger, 1984; Mayfrank, Mobashery, Kimmig, & Fischer, 1986; Remington & Pierce, 1984). Furthermore, in callosotomized patients, the cuing effect seems


542 BRAUN ET AL.

to manifest a pattern similar to the one observed here, including negative CUDs at invalid trials (Holtzman, Volpe, & Gazzaniga, 1984). However, the slope of the SOA-RT relation is abnormal (precues were location neutral) in the split-brain monkey (Beaubaton & Requin, 1973), leading the authors to conclude that the motor contribution to the SOA effect is critical. So to sum up, although many indirect lines of evidence suggest that there is indeed a motor modulation by the response-emitting hemisphere of the relationships between cue, SOA, hand, and field, the present study, as far as we know, is the first to demonstrate a significant such four-way interaction in the normal human.

CONCLUSION

The evidence from this study and others suggests that CUDs are a net result of contributions of multiple and complex modular brain circuits located in various cortical and subcortical sites of each hemisphere, actively processing the visual environment and preparing unimanual responses to it, and communicating with each other via multiple intra- and interhemispheric pathways and commissures.

This conclusion is compatible with the findings of Posner and colleagues whose patients with focal parietal, thalamic, and midbrain lesions have deficits of specific components of the validity effect, namely the "disengage," "move," and "engage" components, respectively (Rafal & Posner, 1987). Detailed analysis of various other aspects of the paradigm suggests, however, that the brain microgenesis of cued RT may involve many more brain circuits than these three, as well as several other cognitive operations including "expectancy" subtending the SOA effect and segregation of the disengage, move, and engage components into sensory and motor modules.

Notes

This research was supported by a grant to Claude M. J. Braun by the Fonds Concerte d' Action et de Recherche granting agency of Quebec, and the Natural Science and Engineering Research Council granting agency of Canada. This article is dedicated to Justine Sergent, whose untimely death we much regret.

Correspondence concerning this article should be addressed to Claude M. J. Braun, Laboratoire de �eurosciences Cognitives, Departement de Psychologie, Universite du Quebec a \fontreal, Case Postale 8888, Succursale Centre-Ville, Montreal, Quebec, Canada H3C 3P8. Received for publication March 5, 1 992; revision received June 8, 1994.



References

Aboitiz, F., Scheibel, A. B., Fisher, R. S., & Zaidel, E. (1992). Fiber composition of the human corpus callosum. Brain Research, 598, 143-153.

Bashore, T. R. (1981). V isual and manual reaction time estimates of interhemispheric transmission time. Psychological Bulletin, 89, 352-368.

Baynes, K., Holtzman, J. D., & Volpe, B. T. (1986). Components of visual attention: Alterations in response pattern to visual stimuli following parietal lobe infarction. Brain, 109, 99-114.

Beaubaton, D., & Requin,J. (1973). The time course of preparatory processes in split brain monkeys performing a variable foreperiod reaction time task. Physiology and Behavior, 10, 725-730.

Braun, C. M. ]., Dumas, A., & Collin, I. (1994). Effects of memory load on interhemispheric relay. American journal of Psychology, 107, 537-549.

Braun, D., & Breitmeyer, B. G. (1988). Relationship between directed visual attention and saccadic reaction times. Experimental Brain Research, 73, 546-552.

Brinkman, C. (1984). Supplementary motor area of the monkey's cerebral cortex: Short and long term deficits after unilateral ablation and the effects of subsequent callosal section. Journal of Neuroscience, 4, 918-929.

Buchtel, H. A., & Butter, C. (1988). Spatial attentional shifts: Implications for the role of polysensory mechanisms. Neuropsychologia, 26, 499-509.

Bushnell, M. C., Goldberg, M. E., & Robinson, D. L. (1981). Behavioral enhancement of visual responses in monkey cerebral cortex: I. Modulation in posterior parietal cortex related to selective visual attention. journal of Neurophysiology, 46, 755-772.

Chan, J., & Ross, E. D. (1988). Left-handed mirror writing following right anterior cerebral artery infarction. Neurology, 38, 59-63.

Farah, M. J., Wong, A. B., Monheit, M.A., & Morrow, L.A. (1989). Parietal lobe mechanisms of spatial attention: Modality-specific or supramodal? Neuropsychologia, 27, 461-470.

Fischer, B., & Boch, R. (1983). Saccadic eye movements after extremely short reaction times in the monkey. Brain Research, 260, 2 1-26.

Fischer, B., & Ramsperger, E. (1984). Human express saccades: Extremely short reaction times of goal directed eye movement. Experimental Brain Research, 57, 191-195.

Goldberg, M. E., & Bruce, C.]. (1985). Cerebral cortical activity associated with the orientation of visual attention in the rhesus monkey. Vision Research, 25, 4 7 1-481.

Holtzman, ]. D., Volpe, B. T., & Gazzaniga, M.S. (1984). Spatial orientation following commissural section. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention (pp. 375-394). London: Academic Press.

Jeeves, M. A. (1969). A comparison of interhemispheric transmission times in acallosals and normals. Psychonomic Science, 16, 245-246.

Jonides,J. (1981). Voluntary versus automatic control over the mind's eye's


544 BRAUN ET AL.

movement. In ]. Long & A. Baddeley (Eds.), Attention and performance IX (pp. 187-203). Hillsdale, NJ: Erlbaum.

Kaswan, J., & Young, S. (1965). Effect of luminance, exposure duration and task complexity on reaction time. journal of Experimental Psychology, 69, 393-400.

Klein, R. M. ( 1 978). Chronometric analysis of saccadic eye movements: Reflexive and cognitive control. In D. Landers & R. Christina (Eds.), Psychology of motor behavior and sport (pp. 98-120). Champaign, IL: Human Kinetics.

Klein, R. M. (1980). Does oculomotor readiness mediate cognitive control of visual attention? In R. S. Nickerson (Ed.), Attention and performance VIII (pp. 259-276). Hillsdale, NJ: Erlbaum.

Lines, C. R., Rugg, M. D., & Milner, A. D. (1984). The effect of stimulus intensity on visual evoked potential estimates of interhemispheric transmission time. Experimental Brain Research, 57, 89-98.

Mangun, G. R., Hansen,J. C., & Hillyard, S. A. (1987). The spatial orienting of attention: Sensory facilitation or response bias? Current Trends in Evoked Potential Research (Suppl. 40), 118-124.

Mather, ]. A., & Fisk, J. D. (1985). Orienting to targets by looking and pointing: Parallels and interactions in ocular and manual performance. Quarterly Journal of Experimental Psychology, 37A, 315-338.

Mayfrank, L., Mobashery, M., Kimmig, H., & Fischer, B. (1986). The role of fixation and visual attention on the occurrence of express saccades in man. European Archives of Psychiatric and Neurological Science, 235, 276-281.

Maylor, E. A., & Hockey, R. (1985). Inhibitory component of externally controlled covert orienting in visual space. Journal of Experimental Psychology: Human Perception and Performance, 11, 777-787.

McNabb, A. W., Carroll, W. M., & Mastaglia, F. L. (1988). "Alien hand" and loss of bimanual coordination after dominant anterior cerebral artery territory infarction. journal of Neurology, Neurosurgery and Psychiatry, 51, 218-222.

Milner, A. D. (1986). Chronometric analysis in neuropsychology. Neuropsychologia, 24, 115-128.

Milner, A. D.,Jeeves, M.A., Ratcliff, P.J., & Cunnison,J. ( 1 982). Interference effects of verbal and spatial tasks on simple visual reaction time. Neuropsychologia, 20, 591-595.

Nagel-Leiby, S., Buchtel, H. A., & Welch, K. M.A. (1990). Cerebral control of directed visual attention and orienting saccades. Brain, 113, 23 7-276.

Poffenberger, A. T. (1912). Reaction time to retinal stimulation with special reference to the time lost in conduction through nerve centers. Archives of Psychology, 23, 1 -73.

Posner, M. I. (1980). Orienting of attention. Quarterly journal of Experimental Psychology, 32, 3-25.

Posner, M. 1., & Cohen, Y. (1984). Components of visual orienting. In H.



Bouma & D. Bouwhovis (Eds.), Attention and performance X (pp. 531-556). Hillsdale, NJ: Erlbaum.

Posner, M. I., Cohen, Y., & Rafal, R. D. (1982). Neural sy stems control of spatial orienting. Philosophical Transcripts of the Royal Society B, 298, 187-198.

Posner, M. I., Walker, J. A., Friedrich, F. A., & Rafal, R. D. (1987). How do the parietal lobes direct covert attention? Neuropsychologia, 25, 135-145.

Possamai, C. A. (1991). A responding hand effect in a simple RT precueing experiment: Evidence for a late locus of facilitation. Acta Psychologica, 77, 47-63

Preilowski, B. F. B. (1975). Bilateral motor interaction: perceptual-motor performance of partial and complete "split brain" patients. In K. T. Zulch, 0. Creutzfeldt, & G. C. Galbraith (Eds.), Cerebral localization (pp. 115-132). Berlin: Springer.

Rafal, R. D., & Posner, M. I. (1987). Deficits in human visual spatial attention following thalamic lesions. Proceedings of the National Academy of Science, 84, 7349-7353.

Rafal, R. D., Posner, M. I., F riedman, J. H., Inhoff, A. W., & Bernstein, E. (1988). Orienting of visual attention in progressive supra-nuclear palsy. Brain, 111, 267-280.

Remington, R. W. (1980). Attention and saccadic eye movements. journal of Experimental Psychology: Human Perception and Performance, 6, 726-744.

Remington, R., & Pierce, L. (1984). Moving attention: Evidence for time invariant shifts of visual selective attention. Perception & Psychophysics, 35, 393-399.

Rizzolatti, G. (1979). Interfield differences in reaction times to lateralized visual stimuli in normal subjects. In I. S. Russell, M. H. Van Hof, & G. Berlucchi (Eds.), Structure and function of the cerebral commissures (pp. 390-399). Baltimore: University Park Press.

Rizzolatti, G., Bertoloni, G., & Buchtel, H. A. (1979). Interference of concomitant motor and verbal tasks on simple reaction time. Neuropsychologia, 17, 223-330.

Rizzolatti, G., Bertoloni, G., & De Bastiane, P. (1982). Interference of concomitant tasks on simple reaction time: Attentional and motor factors. Neuropsychologia, 20, 447-455.

Rizzolatti, G., Riggio, L., Dascola, I., & Umilta, C. (1987). Reorienting attention across the horizontal and vertical meridians: Evidence in favor of a premotor theory of attention. Neuropsychologia, 25, 31-40.

St. John, R., Shields, C., Krahn, P., & Timney, B. (1987). The reliability of estimates of interhemispheric transmission times derived from unimanual and verbal response latencies. Human Neurobiology, 6, 195-202.

Tassinari, G., Aglioti, S., Chalazzi, L., Marzi, C. A., & Berlucchi, G. (1987). Distribution in the visual field of the costs of voluntarily allocated attention and of the inhibitory after-effects of covert-orienting. Neuropsychologia, 25, 5 5-71.


546 BRAUN ET AL.

Umilta, C., Frost, N., & Hyman, R. (1972). Interhemispheric effects on choice reaction times to one, two, and three letter displays. journal of Experimental Psychology, 93, 198-204.

Umilta, C., Rizzolatti, G., Anzola, G. P., Luppino, G., & Porro, C. (1985). Evidence of interhemispheric transmission in laterality effects. Neuropsychologia, 23, 203-213.

Verfaellie, M., Bowers, D., & Heilman, K. M. (1988a). Attentional factors in the occurrence of stimulus-response compatibility effects. Neuropsychologia, 26, 435-444.

Verfaellie, M., Bowers, D., & Heilman, K. M. (1988b). Hemispheric asymmetries in mediating intention but not selective attention. Neuropsychologia, 26, 5 21-5 31.

Wurtz, R. M., Goldberg, M. E., & Robinson, D. L. (1980). Behavioral modulation of visual responses in the monkey: Stimulus selection for attention and movement. Progress in Psychobiology and Psychology, 9, 43-83.

Zaidel, E. (1983). Disconnection syndrome as a model for laterality effects in the normal brain. In]. B. Hellige (Ed.), Cerebral hemisphere asymmetry: Method, theory, and applications (pp. 95-151). New York: Praeger.

Documents

Does so-called interhemispheric transfer time depend on attention? Does so-called interhemispheric transfer time depend on attention