Switching

While the above discussion of memory has emphasized its probable bilaterality in many instances, this is by no means a closed subject. Regardless of just where en-grams may lie and how they are accessed, the human hemispheres display a pronounced and unequivocal specialization for different types of neuronal processing. This alone raises the issue as to how attention may be distributed between the hemispheres to deal with the differing modes of operation. There is a large literature on the focusing of attention, mostly involving, ultimately, location in space. This immediately harks back to the point made in the introductory paragraph of this essay: that duplication arose in nervous systems consequent to the desirability of making choices between right and left. It is thus rather surprising that the extensive research on attentive locus is couched purely in terms of space, without reference to hemispheric function.

As a general rule, both hemispheres cannot be attended to concurrently. This situation is particularly well illustrated in split-brain patients, in whom for most circumstances attention still needs to be directed to one or other hemisphere/visual field. Trevarthen (1974), for instance, reports that when such a patient was asked to mark with her left hand a large white square present on a black background, she protested that the square disappeared; that is, when the right hemisphere assumed control of the left hand, the attentive focus passed to that hemisphere, and perceptual processing in the left, speaking hemisphere was curtailed. Other work with split-brain patients confirms this general theme: that attention is primarily a process that is not readily divided and must thus be allotted sequentially, rather than concurrently, to each hemisphere (Holtzman, 1985; Levy and Trevarthen, 1976; Teng and Sperry, 1974).

If attention must then be switched between hemispheres, how long does it take? Although couched merely in terms of right and left visual fields, an ingenious experiment of Weichselgartner and Sperling (1987) suggests an answer, and it is surprisingly long. Their subjects, all normal, were required to watch for a particular letter in a stream of letters being displayed for 18 ms at a rate of 10—12.5 letters/s in the left visual field. When the target letter appeared, the subjects then had to switch attention to a similar stream of numerals in the right visual field and report the first numeral(s) that they could detect and remember. On 50% of the occasions the first remembered numeral occurred 300 ms after the signal to switch, the great majority lying between 200 and 400 ms. On the other hand, if the subject was already attending to the numeral stream and the signal to begin remembering was a brightening or highlighting of one of the numerals, either that numeral or the one immediately following it at 100 ms could be reported; that is, without the switch in visual field, the attentive process was engaged much more rapidly. Interestingly, in this second situation the later engagement of attention also occurred, for subjects then also had a peak of remembering in the 300-ms time frame found in the switching experiment.

A further feature of such engagement of attention is that once activated, it runs a course of 100-300 ms during which switching to other items is impaired (Duncan, Ward, and Shapiro, 1994).

We (Kavcic, Krar, and Doty, 2000) have endeavored to measure such temporal cost in a situation more clearly allied to switching between use of one hemisphere versus the other. Two types of stimuli were used, chosen to tap left and right hemisphere aptitudes, respectively: nonoffensive four-letter English words and complexly patterned multicolored images having no ready verbal descriptors. These items were intermingled pseudorandomly and presented on a computer monitor one at a time for 200 ms followed by a multicolored, patterned masking flash. The subject's assignment was to identify whether a given item had or had not been seen previously in the sequence, that is, a continuous recognition task. In the 240 trials 120 items were new and 120 old; on 20 occasions a word followed an image, and on 20 others a word preceded an image. These latter two conditions were the focus of the experiment, measuring the extra time it took in evaluating whether the item was new or old when a switch (in activation of one versus the other hemisphere?) occurred in comparison with when it did not. With nine subjects, for an image following a word, the added time averaged 78 ms beyond the 850 ms when an image followed an image; and for a word following an image, the added time was 92 ms beyond the 792 ms for a word following a word. Thus, there was a very significant delay (p < 0.001 and 0.005, respectively) in switching between processing the one to the other type of stimulus.

To affirm whether this was indeed a switch between hemispheres, a putatively intrahemispheric switch was assayed between words and nonwords. The switching cost in such circumstance was only @30 ms, thus being consonant with the idea that the comparable times for word/image switching might include an interhemi-spheric cost. However, such simplicity was dispelled by the next finding: that switching between stimuli, presumably both targeted at right hemispheric processes, human faces and colored images, exacted a switching time of 70—79 ms. Thus, to assign a credible cost to in-terhemispheric switching per se, rearrangement of the underlying neuronal processing will have to be more effectively limited to a selected hemisphere. It is of some interest that in no instance were any of the subjects aware that their reaction time was slowed on the switch trials.

An important factor that needs to be considered in such experiments is that attention may be a specialty of the right hemisphere. This is most strikingly illustrated by the fact that lesions of the right human hemisphere are prone to produce neglect of events occurring to the left, whereas the corresponding deficit following left hemisphere lesions is far less (e.g., Robertson and Marshall, 1993). The remarkable uniformity of this right hemispheric propensity for control of attention in the human population has been demonstrated by Spiers and colleagues (1990). In a series of 48 consecutive patients undergoing amobarbital infusion into right versus left carotid arteries, which briefly inactivates the respective hemisphere to test for sidedness of speech before neu-rosurgery, only injection on the right produced neglect (of events to the left). Surprisingly, this was also true of left-handed individuals, even of those in whom speech appeared to be a function of the right hemisphere! The effect was particularly marked by changes in the electroencephalogram of the right frontal cortex.

More subtle evidence of this right hemispheric prepotency for attentive activation appears in the study of Levy, Wagner, and Lu (1990). Using verbal stimuli presented randomly to one or the other visual field for @200 ms in normal subjects, followed by a masking stimulus, they found the expected superiority of the right visual field. However, it was also clear that the trials that were preceded by presentation to the left visual field (right hemisphere), regardless of which hemisphere responded next, were consistently more accurate than when a right visual trial preceded. Thus, activation of the right hemisphere had a lingering effect, perhaps of sustaining a more alert state, leading to augmented performance by either hemisphere. Unfortunately, Levy and colleagues did not report measures of reaction time in these experiments.

Perhaps equally unexpected is the finding that this right hemispheric bias for attentive processes can be seen even in split-brain patients (Mangun et al., 1994). In this instance reaction time was the critical measure, and it was demonstrated that the right hemisphere allocated attention to either visual field (faster or slower response depending on field location of a cue preceding the target), whereas the left hemisphere dealt only with the right visual field. As they note, such findings clearly implicate the participation of subcortical systems in attentive processes, since the asymmetry and bilaterality demonstrated here survive callosotomy.

Was this article helpful?

0 0

Post a comment