This paper has proceeded on the assumption that the concept of “masking” is both inadequate and misleading as an explanation of the perceptual effects produced by the rapid successive presentation of brief stimuli to adjacent or over- lapping retinal areas. The objections offered against the concept of “masking” and, in particular, “backward masking,” can be reduced to three basic arguments:
1. The concept of masking is inadequate because it cannot explain the emergence of new perceptual attributes produced by the stimulus situation.
2. The concept of masking is unclear, as can be shown by its relativistic applications.
3. The concept of masking is misleading because it does not allow for distinctions between the various perceptual causes of performance changes. These three objections deserve further elaborations.
I will discuss [these arguments] one at a time.
1. The masking concept is inadequate because it cannot explain the emergence of new perceptual attributes, e.g., apparent movement, not present in the perceptions produced by the presentation of the separate stimuli in isolation.
These emerging attributes are quite common when the MS consists of a structured pattern or form. The concept of masking may be {64} adequate for explaining a S’s failure to perceive an otherwise suprathreshold TS in the presence of visual noise (“noise” being defined as an unstructured, diffuse stimulation). But noise masks and structured masks do not produce the same perceptual effects. The former produce informational loss; the latter produce increased informational complexity. The concept of masking obscures this difference.
2. Perhaps the most intellectually stifling characteristic of the concept of masking, in experiments employing sequential and spatially overlapping stimuli, is its relativistic application.
If two brief visual stimuli are successively presented to overlapping retinal areas within 100 msec of one another, the S’s ability to perceptually separate the two stimuli decreases as the onsets of the stimuli approach coincidence. If the two stimuli are identical, one will be measuring the S’s two-pulse resolution capacity. The concept of masking is not applied to such an experiment since recognition and detection criteria established by students of two-pulse threshold never require the perception of temporal attributes of the separate stimuli. Nevertheless, if two identical stimuli cannot be temporally resolved, the S cannot strictly be said to have accurately “recognized” either of them, since he is unable to distinguish their respective temporal attributes. In other words, the introduction of the second of two identical stimuli to the same retinal area certainly interferes with the S’s recognition of the first stimulus; yet {65} the first must be said to have precisely the same effect on the second. Therefore, since recognition thresholds are being changed by the introduction of additional stimulation, this situation should qualify as a case of masking.
But because the S is asked to report only on his experience of the interstimulus interval, rather than on his experience of the temporal attributes of the stimuli themselves, the applicability of the masking concept to this situation is generally overlooked. If the two stimuli differ only in wavelength and are presented within 60 msec of one another, besides being unable to resolve the stimuli temporally, the S will be unable to differentiate their respective chromatic attributes. Under such conditions the S experiences a single continuous event whose duration is indistinguishable from that of either of the stimulus components presented alone (Efron, l970a, b), and whose color does not correspond to that of either of the stimulus components, but is instead the result of their chromatic integration.(Efron, 1967). The concepts of forward and backward masking are not applied to the types of experiments just discussed precisely because the obvious reciprocal nature of the stimulus inter- actions makes it impossible to assign the terms “masking stimulus” or “masked stimulus” to either of the stimulus components. Yet this designation is a precondition for deter- mining whether “forward” or “backward” masking has occurred.
It is for this reason that the concept of successive masking {66} is applied only to those situations wherein the S’s perception of one of the stimuli remains relatively unchanged, while his perception of the other is relatively altered. This criterion can generally be obtained by making one of the stimuli considerably larger than the other. An additional perceptual advantage may be given to the larger stimulus by making it greater in duration and/or intensity than the smaller stimulus. Under conditions such as these the S’s perception of the larger stimulus remains relatively un- changed, regardless of whether it is immediately preceded or followed by the smaller stimulus. On the other hand, the S’s perception of the smaller stimulus tends to be radically altered as a function of the lateral temporal proximity to the larger stimulus. By virtue of its relative perceptual stability the larger stimulus is regarded as the “masking stimulus” (MS), while the smaller stimulus, being strongly susceptible to perceptual alteration, is referred to as the “test stimulus” or “target stimulus” (TS).
The above argument explains why, in the experimental portion of this paper, the stimulus that remained relatively unaltered perceptually was assigned the name “dominant stimulus” (DS), rather than the obviously biasing name “masking stimulus” (MS).
3. The masking concept is misleading because it does not allow for distinctions between the various perceptual causes of the S’s change in performance. This is particularly true when masking is defined in behavioral rather {67} than perceptual terms. If, as is often the case, masking is defined in terms of changes in the S’s performance level on perceptual tasks requiring him to respond in exclusively binary (yes or no) terms, the S’s performance on any trial communicates nothing about what he is experiencing. At best such a procedure can supply the experimenter with a stimulus- response correlation, but this method can never yield an understanding of the cause or meaning of such a correlation. When binary response indicators are employed, the causes of the S,s performance are obscured rather than illuminated.
This is because when a S is asked not what he sees but merely whether or not he sees a particular stimulus event, he is forced to ignore those very aspects of his experience which would yield an understanding of how perception is related to the stimulus conditions.
The present paper has proceeded on the assumption that if one seeks a real understanding of the stimulus detection and recognition experiments, the primary question to be answered is not “Under what conditions are the TS and MS responded to differentially?” but rather, “How do the stimulus conditions determine what is experienced?” An adequate answer to the first question requires an adequate answer to the second. How a S responds on a given perceptual task is entirely determined by his conscious decisions concerning what he sees. Restating this in negative terms, there are no non-conscious processes triggering non-conscious responses that the S is helpless to prevent. {68}
Proceeding from this point of view, our analysis of metacontrast and the Broca-Sulzer and Crawford effects has revealed the following general facts [and conclusions]:
(1) What has often been interpreted as failure to perceive a particular stimulus can, in many cases, be understood as failure to perceive a stimulus in a particular way.
(2) While a TS may not be perceptually distinguishable from a later or prior MS, it cannot necessarily be concluded that the information contained in the TS has no effect on what is perceived. The fact is that a trained S can use subtle or obvious perceptual effects produced by TS as cues for the latter’s presence illustrates this second point and makes it necessary to differentiate between how the stimulus is perceived and how it is interpreted by the S, i.e., between the S’s experience of the perceptual evidence and his inference from the perceptual evidence.
(3) There is always a reciprocal interaction between each of the components of spatially or temporally contiguous stimuli. This reciprocity is not e1iminate< but is merely obscured when one of the stimulus components is allowed to dominate the final perception produced by the stimulus complex. Instead of appealing to the concept of backward visual masking, this paper has attempted to show that the visual phenomena historically “explained” by this concept can be made far more understandable by appealing jointly to the following facts and assumptions:
(4) There exists a neural processing period the average {69} duration of which is 60-70 msec for stimuli of moderate intensity. The duration of this processing period is increased by approximately 10 msec for each log unit reduction in stimulus intensity. The nervous system, when processing sensory information, does not correct for neural latency differences caused by stimulus intensity differences.
(5) The processing time for each element of a stimulus complex depends upon the intensity of the stimulus energy associate with that particular element. The extent to which information from one stimulus will be integrated with information from a second stimulus to produce a single perception depends upon (a) the extent to which the processing periods of the two stimuli overlap, (b) figural and spatial relationships between the two stimuli, (c) attentional factors, and (d) other factors relating to internal patterning, informational congruity, e.g., color, orientation, and form. {70}
Copyright © 2014 by bioperipatetic. Published on April 8, 2014 @ 9:23 pm
Latest revision: April 30, 2014 @ 5:45 pm
bioperipatetic 