Use this URL to cite or link to this record in EThOS:
Title: Visual perception and metacontrast at rapid input rates
Author: Cumming, Geoff D.
Awarding Body: University of Oxford
Current Institution: University of Oxford
Date of Award: 1971
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Metacontrast is the type of visual masking which occurs when the normal perception of one stimulus, the test stimulus (TS), is disrupted by a second stimulus, the masking stimulus (MS), presented 50-100 msec later in time and at an adjacent but not overlapping position. Mayzner has described metacontrast in displays consisting of several letters forming a horizontal row but displayed very briefly in a scrambled temporal order; the first few letters presented are enclosed by adjacent letters presented later in time and are reported as 'not seen'. He called this metacontrast phenomenon 'sequential blanking'. In this thesis a series of experiments is reported which investigated just what information about such suppressed letters, those which are 'not seen', is available to the observer. A review of our present knowledge of masking and metacontrast is given in Chapter 2 and then in the following chapter the observations of Mayzner and his associates are described and their model of visual information processing, baaed on their work with sequential blanking, is discussed. The model postulates that letters do not reach consciousness ('subjective visual experience', SVE) until at least about 130 msec after presentation and that interactions in that time among the responses of visual analysers determine whether any particular letter will arrive in SVE or will be suppressed. It is suggested in Chapter 3, however, that suppression is not all-or-none and that investigation of the different perceptual analyses required for different responses to suppressed letters, rather than simply asking whether the letters are seen or not, would be valuable. Chapter 5 describes experiments in which S classified whole displays by making fast keypress responses. The results showed that suppressed letters can be analysed sufficiently quickly and in sufficient detail to control a choice motor response of normal speed and accuracy, even when the analysis had to be sufficiently detailed to distinguish letters and digits. In the experiments of Chapter 6 S searched through displays of random letters looking for a single target letter and gave keypress responses to indicate whether each display did or did not appear to contain a target. Suppressed targets were usually detected quite efficiently, but if S adopted a strict criterion for target presence it was mainly the suppressed target letters which were missed. The first two experimental chapters described situations in which suppressed letters were analysed at least almost as efficiently as other letters. In the next, Chapter 7, experiments are described in which S identified letters at various positions in the displays. Suppressed letters were found to be generally difficult to identify. In addition, if position intrusion errors were excluded the remaining errors, item errors, were more likely to be acoustically confusible with the correct letter if the correct letter had been suppressed. This result suggested that suppression is the degradation of visual information in particular and that sometimes S could name the letter before it suffered suppression. In the experiments described in Chapter 8 a switching task was used. S gave a keypress response signalling either that two successively presented letters were the same or were different, or that two entire displays were the same or were different. Acoustic and visual confusibility was manipulated in the 'different' trials in an attempt to determine whether the memory and watching processes employed depended mainly on acoustic or visual coding. The results suggested once again that suppressed letters are more difficult to process than other letters and that in some conditions there appears to be a tendency for acoustic to be relatively more important for these letters. The final experimental chapter, Chapter 9, reports investigations of the appearance of displays. Ss gave ratings of the brightness, apparent duration and perceived clarity of individual letters in metacontrast and control displays. Suppressed letters were judged as appearing less bright and shorter, but no less clearly defined than other letters. The results were interpreted as showing that suppressed letters are available for processing for a short time before suppression occurs. Metacontrast suppression appears to be a reduction in apparent brightness caused by fairly peripheral neural processes. But whether suppressed letters are reported as 'seen' or not depends on whether an appropriate perceptual analysis can be applied to them sufficiently early and quickly to give a confident categorisation, which can preserve information about them despite suppression of their apparent brightness. For example a letter-digit discrimination can be made sufficiently quickly to trigger a choice motor response of normal speed and accuracy. There is also some evidence that a word, all of whose letters are suppressed, might even so be identified because wholistic Word recognition is so fast. Organisation of the perception of the whole display into a good Gestalt is another way in which information about suppressed letters may be preserved. Analysis of letters before they suffer suppression improves with long practice. The conclusion that information about suppressed stimuli is available for a short time and then lost unless recorded and stored applies also to standard metacontrast displays, of squares or circles for example. It disagrees with theories that metacontrast consists simply of immediate, irretrievable, partial loss of information. It also suggests that we should not postulate that stimuli either reach SVE and are seen, or are entirely suppressed, but rather we should investigate the the course of the brightness reduction that is metacontrast and investigate the perceptual analyses which allow information about these stimuli to be preserved through suppression.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available