Academic journal article Perception and Psychophysics

Backward Recognition Masking as a General Type of Interference in Needed Poststimulus Processing

Academic journal article Perception and Psychophysics

Backward Recognition Masking as a General Type of Interference in Needed Poststimulus Processing

Article excerpt

Auditory backward recognition masking (ABRM) has been argued to reflect interference in the storage and/or processing of a short-lived sensory form of information and has been viewed as a relatively invariant attribute of auditory pitch processing for very brief stimuli that are minimally separated in frequency (ΔF). In contrast, the present study demonstrates that ABRM reflects interference with several basic principles of auditory processing. Measured in terms of target tone duration, rather than ΔF, ABRM is demonstrated for target stimuli representing the interval of a musical fifth and masker-target stimulus intervals of a musical third, with thresholds ranging from approximately 22 to 55 msec and psychometric functions that are indicative of more than one contributing factor. On the basis of common underlying principles, the possibility that the threshold for the identification of temporal order of onset reflects ABRM and possible implications for the perception of complex stimuli, including speech, are discussed.

Auditory backward recognition masking (ABRM) is the interference in the recognition of some sound attribute due to the presentation of a subsequent sound. In a typical ABRM paradigm, illustrated in the upper panel of Figure 1, the listener is asked to identify the pitch of alternative brief (10- or 20-msec) stimuli that, when presented in isolation, are close to the difference limen (DL) for recognition. When the target stimulus is followed by a fixed-property contextual stimulus, recognition accuracy is reduced by an amount inversely related to masker delay. Investigated over the last decade largely in applied contexts as a fixed characteristic with neurological correlates (e.g., McArthur & Bishop, 2002) or individual differences that may be correlated with specific cognitive attributes-for example, reading disabilities (Griffiths, Hill, Bailey, & Snowling, 2003; McArthur & Hogben, 2001) or intelligence (Deary, 1995)-ABRM would appear to be of relatively limited relevance to the perception of complex stimuli. In fact, the gating paradigm (e.g., Grosjean, 1996), commonly used in the speech literature to evaluate when auditory information is first perceptually available, makes the implicit assumption that later occurring portions of the stimulus are irrelevant and, thus, assumes an absence of any ABRM. Rather than being a relatively minor form of specific, relatively invariant interference with limited general relevance, we believe that ABRM reflects the interaction of several general principles of auditory information processing that are common to simple and complex listening situations. The present study is an initial attempt to develop the broader generality and importance of the principles underlying this phenomenon.

Auditory Backward Recognition Masking

ABRM was originally investigated from two different research perspectives. The goal for Massaro (1970) was to evaluate the duration of the short-term auditory sensory store, with the notion that the contextual stimulus disrupted the sensory information still being processed in the pitch identification task. The goal for Ronken ( 1971,1972) was to investigate the frequency DL of brief 10-msec tones. These and the other past ABRM studies always investigated brief target stimuli that were very similar in frequency, measuring either the frequency DL (thus, constant accuracy) or accuracy for a constant (near the DL) frequency difference. For relatively naive listeners, accuracy is significantly reduced at relatively short masker delays (e.g., interstimulus interval [ISI] < 50 msec), systematically improving to an asymptote that often still indicates some interference at ISI values of 250-350 msec (e.g., Massaro, Cohen, & Idson, 1976). When the frequency DL is measured, there is a relatively small change in the DL, and interference is found only up to ISI values of 100-150 msec (e.g., Foyle & Watson, 1984; Leshowitz & Cudahy, 1973). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.