Tutorials in Visual Cognition

Veronika Coltheart

• Publisher: Taylor & Francis
• Year: 2011
• eBook ISBN: 9781136940347
• Edition: 1
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

In the late-1980s, visual cognition was a small subfield of cognitive psychology, and the standard texts mainly discussed just iconic memory in their sections on visual cognition. In the subsequent two decades, and especially very recently, many remarkable new aspects of the processing of brief visual stimuli have been discovered -- change blindness, repetition blindness, the attentional blink, newly-discovered properties of visual short-term memory and of the face recognition system, the influence of reentrant processing on visual perception, and the surprisingly intimate relationships between eyeblinks and visual cognition.

This volume provides up-to-date tutorial reviews of these many new developments in the study of visual cognition written by the leaders in the discipline, providing an incisive and comprehensive survey of research in this dynamic field.

1
Introduction

VERONIKA COLTHEART

INTRODUCTION

Tutorials in Visual Cognition is the outcome of a meeting held at the Macquarie Centre for Cognitive Science in Sydney, Australia, to discuss current topics in a variety of areas of visual cognition. The chapters of the book are based on the contributions presented at this meeting and intended to provide a tutorial review of each topic and the methods used to study it. The readership includes postgraduate and postdoctoral researchers who want to become acquainted with research in various areas of visual cognition.

THE REENTRANT PROCESSING ACCOUNT OF PERCEPTION AND COGNITION

A highlight of this volume is the account of iterative reentrant processing proposed by Vincent Di Lollo in Chapter 2. Di Lollo presents a model that accounts for the mechanisms of information processing by all the sensory modalities (though he focuses on vision) and by higher level cognitive processes. He points out that visual perception has been attributed to the results of feed-forward mechanisms but that more recent neurophysiological investigations have demonstrated the existence of feedback mechanisms as well; it is the feedback mechanisms that constitute reentrant processing.

The ideas of feedback and feed-forward activation and cascaded processing have been a feature of various cognitive theories (e.g., of word recognition), but only now are such ideas being formulated in terms of explicit brain mechanisms. Di Lollo presents a lively review of ideas throughout the ages about how the brain works; these ideas were heavily influenced by the technology of the day—aqueducts, telephone switchboards, and computer hardware. All of these function solely in feed-forward mode.

Di Lollo points out that the integration of separate sorts of visual information, color, form, orientation, etc. (the “binding problem”) presents a major theoretical obstacle to the understanding of visual perception and cognition. He goes on to argue that the feed-forward and feedback reentrant processing concepts remove the binding problem. The chapter considers the possible mechanisms and evidence for them in detail. The reentrant processing model is applied to the explanation of object recognition, inhibition of return, motion perception, visual search, object substitution masking, and other phenomena.

SPATIAL VISUAL ATTENTION

A number of the chapters in this volume are concerned with various aspects of visual attention—specifically, spatial attention. In Chapter 3, interest has focused on voluntary and involuntary attention, a distinction made by Wundt (1902) as cited by Bill Prinzmetal and Ayelet Landau, who argue that these forms of attention have different mechanisms. They use the spatial cueing paradigm devised by Posner (1980) in which observers have to detect or identify a target in one of two locations. The target is preceded by a valid cue (one that predicts the target location), an invalid cue, or no cue at all. Typically, the majority of the cues are valid so that the cue will encourage subjects to shift attention to the cued location. Voluntary attention can be deployed only when the cue–target SOA (stimulus onset asynchrony) is at least 200 ms.

Prinzmetal and Landau report that when task conditions were manipulated so that either response accuracy or reaction time was used as the response measure, different conclusions were indicated concerning voluntary and involuntary attention. They report experiments using functional magnetic resonance imaging (fMRI) measures indicating different patterns of brain activity for voluntary attention to a predictive cue compared to those for involuntary attention.

CONTROL OF VISUAL SELECTION

Effective response to the complex visual environment we inhabit requires effective attentional selection, which may be controlled by the observer (top-down processing) or may be elicited by properties of visual stimuli in the field of view (bottom-up processing). The extent to which these two processes determine visual selection is the concern of Chapter 4 by Jan Theeuwes and Artem Belopolsky. They investigated this question using a feature singleton visual search task. The target possesses a unique basic feature (e.g., it may be the sole red item in the display, or the red item may be an irrelevant distractor—the “irrelevant singleton” paradigm). Their chapter reviews the extensive evidence concerning the long-standing debate in this field and concludes that the size of the attentional window determines the likelihood of visual capture by a salient singleton.

VISUAL SEARCH

The paradigms used by Prinzmetal and Landau and by Theeuwes and Belopolsky involved small arrays of visual stimuli; however, the visual field frequently contains a large array of items. The observer frequently has to search the array to determine whether a required item is present. The processes involved in visual search have been the subject of intensive study since Treisman’s research and theorizing ignited interest in this field in the 1980s. An impressive and substantial body of research and theorizing has been assembled by Jeremy Wolfe and his colleagues.

In Chapter 5, Wolfe, Horowitz, Palmer, Michod, and Van Wert present the 4.0 version of their guided search model, focusing on the nature of the visual-search guidance process, specifically on its temporal features. Early investigations reported evidence of parallel search for targets possessing a simple unique basic visual feature such as color, size, or orientation; targets defined by a conjunction of features could be found only after a slow serial search process. However, as Wolfe has noted in an earlier review (1998), search is frequently neither clearly parallel nor clearly serial, and it might more properly be termed more, or less, efficient. Processes of guidance are hypothesized to play a role in search, and object recognition is represented as a diffusion process in the current 4.0 model.

In many studies of visual search, a target definition is given at the outset of the task. In contrast, Wolfe et al. asked subjects to search for the unique item in an array without telling them about the possible target attributes. They introduced some clever manipulations to control guidance and contrasted unguided conditions with those in which different types of cue were provided at varying intervals before the array appeared. Guidance can be viewed as the “reentrant process” postulated by Di Lollo in his earlier papers (e.g., Di Lollo, Enns, & Rensink, 2000) and further developed and extended in his theoretical review chapter in this volume. Wolfe et al. present evidence to suggest that it takes time (a few hundred milliseconds) for search guidance to be established.

DISRUPTING VISION: EYE BLINKS AND COGNITION

In Chapter 6, David Irwin and Laura Thomas took as an index of cognitive processes the mundane occurrence of frequent eye blinks whose disruption to vision typically goes unnoticed. People usually blink 12–15 times per minute in normal daily conditions, with vision obscured for 100–150 ms each time. Vision is also suppressed briefly just before and after each blink.

Irwin and Thomas point out that people blink less often during difficult visual tasks and generally less often during cognitively taxing tasks. Their research investigated the question of whether blinking affects thinking—specifically, whether it impairs cognition. They report that blinking disrupted information maintained in iconic memory. Some evidence indicates that blinks suppress activity in primary visual cortex (V1) as well as in posterior parietal cortex. The latter is thought to be involved in spatial working memory and spatial attention, and Irwin and Thomas report some of their findings on effects of blinks on allocation of attention in space.

VISUAL SPATIAL ATTENTION AND VISUAL SHORT-TERM MEMORY

Chapter 7 by Pierre Jolicœur (with colleagues Dell’Acqua, Brisson, Robitaille, Sauvé, Leblanc, Prime, Grimault, Marois, Sessa, Grova, Lina, and Dubarry) is concerned with the interaction between early and late attentional mechanisms in selection and retention of visual information. They report research using visual cognition tasks with electrophysiological, magnetoencephalographic (MEG), and fMRI measures to study visuospatial attention and visual short-term memory. The N2pc has been shown to be a correlate of visuospatial attention to lateralized visual stimuli and the SPCN (sustained posterior contralateral negativity) has been associated with the use of visual short-term memory.

These authors studied visual attention in a temporal visual search task using rapid serial visual presentation (RSVP). In the typical task, two targets are inserted at various distances apart in a long sequence of 10 or more distractors; each item is shown for about 100 ms. The first target (T1) is usually easily identified, but the second target (T2) is frequently missed if it occurs within an interval of about 500 ms after T1. This difficulty in T2 detection or identification is known as the attentional blink; it is a transitory deficit because the T2 report recovers as the interval between T1 and T2 increases beyond 500 ms.

Jolicœur et al. used a peripherally presented colored T2 at the left or right of the central stream (as well as an irrelevant distractor on the opposite side). This procedure allowed them to measure the N2pc used to track the deployment of visual spatial attention. In a series of experiments, they found N2pc attenuation during the attentional blink interval (when T2 was missed). In a single target condition (report T2 only), the SPCN thought to reflect visual short-term memory (STM) activation was large. Noting subjects’ difficulty in ignoring a T1 that was the same color as T2 in the single target condition, they explored the mechanisms of contingent capture using event-related potential (ERP) N2pc measures. In addition to the likelihood that a red target would not be reported when a peripheral distractor was red, there was an N2pc that did not occur when the peripheral distractor was another color.

REPETITION BLINDNESS

It is difficult not only to detect and identify two targets in an RSVP sequence of items shown at a rate of about 10 per second, but also to detect and report a second (repeated) occurrence of an item shown in similar temporal conditions. This phenomenon was termed repetition blindness by Kanwisher (1987) and is reviewed in Chapter 8 by Veronika Coltheart. Repetition blindness (RB) is large when the first item and its repetition are close in a sequence and diminishes with increased lag (or SOA). However, the time course of RB is usually shorter than that of the attentional blink.

Repetition blindness is also found when distractor/mask similarity to the to-be-reported items is low; the attentional blink is likely to be diminished or absent under these conditions (Chun, 1997). The paradigms in which RB occurs can include short or long sequences of to-be-reported items. When long sequences are presented, the items are typically words that form a grammatical sentence if all words are reported. Several accounts of RB have been proposed with varying attributions of the role of memory in causing the repetition deficit, and these are evaluated in the chapter.

DETECTION OF WEAK VISUAL SIGN...

Table of contents

• MACQUARIE MONOGRAPHS IN COGNITIVE SCIENCE
• Contents
• Contributors
• 1 Introduction
• 2 Iterative Reentrant Processing
• 3 Dissecting Spatial Visual Attention
• 4 Top-Down and Bottom-Up Control of Visual Selection
• 5 Getting Into Guided Search
• 6 Eyeblinks and Cognition
• 7 Visual Spatial Attention and Visual Short-Term Memory
• 8 A Review of Repetition Blindness Phenomena and Theories
• 9 Spatial Attention and the Detection of Weak Visual Signals
• 10 Face and Object Recognition
• 11 Is Face Processing Automatic?
• 12 Visuospatial Representation of Number Magnitude
• 13 Visual Memories
• Author Index
• Subject Index