Cognitive Approaches to Human Perception

Soledad Ballesteros, Soledad Ballesteros, Soledad Ballesteros

• Publisher: Psychology Press
• Year: 2014
• eBook ISBN: 9781317782599
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

Examining the current state of the research in perception stressing contributions in visual information processing, this volume provides an original and timely account of recent results obtained in this and other related areas of cognitive psychology. The scope of the book is intended to be broad, featuring state-of-the-art contributions from a number of outstanding researchers from different parts of the world -- the United States, Europe, and Australia. The intention is to update areas of considerable theoretical implications and active experimental investigation in this broad field called the "psychology of perception." This volume's main purpose is to highlight, from a cognitive position, a selected number of important theoretical and empirical topics which deal with critical issues in perception and other high level, related cognitive processes such as attention, mental representation, memory, word naming and semantic categorization.

The studies reported were designed to answer many far-reaching questions including:

* Is the global precedence effect due to low or high level processing?

* Can veridical and illusory perception be explained by the same theory?

* What is the relationship between attention and perception?

* Is perception "direct" or an inferential process?

* What mechanisms are involved in picture and word naming and categorization?

* How can word and picture processing be modeled?

The answers to these questions seek to unite theoretical perspectives on very important areas of cognitive psychology such as attention, perception, representation of visual objects and words, and human memory.

1	Cognitive Approaches to Human Perception: Introduction

Soledad Ballesteros

Universidad Nacional de Educatión a Distancia, Madrid, Spain

The study of perception has appealed to psychologists for more than a century and philosophers and neurologists even earlier (for a review, see Hochberg, 1988). The purpose of this introductory chapter is to consider how important topics in perception are approached in the chapters that follow, trying to place them along current lines of inquiry in cognitive psychology.

The cognitive approach to the study of mental processes, specifically attention, perception, mental representation, and memory, is the theoretical stance favored throughout the chapters of this book. This approach, however, is not the only theoretical position that can be adopted in the study of perception but only one of the various ways of addressing it (Dretske, 1990). Since the beginning of experimental psychology, psychophysical investigation equipped with a rigorous and sophisticated methodology has studied how human perception is influenced by the systematic variation of different stimulus variables. As Hochberg points out (this volume), even though this kind of research is very useful in showing how the receptive structures of the visual system function, this line of work provides little support in explaining human perception and constructing useful theories that could be used to understand a wide variety of perceptual phenomena. Hochberg recognizes, however, that recently, a number of important conceptual tools such as sensory networks and acquisitive neural nets have appeared that can help in such an effort.

There are other positions as well, one of which is the behavioristic approach, which focuses only on observable behavior. This theoretical stance, no longer in vogue, dominated practically all research conducted in experimental psychology during the first half of this century (see Amsel, 1989; Bechtel, 1988).

A more updated approach, however, characterizes mental activities as the underlying neural processes. Neuropsychological studies focusing on people who suffer different kinds of brain injuries have proved very useful in understanding not only impaired perception but normal perception as well (e.g., Farah, 1990; Humphreys & Riddoch, 1984; Lamb, Robertson, & Knight, 1990; Shimamura & Squire, 1987; Squire & McKee, 1992). These studies in conjunction with other investigations consisting of single-cell recording in the visual cerebral cortex and visual pathways in animals (especially in monkeys) have provided important information for knowing how normal visual perception works (e.g., Cowey & Gross, 1970; Desimone, Albright, Gross, & Bruce, 1984; Gross, 1978; Plaut & Farah, 1990).

Today, the neural approach is a very active field of investigation that provides a way of testing theories constructed from data obtained with normal as well as impaired populations. At the same time, this approach can be very valuable in proposing new hypotheses about cognitive functioning (e.g., Biederman & Cooper, 1992; Desimone & Ungerleider, 1986; Ellis & Young, 1988; Farah, 1990; Nadel, Cooper, Culicover, & Harnish, 1989; Shimamura & Squire, 1987; see Cooper, chapter 8, this volume).

The cognitive approach, by contrast, is characterized by its interest in identifying mental states functionally, considering their causal interactions with other mental states as a way of overcoming the strictures of behaviorism. At the same time, these mental states can be thought of independently of their material realization in the brain. This characteristic ensures the necessary independence of psychology from neuroscience.

Cognitive scientists investigate the nature of the human mind and its underlying structures and processes. For this purpose they try to understand the nature of the mental representations that underly perception and other cognitive processes that support our interactions with the external world. Much of the recent research in the field has been directed at the understanding of the inferential or computational character of perception and to the study of the main qualities that perception exhibits. Are those qualities comparable to those of reason and intelligence? Are perceptual processes top-down, directed by reason, or are they bottom-up, directed by the data? Does the perceptual system postulate hypothesis or solve problems? These are just some of the questions that arise in the field of human perception and are addressed in this volume.

Following Helmholtz’s tradition, a large number of researchers maintain that the correspondence between proximal and distal stimulation is never perfect (e.g., Gregory, 1978; Hochberg, 1988, chapter 11, this volume; Rock, 1983; Ulman, 1980). Hochberg, (chapter 11, this volume) distinguishes between two kinds of theories: (a) theories about edges, contours, motions, and so on (the attributes of the retinal image); and (b) theories about visual perception. The latter are concerned with the distal properties of the stimuli, whereas the former are interested in its proximal attributes. Because there is never a total (1:1) correspondence between proximal and distal properties of stimulation, higher processes are supposed to do the job.

Others scientists, like Gibson (1950, 1966, 1979; Michaels & Carello, 1981), however, challenged this position and defended that all the information necessary to specify the distal object is on the stimulus. That is, the stimulus carries out enough information to determine the character of the distal object. Contrary to constructivist theorists, following Gibson’s leadership direct perception psychologists discard all kinds of intervening cognitive variables from the act of perception. Constructivist scientists, however, believe that the sensory stimulation carries a great burden of ambiguity because a large variety of distal stimuli could be responsible for the same pattern of stimulation (see Hochberg, chapter 11, this volume). Immersed in this ambiguous situation, the perceiver has to add further information to the stimulus in order to achieve a meaningful percept.

Attending, recognizing, remembering, and reasoning about the environment is something that people do quite easily, but we are still very far from achieving a complete understanding of how these processes work. The main interest today rests on understanding and explaining the structure of mental representations underlying these psychological processes, as well as how these processes are carried out by the human perceiver. The chapters throughout this volume deal directly with these two issues.

SELECTED TOPICS AND THEORETICAL ISSUES

Most of the relationships that we establish with the environment are carried out through perception, especially visual perception. Vision is very important for human beings because it provides the perceiver with crucial information about individual objects and spatial layout. Vision helps human perceivers to distinguish a friend from a stranger, a hostile landscape from a familiar space, novel visual forms and objects from familiar ones, and so on. Nevertheless, other perceptual modalities such as audition and touch are also very important in our daily relationship with the external world. Lets take haptic perception as an example. Through purposive touch the haptic perceiver can obtain accurate and valid information about objects and surfaces, some of which can not be captured through other perceptual systems (Klatzky, Lederman, & Metzger, 1985; Loomis & Lederman, 1986; Millar, 1978). Important volumes in tactual (Heller & Schiff, 1991; Katz, 1989) and auditory perception (Bregman, 1990; Buser & Imbert, 1992) have been published in recent years, and the interested reader will find in these books important studies about the properties, characteristics, and capabilities of touch and audition.

Cognitive scientists have as a goal the understanding of human minds, and a good way to achieve this goal is to formulate hypotheses and test theories about how attentional and perceptual mechanisms work, as well as to try to discover the nature of representations underlying object perception and memory. This understanding will eventually help us in building machines provided with similar capabilities that could be used in industry and by disabled populations in their interactions with the world.

Attention and Visual Perception

The chapters in Part I of this volume deal with fundamental issues in attention and its relation to perceptual organization. Most of the recent developments in attention are due to research trying to determine the locus of the attentional bottleneck. Traditionally, the two most important issues in attentional investigation have been: (a) the problem of the limited capacity of attention, also called the limitation of resources; and (b) where the locus of selection in the information processing system must be located, that is, the problem of early versus late selection.

This line of enquiry started with the pioneering work of Broadbent (1958), who argued that there is a selective filter working at an early stage of the sensory analysis of the stimuli, before perceptual recognition or categorization of the stimuli takes place. According to Broadbent, the main purpose of this filter is to avoid overload inside the limited capacity mechanism. Some physical characteristics of the sensory input are processed in parallel until its arrival to this filter, but information not selected at this point is excluded from further analysis. This excluded information is neither recognized nor categorized. Soon after the publication of Broadbent’s work, some investigators, proposed based on empirical results, that the filtering occurs quite late, after categorization, or even not at all (for reviews, see Johnston & Dark, 1986; Kahneman & Treisman, 1984).

It is well accepted that phenomena such as perceptual organization and figure-ground segregation based on the detection of homogenous regions and discontinuities may occur automatically (or preattentively) at the beginning of a visual presentation (e.g., Shiffrin, 1988). However, in addition, these phenomena seem to be affected by attentional processes, as results show in chapter 2 of this volume by Rock and Mack. These findings can be taken as new evidence that when perceptual patterns are viewed without attention, perceptual grouping does not seem to take place.

Rock and Mack’s results pose two main problems for previous research. The first problem is how to reconcile these negative results on perceptual organization with those obtained from the search paradigm indicating parallel processing. The second problem is how to explain the early stage of preattentively visual field organization if it is not based on Gestalt laws. Rock and Mack, in an attempt to reconcile these two sets of evidence, assume the existence of three different levels of attentional processing: (a) a nonattentional level at which some stimulus properties may be perceived without attention or intention to perceive; (b) a distributed attentional level in which multiple elements can be detected in parallel over the visual field; and (c) a highest level in which focal attention to the stimuli is necessary.

In summary, Rock and Mack, taking into account new findings from a number of recent studies reported in their chapter, as well as on previous results, argue that shape perception always requires attention and without it perception cannot be achieved (Rock 1983; Rock & Gutman, 1981).

Ballesteros and Manga (chapter 3) try to understand the way in which irrelevant information present in the experimental situation influences the processing of relevant characteristics of the stimulus. In other words, the question is whether selective attention to the relevant information is possible when irrelevant information is also present in the visual field. Several lines of research are presented that seem to support the idea that various types of information about the stimulus, or the context in which the stimulus appeared, are available to the perceiver shortly after the stimulus presentation.

Nevertheless, after years of investigation the controversy is still present (see Allport, 1989; Broadbent, 1991; Shiffrin, 1988). At the moment, it seems reasonable to propose that the extraction of information is carried out by attentional processes prone to limitations that may appear at any stage of processing.

Form Perception

Chapters in Part II are mainly concerned with several important issues in form perception. Pomerantz, Carson, and Feldman present several lines of experimental results related to the diagnosis of the psychologically real parts of visual forms. The leading hypothesis is that if two elements function as separate parts, the perceiver should be able to attend to each element selectively, otherwise interference will occur. Pomerantz, Carson, and Feldman (chapter 6, this volume) present differences between Garner and Stroop interference, asymmetries of interference, and how perceptual interactions can be diagnosed in the laboratory by performance in visual information processing tasks.

The Processing of Global and Local Information in Form Perception. An old but still pervasive topic in visual perception and cognition is whether the recognition of a global, overall pattern precedes the recognition of its component elements.

Throughout the history of experimental psychology opposed positions over this issue have appeared. In the early times, structuralists such as Titchener (1909) proposed that the extraction of components in a visual pattern is carried out first, followed by the building up of the total percept from these elementary units. According to the structuralist position, the processing of local units takes place first. In contrast, Gestalt psychology holds that the perceptual whole is qualitatively different from the composition of its individual parts or units, maintaining a holistic view.

More recently, cognitive researchers presented experimental results that are in agreement with the idea that perceptual processes proceed from global to local, that is, from more general to more detailed extraction of information. This is known as the global precedence hypothesis, first proposed by Navon (1977). The common way of testing this proposal has been to observe the effect of hierarchically constructed patterns over behavioral measures such as the reaction times.

The visual forms are usually large letters composed of a certain number of smaller letters. In a typical experiment, participants are asked to respond to the global or to the local level of the perceptual pattern according to experimental conditions. The results obtained by Navon and others favored a global-to-local order of processing. Global letters are processed before its component local elements (the global precedence effect); and global information produces an inhibitory effect in the condition of local directed attention when global conflicting information is present in the display (the interference effect). On the contrary, local conflicting information does not affect the processing of global information. These two effects taken together were interpreted by Navon (1977) as a sufficient demonstration of the inevitability of global precedence in visual information processing. This interpretation, however, has been questioned on several grounds, for example, attributing the results either to the differential salience of the stimuli (Garner, 1983; Pomerantz, 1983) or stating that the phenomenon is not perceptual in nature but postperceptual (e.g., attentional and response competition; Boer & Keuss, 1982; J. Miller, 1981).

More recent studies (Navon, 19...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Dedication
• Contents
• Preface
• Acknowledgments
• 1 Cognitive Approaches to Human Perception: Introduction
• Part I Attention and Visual Perception
• Part II Form Perception
• Part III Mental Representation of Pictures, Words, and Objects
• Part IV Perceptual Theory
• Author Index
• Subject Index