UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A contextual effect in feature detection Womersley, Marcus David 1975

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1975_A8 W64.pdf [ 3.74MB ]
Metadata
JSON: 831-1.0100081.json
JSON-LD: 831-1.0100081-ld.json
RDF/XML (Pretty): 831-1.0100081-rdf.xml
RDF/JSON: 831-1.0100081-rdf.json
Turtle: 831-1.0100081-turtle.txt
N-Triples: 831-1.0100081-rdf-ntriples.txt
Original Record: 831-1.0100081-source.json
Full Text
831-1.0100081-fulltext.txt
Citation
831-1.0100081.ris

Full Text

A CONTEXTUAL EFFECT IN FEATURE DETECTION by MARCUS DAVID WOMERSLEY B.A., Simon Fraser University, 1973 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n the Department of Psychology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 1975 Rights of Publications and Loan In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the Univ e r s i t y of B r i t i s h Columbia, I agree that the Lib r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I further agree that permission f o r extensive copying of t h i s thesis for scholarly purposes may be granted by the Head of my Department or by h i s representative. It i s understood that copying or pu b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permis-sion. MARCUS DAVID WOMERSLEY September 29, 1975 Psychology Department of The University of Brit i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Abstract The question i s addressed of whether the perception of a form i s e x c l u s i v e l y determined by a p r i o r analysis of i t s elements, and two major paradigms are reviewed, v i z . , Gestalt and information processing. Three experiments were c a r r i e d out. Experiment 1 employed a s i g n a l detection task to t e s t the hypothesis that embedding a l i n e segment feature i n a unitary f i g u r a l context would f a c i l i t a t e i t s detection. The contextual e f f e c t found f a l s i f i e d the theory of a one-way causation between analysis of f i g u r a l elements and form perception. Experiment 2 showed that a necessary condition of t h i s context e f f e c t on feature detection i s the three-dimensionality of the unitary context. With bi-hemiretinal stimulus presentation Experiment 3 showed a s i g n i f i c a n t context e f f e c t i n the RVF, but not i n the LVF. Some current paradigms are applied to these r e s u l t s ; i t i s argued concurrently that the explan-at i o n of phenomena c a l l e d "perceptual" e n t a i l s the s o l u t i o n of two problems:, that of determining what constitutes such an explanation, and an adequate theory of the e x p e r i e n t i a l aspect of perceptual phenomena. These are addressed i n Appendices A arid B. Supervisor i i i TABLE OF CONTENTS Page LIST OF TABLES i v LIST OF FIGURES v Introduction 1 Experiment 1 9 Method 11 Results 14 Experiment 2 25 Method 26 Results 28 Experiment 3 39 Method 42 Results 46 Discussion 59 References 68 Appendix A , 75 Appendix B 77 i v LIST OF TABLES Table Page I Data from N = 10 subjects. Experiment 1 15 II t - t e s t f o r correlated means on E^, Experiment 1 25 III Data from N = 10 subjects. Experiment 2 29 IV t - t e s t f o r correlated means on E^, Experiment 2 39 V Data from N = 10 subjects. Experiment 3 47 VI Summary of analysis of covariance. Experiment 3 57 LIST OF FIGURES Figure Page 1 Stimuli for the signal detection task—Experiment 1 11 2 Fixation point 12 3 Stimuli for the signal detection task—Experiment 2 27 4 Stimuli for the signal detection task—Experiment 3 43 -5 Visual acuity test 44 v i Acknowledgement I thank Larry Ward of the Department of Psychology, U.B.C., for advice and assistance at every stage of th i s w o r k — p a r t i c u l a r l y for a careful and c r i t i c a l reading of what I had fondly hoped would be the f i r s t and l a s t draft; also Stan Coren of the Department of Psychology, U.B.C., for his help, and the following persons for discussions which, perhaps unbeknown to them, were pertinent: Garry Ford of the Department of Physics, U.B.C., Malcolm Greig of the Computing Centre, U.B.C., and Len Theodor of the Department of Psychology, York University. This work was p a r t i a l l y supported by Grant A9958 from NRC of Canada to Lawrence M. Ward. 1 Introduction The history of attempts to examine s c i e n t i f i c a l l y the phenomena of v i s u a l perception, and to explain them adequately, shows a number of disparate theoretical approaches. They may usefully be subsumed, however, under the headings of "macro-analysis" (or functional analysis) and "micro-analysis," to p a r a l l e l a r e l a t i v i s t i c d i s t i n c t i o n which Fodor (34) makes. The macro-analytic approach tends to be cognitively oriented. Helmholtz (49, 50) takes an active view of the perceiving organism; for him, "unconscious conclusions" and "inferences" are basic processes of perceiving, and of thinking i n general. S i m i l a r l y , l a t e r workers such as Beck (7), Hochberg and Beck (54) and Festinger et a l . (32) are led to postulate active i n f e r e n t i a l and attentional processes to account for the way sensory input i s perceived; while for the Gestalt psychologists, such as Kohler (81) and Koffka (79, 80), the important problem i n per-ception i s that of organization, and the solution to i t has to come at the l e v e l of the whole v i s u a l f i e l d , i n terms of laws of organization. Koffka (79) permits "forces i n the Ego" such as attitude a causal part i n the organization of the perceptual f i e l d (p. 149). The micro-analytic approach to perception can be termed "mechanis-t i c , " a term carrying few derogatory connotations i n psychology, despite i t s demise as an explanatory paradigm i n physics (see, for example, Einstein and Infeld 30, De Broglie 19). For Hering (51), Mach (88) and their successors (e.g., R a t l i f f 107), perceptual phenomena must have a physiological basis. The workings of the v i s u a l system cannot be explained by constructs l i k e "interpretation" or "inference," unconscious 2 or otherwise; rather processes such as reciprocal physiological i n t e r -action are held to be basic modi operandi, perception depending on the operation of complex machinery which i s p u b l i c l y observable, i n p r i n c i p l e , and (by implication) not controllable by the percipient. Both these broad frameworks, the molar and the molecular, share two defects. One i s a general f a i l u r e to explicate just what consti-tutes an explanation of a perceptual (or any other) phenomenon. The second i s that any adequate account of phenomena which have a conscious aspect w i l l have to subsume some explanation of consciousness. While the l a t t e r has been an intractable problem hitherto, i t s solution w i l l probably depend on a p a r t i a l re-orientation i n psychology. What seems to be needed i s a s c i e n t i f i c account of consciousness i n terms of an unobservable having observable consequences—hardly a unique s i t u a t i o n i n science—the "prediction of behavior," on this view, being s i g n i f i -cant only insofar as i t bears on a theoretical account of subjective states. Natsoulas (99) has recently drawn attention to the experiential element i n perception as a phenomenon to be explained. (see Appendices A and B). The issue of micro- or macro-analytic explanation i s s t i l l current; Gibson (39) poses the question of whether perception i s composed of elements, or organized into structure, and t r i e s to deal with i t by sug-gesting that invariant aspects of stimulation are "registered" by the perceptual system and that, rather than "constructing" or "organizing" input, the centers of the nervous system "resonate to information" (p. 267). Gibson's formulation seems to be based on the assumption that perception i s d i r e c t , the postulation of mediating processes not being 3 required. Neisser (101) argues, to the contrary, that elaborate pro-cesses must be required between the information i n the optic array and the perception of the environmental layout which i t speci f i e s . In the analysis of the processes by which we come to v i s u a l l y perceive a form, two paradigms are s a l i e n t — G e s t a l t theory, and the information-processing approach (of which Neisser i s one proponent). For the G e s t a l t i s t s , things look as they do because of the organi-zation to which the proximal stimulus gives r i s e (Koffka 79). The pr i n c i p l e that what happens to a part of a whole i s determined by laws inherent i n the whole i s held by Gestalt theorists to be of wide a p p l i -cation, even beyond i t s relevance to the psychology of perception. In the matter of form perception, the problem as the G e s t a l t i s t s see i t i s to specify the laws by means of which description and prediction may be made of how a set of elements i n the v i s u a l space i s organized into a f i g u r a l percept. The s c i e n t i f i c understanding of perceptual phenomena i s to come from s t a r t i n g at the l e v e l of perceptual organization, and analyzing which parts belong as parts to functional wholes. In view of l a t e r postulations of a hierarchic organization of psychological phenomena (e.g., Neisser 100, Koestler 77, 78) i t i s noteworthy that the Gestalt view of organization i s that i t i s fundamentally h i e r a r c h i c — a point which i s quite clear from Koffka's discussion of wholes and sub-wholes (80). The s i x Gestalt laws (see Koffka 79, Katz 69) may be regarded as special cases of the general law of "pragnanz," which states (as formu-lated by Koffka 79) that psychological organization w i l l always be as good as the prevailing conditions allow. "In this d e f i n i t i o n the term 'good' i s undefined. I t embraces such properties as regularity and 4 symmetry, s i m p l i c i t y and others" (p. 110). Central to the Gestalt appraoch i s the doctrine of isomorphism, formulated by Kohler (81), which asserts that our experiences have the same structure as the brain processes with which they are associated (Miles 93). Under isomorphism, the motions of atoms and molecules are i n their molar aspect i d e n t i c a l to thoughts and feelings; physiological processes, whether or not accom-panied by consciousness, are h o l i s t i c , such that the a c t i v i t i e s of the parts are determined by the whole (Koffka 79). Gestalt theory, and c r i t i c a l reaction to i t , has given r i s e to research into quantifying such f i g u r a l properties as "goodness" (Hochberg and McAlister 53), and " s i m p l i c i t y , " "good continuation" and "symmetry" (Hochberg and Brooks 55). Unlike some e a r l i e r work, such as Attneave's application (2) of information theory techniques (see also Haber and Hershenson 45), the study of Hochberg and Brooks provides a quantifica-t i o n of Gestalt laws which leads to predictions—though over a very limited domain—about perceptual organization, on the basis of the measurement of physical attributes of the stimulus. Other workers (Zusne 133, Michels and Zusne 92) propose the use of areal and perimetric moments to quantify form, seen as a two-dimensional d i s t r i b u t i o n of elements—an approach which derives some support from the attempt by Brown and Owen (11) to specify f i g u r a l organization under a set of fi v e p r i n c i p a l axes. With a few exceptions, however, attempts to apply information measurement to perceptual s t i m u l i have been a disappointment; the measured amount of information i n a v i s u a l form has f a i l e d , i n general, to predict such dependent measures as r e c a l l , recognition or speed of 5 perception (see Haber and Hershenson 45, Corcoran 17). The thrust of these studies, nonetheless, i s quite i n keeping with the Gestalt axiom— embodied i m p l i c i t l y i n the Gestalt laws—that v i s u a l form perception depends largely on the organization which the proximal stimulus d i s t r i -bution gives r i s e to, this being determined, i n turn, by external stimulation. What tends to be disregarded i n attempts to specify the variables contributing to Gestalt organization i s the place G e s t a l t i s t s assigned to the ego i n the organization of perception. While this seems to involve some complex hierarchic notions (see Koffka 79, p. 319 f f . ) , what i s pertinent here i s that Koffka c l e a r l y finds i t impossible to avoid ascribing to the observer—as opposed to the exter-n a l l y given array—forces which are causally involved i n perceptual organization (p. 149). Whatever the present value of the neurological hypotheses advanced, Gestalt theory c l e a r l y includes the notion of an active observer. In contrast to the Gestalt view that " . . . summing i s a meaningless procedure, whereas the whole-part relationship i s meaningful" (Koffka 79, p. 176), the information-processing approach (Neisser 100, Haber and Hershenson 45, Lindsay and Norman 87) seeks to analyze perception into a set of elements, which are regarded as processes of "information trans-formation." By assumption, perception i s a re s u l t of a series of events over a time i n t e r v a l divided into stages, each stage corresponding to a transformation of information. The information-processing approach may be distinguished from information theory and i t s applications, b r i e f l y mentioned previously (see discussions by Haber and Hershenson 45, Neisser 100, and M i l l e r 94). In contrast to information theory, 6 "information" tends to be given either a c i r c u l a r d e f i n i t i o n (Neisser 100, p. 8), or none at a l l (a vagueness c r i t i c i z e d by Koffka 79 as early as 1935). I t seems clear, however, that i n the case of v i s u a l percep-tion the information being processed i s the conduction of energy i n the nervous system which, at some set of hitherto unspecified points and i n completely mysterious fashion, becomes associated with a conscious per-cept. Despite the f a i l u r e of the paradigm, so f a r , to provide even a glimmering of a theory of consciousness—when the perceptual and other cognitive phenomena i t i s intended to deal with r e l i a b l y y i e l d a con-scious a s p e c t — t h i s should be taken less as a f i n a l c r i t i c i s m , perhaps, than as a problem to be solved. No more i s i t f a i r to deride the seeming ad hoc-ness of "image demons," "computational demons" and "cognitive demons" (see, for example, Selfridge 113) without regard to possible h e u r i s t i c value and supporting evidence. An information-processing model no longer i n favor i s the template model, which proposed the storage of pattern representations and their subsequent a v a i l a b i l i t y for comparison with new input; such an approach f a i l s to deal with the problem of constancy of form perception (Neisser 100). Within the several conceptual models of information processing that are current, a d i s t i n c t i o n i s usually made between s e r i a l and p a r a l l e l processing (a d i f f e r e n t i a t i o n which, however, has been c r i t i -cized by Rabbitt 106). Pandemonium models populate a conceptual nervous system with a series of demonic e n t i t i e s , such that "image demons" process i n i t i a l v i s u a l input, which passes i n turn to a set of "feature demons," thence to "cognitive demons" and f i n a l l y to a "decision demon" (Selfridge 113, Neisser 100). The amplitude of the output of each demon 7 depends on the degree to which a part i c u l a r feature i s i n i t s input. The scheme i s hierarchic, with each group of demons working i n p a r a l l e l ; the organization of the whole, however, i s by implication s e r i a l — although feature-processing systems with multiple read-off have been proposed (Ward and Wexler 124), under which the decision demon has access to the output of more than one lower l e v e l . Pandemonium-type models have gained much impetus from the pioneering physiological work of Hubel and Wiesel who, i n a series of papers, out-l i n e the v i s u a l receptive f i e l d organization of the cat (59, 60) and the monkey (61). On the evidence of s i n g l e - c e l l recordings and the putative functional architecture of the mammalian cortex, Hubel and Wiesel present a "possibly oversimplified concept" (61) of a hierarchic system of receptive f i e l d s dependent on anatomical wiring. This work has been taken to provide an explanation of how information i s processed i n the human v i s u a l system, given the additional evidence for the general simi-l a r i t y of the system to that of other mammals (Marg and Adams 89, Wein-s t e i n , Hobson and Baker 125). As described by Sekuler i n a c r i t i c a l review (112), the standard gospel i s based on an assumption of a predom-inantly s e r i a l organization of the v i s u a l system. I t teaches that receptive f i e l d structure at the bipolar l e v e l determines receptive f i e l d structure at the r e t i n a l ganglion l e v e l , which i n turn determines that at the l a t e r a l geniculate nucleus; s i m i l a r l y , receptive f i e l d organiza-tion at each l e v e l up the hierarchy i s determined by a series of c e l l u l a r convergences—of l a t e r a l geniculate onto simple c o r t i c a l , simple c o r t i c a l onto complex c o r t i c a l , complex c o r t i c a l onto hyper-complex, etc. The deficiencies of the interpretation include a f a i l u r e to recognize the 8 role i n s p a t i a l v i s i o n of midbrain structures such as the superior c o l l i -culus, the apparently non-homogeneous nature of the retino-geniculate pathway (Sekuler 112), and the more recent emphasis on the capacity of the v i s u a l system for p a r a l l e l information processing (Glezer et a l . 41, Ward and Wexler 124), for which there i s supporting physiological e v i -dence (Hubel and Wiesel 62). Pandemonium and related models provide a general model, however, of which the most important notion perhaps i s that there are units i n the v i s u a l system—"holons" i n Koest-ler's sense (78)—which are s p e c i f i c to the detection of salient aspects of a pattern (see discussion by Corcoran, 17). There i s evidence that there are c e l l s i n the mammalian cortex which respond maximally to l i n e segments with a pa r t i c u l a r orientation and i n a par t i c u l a r r e t i n a l location (Hubel and Wiesel 59). This has led to proposed explanations of v i s u a l pattern perception i n which such l i n e orientation detectors are one stage i n a hierarchic feature i d e n t i f i c a -t i o n process (Weisstein 126, Weisstein and Harris 127). Under the usual s e r i a l i t y assumption, elementary features, on this view, are detected i n the processing hierarchy as a subset of a series of information transfor-mations which i s a precondition of the perception of organized structure or pattern. According to this approach—unquestionably more i n vogue than the Gestalt paradigm—the perception of the whole i s subsequent to and dependent on an analysis of i t s parts. A p o t e n t i a l l y f a l s i f y i n g test of this theory (Popper 105) would be to manipulate the contextual pattern i n which an elementary feature, such as an oriented l i n e segment, i s embedded. On the linear causation view outlined above, the a b i l i t y to detect such a feature should not 9 vary with i t s f i g u r a l context. Weisstein and Harris (127) found that the detection of a l i n e segment oriented i n the v i s u a l space i s f a c i l i -tated i f i t i s within a unitary context. Experiment 1 was an attempt to repl i c a t e this finding, using a tachistoscopically presented v i s u a l detection task; Experiment 2 was an attempt to analyze the pertinent aspects of the unitary pattern i n producing the ef f e c t , and i n Experiment 3 the aim was to relate any finding of a contextual effect on feature detection to the burgeoning l i t e r a t u r e on functional cerebral hemispheric dominance. Experiment 1 This experiment was intended to set up conditions under which the theory that the perception of a form i s a function of a prior analysis of i t s elementary features could be f a l s i f i e d . The log i c of the approach i s that i f the above theory i s adequate, the d e t e c t a b i l i t y of a l i n e segment should not be any easier when the segment i s part of a functional whole than when i t i s part of a non-unitary context. Every attempt was made to equate the unitary and non-unitary s t i m u l i on the number of lin e s constituting the pattern, th e i r area, and the position and orientation i n the observer's v i s u a l space of both the whole pattern and the target l i n e . The measure of d e t e c t a b i l i t y was derived through signal d e t e c t a b i l i t y theory (for detailed coverage of which, see Green and Swets 42, Egan and Clarke 29, or Nachmias 98). B r i e f l y , the use of signal d e t e c t a b i l i t y theory (SDT) yields a sensitive measure of a person's a b i l i t y to detect a signal from a background of noise, and also permits a determination of the c r i t e r i o n , which i s taken to r e f l e c t variables of bias, set, attitude 10 or perception of goal (Green and Swets 39). In the fundamental SDT problem, noise (N) i s assumed always present, signal (S) i s sometimes present. A set of observations (x) i s assumed to produce two d i s t r i b u -tions of in t e r n a l states i n the observer, and two corresponding probability density functions: ff[( x) a n c* ^ g j j ^ x ^ » ^ o r ^ a n <* s + N respectively. In the detection s i t u a t i o n a decision by the observer, "S" or "NS", has four possible outcomes: H i t , SN|A; miss, SN|B: correct reje c t i o n , N|B, and false alarm, N|A—where A i s a response "SN" ("yes", "s i g n a l " ) , and B i s a response "N" ("no", "no si g n a l " ) . According to some optimum, the observer finds a balance among these four p r o b a b i l i -t i e s , giving two pertinent conditional probabilities:pp(yes|signal + noise) = Pg^(A), h i t ; and p (yes|noise alone) = p^(A), false alarm. These correspond to areas under the two density function curves to the right of the c r i t e r i o n , and the proportions of h i t s and false alarms obtained i n the experiment are used as estimates of P^(A) and Pg^(^) a m * to calculate the index of d e t e c t a b i l i t y . The SDT model implies that sensory information i s p r o b a b i l i s t i c , detection being an active perceptual process. The observer i s given a sample and two hypotheses each determining a sampling d i s t r i b u t i o n , and must make a "best choice" between them according to some optimum. Two usual assumptions are the normality and homogeneity of variance of the two d i s t r i b u t i o n s ; for this experiment, such r e s t r i c t i v e assump-tions about the underlying density functions seemed un j u s t i f i e d (see Pastore and Scheirer 103), and a non-parametric index of d e t e c t a b i l i t y , NE, derived by Simpson and F i t t e r (114) after Sakitt (111) was used. For the purposes of this experiment two indices E. were derived for each 11 subject; this yielded a measure of the subject's a b i l i t y to detect the target i n both a unitary and a non-unitary context. In addition, each subject made confidence ratings, corresponding to the use of f i v e c r i t e r i a rather than one—a device which increases the s e n s i t i v i t y of the measure. Method Subjects. Subjects were 10 members of the University community, 8 females and 2 males, a l l adults. A l l had normal or corrected v i s i o n (as determined by their a b i l i t y to eas i l y see the s t i m u l i i n pretesting) and a l l volunteered for the experiment, which comprised one session of approximately 40 minutes. Apparatus. The four s t i m u l i for the signal detection task were based on, but incorporate some changes from, those used by Weisstein and Harris (127). They are shown i n figure 1. Unitary l e f t -oriented ~1 Unitary r i g h t -oriented Non-unitary l e f t oriented Non-unitary r i g h t -oriented Figure 1. Stimuli for the signal detection task—Experiment 1 They were drawn on white cards and within 2.5 degrees v i s u a l angle of a f i x a t i o n point (figure 2), so that v i s i o n was predominantly foveal. Each 12 stimulus contained one target l i n e , oriented either (up to the) l e f t or right at a 45 degree angle from the v e r t i c a l . A l l s t i m u l i , unitary (C^) or non-unitary (Vy) > comprised the same set of l i n e elements—six one-unit l i n e s , two half- u n i t l i n e s and a dot, i n addition to the target l i n e . The f i x a t i o n point with masking stimulus surround was made i n a similar fashion (figure 2). The masking stimulus was to reduce the a v a i l -a b i l i t y of the icon, so that performance would ' not be a function of a subject's capacity for . ______ vis u a l imagery. The s t i m u l i were presented i n a Gerbrands Figure 2. two-channel tachistoscope. The f i x a t i o n point Fixation point, was i n f i e l d one; the st i m u l i were presented i n f i e l d two, and designed such that for the l e f t oriented targets the target l i n e segments and their contexts were i n the same position i n r e l a t i o n to the f i x a t i o n point, and covering the same area i n the v i s u a l space. This also held for the right oriented target l i n e segments, which were mirror images of the l e f t oriented ones. Procedure. The experiment was carried out i n a dimly-illuminated room, and at the start of each session the subject was dark-adapted for approximately 10 minutes. After having been shown the tachistoscope, the subject was read the following instructions: This i s a task i n v i s u a l perception. You w i l l be presented with a series of s t i m u l i consisting of li n e s i n a pattern. A l l these l i n e s w i l l be v e r t i c a l or horizontal except one, which you w i l l try to detect; i t w i l l slope up to the l e f t or up to the ri g h t . Here are some examples. (Subject i s shown several.) On each t r i a l you w i l l see a f i x a t i o n point, also surrounded by l i n e s . (This i s displayed.) When I say 13 " f i x a t e " look at th i s point, then say "ready," and continue f i x a t i n g u n t i l you see the stimulus. Then say " l e f t " or "r i g h t " depending on which way you think the slanted l i n e slopes up, and "one," "two" or "three" to indicate your confidence i n your judgement. Say "one" i f you are uncer-t a i n , "two" i f you are f a i r l y confident, or "three" i f you are quite confident about your judgement. (Subject.is made fami l i a r with the routine.) Please note that i t i s very important that you keep looking at the f i x a t i o n point after you have said "ready." Every twenty t r i a l s or so we'll pause -forra rest. Are there any questions? Then followed some practice t r i a l s , during which the exposure time was manipulated u n t i l the subject was responding with approximately 70% accuracy. The tachistoscopic presentation time (the same for a l l stimuli) was thus b r i e f enough to ensure an adequate degree of "noise" for the signal detection task, and also to e f f e c t i v e l y rule out saccadic eye movements. Saccadic reaction time i s t y p i c a l l y about 200 ms (Alpern 1)—White, however, gives a figure of 150 ms (129)—while, for this study, exposure times for a l l subjects were between 17 and 44 ms. The importance of f i x a t i n g on the f i x a t i o n point after the "ready" was emphasized, and no subject reported any d i f f i c u l t y i n so doing. Illumin-ation of the st i m u l i and the f i x a t i o n point was not varied, being constant for a l l subjects under a l l conditions. The tachistoscope was operated i n a mode such that the f i x a t i o n point (and masking f i e l d ) was on constantly, except during the b r i e f exposure of the stimulus. Thus, for each t r i a l , the subject would be to l d " f i x a t e " ; the response of "ready" would be followed, after a b r i e f indeterminate pause, by one of the target s t i m u l i which would replace the f i x a t i o n point i n the sub-ject's v i s u a l space, tojbeereplaced i n i t s turn by the f i x a t i o n point and i t s surround, which was large enough to mask the target stimulus. The subject then verbalized a decision on the direction of the target l i n e 14 segment and gave a confidence rating. Each of the 4 s t i m u l i was pre-sented on 24 t r i a l s for a t o t a l of 96 t r i a l s , done i n irregular order, for each subject. Results For each context C\ the number of responses under the ratings were summed, then the sums accumulated to get a cumulative d i s t r i b u t i o n ; by the arbitrary selection of the l e f t target l i n e orientation as "si g n a l " (the right orientation then being "noise"), t h i s becomes the Receiver Operating Characteristic (ROC) curve. The accumulation over confidence ratings of responses to a l e f t oriented target l i n e , therefore, corre-sponds to the " h i t rate," while the "false alarm rate," i n the usual terminology, i s given by the accumulation of responses to a righ t oriented target. Table I shows the following for each subject: the number of responses i n the rating categories for both targets and under each contextual con-d i t i o n , the derived points of the ROC curve as proportions, a graph of the curve, and the index of detection performance. This index i s given by: E = (^ - y [2/(4L + s * ) ] 1 ! 2 where and are the means of the grouped data for l e f t oriented and right oriented targets respectively, and s 2 , s 2 are the variances. As a Li R non-parametric index E i s equivalent to the ROC subjacent area (Simpson and F i t t e r 107). A t-test for correlated means (Ferguson 29) was performed on the E^ (see table II.'.). Table I. Data from N - 10 subjects. Experiment 1. Subject 1. unitary context non-unitary context Response freq ROC Left Right Left Right ! Rating target target target target 5 L-3 21 0 .875 0 L-2 1 1 .917 .042 .& L-1 1 0 .958 .042 -•<-R-1 1 0 1.0 .042 R-2 0 1 1.0 .083 . i • R-3 0 22 1.0 1.0 24 24 o < L-3 17 1 .708 .042 I •« a. L-2 0 3 .708 .167 . k L-1 3 1 .833 .208 R-1 1 3 .875 - .333 '• R-2 0 6 .875 .583 . 7 ' R-3 3 - 10 1.0 1.0 24 24 O = 5.7742 E_ = 1.5904 °2 Subject 2. Response freq. ROC  Left Right Left Right Rating target target target target L-3 19 1 .792 .042 2 L-2 2 2 .875 .167 M unitary context \'\ I ? ^° - 5 0 0 . R-l 0 7 1.0 .791 R-2 0 2 1.0 .875 R-3 • 0 3 1.0 1.0 24 24 L-3 9 0 .375 0 *^ L-2 5 0 .583 0 / L - l 4 2 .750 .084 non-unitary context R _ ± g 1 & > 7 5 Q R-2 0 4 1.0 .917 R-3 0 2 1.0 1.0 24 24 Subject 3, ^ unitary context Response freq, ROC Left Right Left Right Rating target target target target L-3 17 4 .708 .167 L-2 3 6 .833 .417 L - l 3 7 .958 .708 R-l 1 6 1.0 .958 R-2 0 1 1.0 1.0 R-3 0 0 1.0 1.0 24 24 2 non-unitary context L-3 7 5 .292 .208 L-2 3 3 .417 ..333 L - l 11 6 .875 .583 R-l 2 5 .958 .792 R-2 1 2 1.0 .875 R-3 0 3 1.0 1.0 24 ' 24 Subj ect 4. C. unitary context C„ non-unitary context Response f req. ROC Left Right Left Right Rating target target target target t L-3 19 0 .792 0 L-2 2 0 .875 0 L-1 3 2 1.0 .083 R-1 0 0 1.0 .083 " R-2 0 4 1.0 .250 R-3 0 18 1.0 1.0 24 24 D 1 « i L-3 19 0 .792 0 £ L-2 1 0 .833 0 L-1 3 5 .958 .208 x. R-1 1 5 1.0 .417 R-2 0 6 1.0 .667 R-3 0 18 1.0 1.0 v 24 24 o E = 5.3374 E = 3.1961 °2 Subject 5, ^ u n i t a r y context Response f r e q . ROC Left Right L e f t Right Rating target target target target L-3 19 2 .792 .083 L-2 3 8 .917 .417 L - l 1 2 .958 .500 R-l 0 3 .958 .625 R-2 0 5 .958 .833 R-3 1 24 4 24 1.0 1.0 . non-unitary context L-3 13 0 .542 0 L-2 4 1 .708 .042 L - l 3 6 .833 .292 R- l 3 13 .958 .833 R-2 1 24 • 4 24 1.0 - 1.0 E C 1 = 1.4726 p„cM E = 1.7932 °2 I/O Sub j ect 6 unitary context Response freq. ROC Left Right Left Right Rating target target target target L-3 13 4 .542 .167 L-2 ; 2 5 .625 .375 L-1 4 2 .792 .458 R-1 2 6 .875 .708 R-2 2 4 .958 .875 R33 1 24 3 24 1.0 1.0 non-unitary context L-3 5 1 .208 .042 * % L-2 6 7 .458 .333 L-1 5 6 .667 .583 R-1 5 7 .875 .875 4 R-2 2 0 .958 .875 R-3 1 24 3 24 1.0 1.0 Er = .7248 °1 E_ = .3272 °2 Sub j ec t 7 . unitary contc::t non-unitary context Resoonse freq. ROC Left Right Left Right Rating target target target target L-3 21 4 .875 .167 L-2 2 6 .958 .417 L - l 0 3 .958 .542 R - l 1 3 1.0 .667 R-2 0 4 1.0 .833 R-3 0 4 1.0 1.0 24 24 L-3 7 0 .292 0 L-2 3 1 .417 .042 L - l 1 3 .458 .167 R-l 5 9 .667 .542 R-2 8 11 1.0 1.0 R-3 0 0 1.0 1.0 24 24 = 1.6074 E = .8Q2£ 1 unitary context Subject 8  Response freq. ROC Left Right Left Right Rating target target target target L-3 3 0 .125 0 L-2 11 0 .583 0 L-1 6 7 .833 .292 R-1 4 6 1.0 .542 R-2 0 8 1.0 .875 R-3 0 3 1.0 1.0 24 24 non-unitary context L-3 0 0 0 L-2 0 0 0 L-1 6 0 .250 R-1 14 13 .833 R-2 4 11 1.0 24 24 0 0 0 .542 1.0 Subject 9 unitary context Response freq, Left Right Left Rating target target target L-3 5 0 .208 L-2 14 0 .792 L - l 5 0 1.0 R-l 0 10 1.0 R-2 0 13 1.0 R-3 0 1 1.0 24 24 C. non-unitary context L-3 1 0 .042 L-2 13 0 .583 L - l 10 3 1.0 R-l 0 19 1.0 R-2 0 2 1.0 R-3 0 0 1.0 24 ' 24 Right target <\ 0 3 0 •« 0 .417 .958 •« 1.0 E = 4.2408 L l *• .« i.« 0 0 .125 .917 J 1.0 ' 1.0 E = 3.0273 °2 f-CA) N3 unitary context non-unitary context Subject 10 Response freq. - ROC Left Right - Left Right la t i n g target target target target <v L-3 20 1 .833 .042 i L-r2 1 4 .875 .208 •» L-1 2 0 .958 .208 t R-rl 1 3 1.0 .333 R-2 0 10 1.0 .750 + R T 3 0 6 1.0 1.0 24 24 0 5" X L-3 2 0 .083 o L-2 8 1 .417 .042 L-1 6 13 .667 .583 * R-1 7 5 .958 .792 R-2 1 4 1.0 .958 R-3 0 1 1.0 1.0 -J 24 24 25 H, 0 D = 0 a = .05 Test s t a t i s t i c t = D/s-C r i t i c a l region: { t < t a/2, 9 = -2.262 { { { t > t 1 - a/2, 9 = 2.262 t = 3.013 p(t = 3.013) < .05 H i s rejected Table I I . t - t e s t f or correlated means on E., Experiment 1 The s i g n i f i c a n t difference between the means of E^ under the two conditions, for N = 10, indicates a strong contextual e f f e c t on the d i s c r i m i n a b i l i t y of the target l i n e segment; the f a c i l i t a t o r y e f f e c t on detection performance of the unitary f i g u r a l context i s c l e a r from the ROC graphs i n Table I. This i s s u f f i c i e n t to f a l s i f y the theory that the perception of a form i s determined e x c l u s i v e l y by the p r i o r analysis of i t s elements since, under the conditions p r e v a i l i n g i n the experiment, the l a t t e r has been shown to depend on the former. The unitary s t i m u l i i n Experiment 1 contained a superposition cue giving them a f a i r l y convincing three-dimensional appearance (see f i g u r e 1). In Experiment 2 a d i f f e r e n t i a t i o n was made between unitary and Experiment 2 26 three-dimensional information to discover i f the l a t t e r i s a necessary condition for the contextual effect on feature detection, or whether a unitary context alone i s s u f f i c i e n t . As i n the previous experiment, the d i s c r i m i n a b i l i t y of a l i n e segment feature was tested under two c o n d i t i o n s — a unitary and a non-unitary context. Again, the target feature was a l i n e segment oriented at a 45 degree angle; but the unitary figure was designed to minimize depth information, and to be interpreted as two-dimensional. To t h i s end, both the superposition cue and the dot were omitted. A l l the s t i m u l i were designed to be equivalent i n such properties as number of constituent l i n e segments, area, and position and orientation i n the observer's v i s u a l space. A signal detection task was employed, and conditions were generally as those i n Experiment 1. Method Subjects. Subjects were 10 members of the University community, 9 females and 1 male, a l l adults. A l l had normal or corrected v i s i o n (as determined by their a b i l i t y to e a s i l y see the s t i m u l i i n pre-testing), and a l l volunteered for the experiment, which comprised one session of approximately 45 minutes. Apparatus. The s t i m u l i for the signal detection task are shown i n figure 3. 27 C2 Non-unitary r i g h t -oriented Figure 3. Stimuli for the s i g n a l detection task—Experiment 2. They were drawn on white cards and within 2.5 degrees v i s u a l angle of the f i x a t i o n point, so that v i s i o n was predominantly foveal . As i n Experiment 1, each fi g u r e contained one target l i n e segment, oriented (up to the) l e f t or r i g h t at a 45 degree angle from the v e r t i c a l . A l l s t i m u l i , unitary (C^) or non-unitary (C^) comprised the same set of l i n e elements—four one-unit l i n e s and ei'ght h a l f - u n i t l i n e s (plus the target l i n e ) . The f i x a t i o n point and masking surround were the same as i n Experiment 1 (see fig u r e 2). The s t i m u l i , again, were designed and presented such that for each o r i e n t a t i o n , target l i n e segments and both context patterns were i n the same p o s i t i o n i n r e l a t i o n to the f i x a t i o n point, and covering the same area i n the subject's v i s u a l space. In each context, the l e f t and r i g h t oriented patterns were mirror images of each other. Procedure. The experiment was c a r r i e d out i n a dimly illuminated room, each subject being dark-adapted f o r approximately 10 minutes at the beginning of each session. After having been shown the tachistoscope, the subject was read the same in s t r u c t i o n s as those of Experiment 1. Pra c t i c e t r i a l s were used to get an approximately 70% accuracy rate f o r F i l l Unitary l e f t -oriented Unitary r i g h t -oriented Non-unitary l e f t -oriented 28 each subject, and the range of tachistoscopic exposure times was w e l l within saccadic reaction time (24 to 44 ms). The illumination of both s t i m u l i and f i x a t i o n point was held constant for a l l subjects under a l l conditions. The presentation procedure was l i k e that of Experiment 1; the b r i e f presentation of the target l i n e segment and context was masked by the pattern around the f i x a t i o n point, and the subject verbalized a decision about the target and gave a confidence r a t i n g . Each of the 4 sti m u l i was presented on 24 t r i a l s for a t o t a l of 96 t r i a l s per subject, done i n irre g u l a r order. Results The results are tabulated i n Table If-I. For the two contextual con-d i t i o n s , the accumulation for each subject of responses over confidence ratings yields the points of the ROC curve, i t s graph, and the index of detection performance E_^ . A t-test for correlated means was performed on the E., as shown i n Table IV. I t can be seen from the result of th i s l test and from the ROC curves that the absence of depth information i s s u f f i c i e n t to n u l l i f y the contextual effect on l i n e segment d e t e c t a b i l i t y . Table I I I Data from N = 10 subjects. Experiment 2 Subject 1 Response freq ROC C^ unitary content C_ non-unitary context Left Right Left Rating target target target L-3 0 0 0 L-2 0 0 0 L - l 11 13 .458 R-l 12 9 .958 R-2 •0 2 .958 R-3 1 0 1.0 24 24 L-3 0 0 0 L-2 0 0 0 L - l 9 9 .375 R-l 12 13 .875 R-2 3 2 1.0 R-3 0 0 1.0 24 24 Right target 0 i 0 •« .542 .4 .917 1.0 •• 1.0 E. = -.1217 °1 .0641 Subject 2 Response freq. ROC Left Right Left Right Rating target target target target L-3 2 1 .083 .042 L-2 6 1 .333 .083 L-1 0 1 .333 .125 R-1 2 1 .417 .167 R-2 6 5 .667 .375 R-3 8 15 1.0 1.0 24 24 C. unitary context " j " • - , J J • J - A J . t C2 non-unitary context R ^ L-3 5 1 .208 .042 L-2 6 5 .458 .250 L-1 4 7 .625 .542 •* 2 .750 .625 R-2 4 8 .917 .958 R-3 2 1 1.0 1.0 24 24 E_ = .6362 °1 1 0 V 1.0 E p = .3555 L2 o Subject 3 Response freq. ROC  Left Right Left Right Rating target target target target •. L-3 3 0 .125 0 \ M l L-2 12 0 .625 0 L - l 9 1 1.0 .042 *j unitary context R_. Q ^ 1 Q ^ R-2 0 10 1.0 1.0 R-3 0 0 1.0 1.0 *1 24 24 3' 0 I .042 .125 -H non-unitary context R_]_ - -- l Q > 5 8 3 1.0 1.0 AJ L-3 1 0 .042 L-2 20 1 .875 L - l 3 2 1.0 R-l 0 11 1.0 R-2 0 10 1.0 R-3 0 0 1.0 24 24 Subject 4 Response freq. ROC  Left Right Left Right Rating target target target target *>•: L-3 18 7 .750 .292 i L-2 2 2 .833 .375 _ L-1 0 1 .833 .417 unitary context R _ ± ^ ± ^ ^ R-2 1 6 .958 .708 R-3 1 7 1.0 1.0 24 24 . -0A2i t 1 . .125 .208 . non-unitary context R _ x 3 7 . 8 7 5 . 5 0 o.J .750 1.0 L-3 15 1 .625 L-2 2 2 .708 L-1 1 2 .750 R-1 3 7 .875 R-2 3 6 1.0 R-3 0 6 1.0 24 24 unitary context Response freq. Left Right Rating target target L-3 0 0 L-2 20 16 L - l 2 3 R-l 1 4 R-3 1 1 . 24 24 Subject 5 ROC Left Right target target 0 0 .833 .667 .917 .792 .958 .958 1.0 1.0 non-unitary context L-3 0 0 . 0 0 L-2 7 2 .292 .083 L - l 4 4 .458 .250 R-l 12 17 .958 .988 R-2 . 1 1 1.0 1.0 R-3 0 0 1.0 1.0 24 24 I.* AA1 E r = .3454 .5002 00 Subject 6 C. unitary context C_ non-unitary context Response freq. -ROC Left Right Left Right ^ Rating target target target target <,,. L-3 0 0 0 0 2 L-2 1 4 .042 .167 •* L - l 5 3 .250 .292 v R-l' 9 8 .625 .625 R-2 7 9 .917 1.0 R-3 2 0 1.0 1.0 24 24 0 L-3 0 O \j 0 0 4t L-2 1 0 .042 0 L - l 4 0 .208 0 R-l 6 10 .458 .417 + R-2 9 9 .833 .792 R-3 4 5 1.0 1.0 *• 24 24 0 . u n i t a r y context Subject 7  Response f r e q . ROC L e f t Right L e f t Rating target target target L-3 14 2 .583 L-2 1 1 .625 L-1 1 1 .667 R-1 2 1 .750 R-2 1 2 .792 R-3 5 17 1.0 24 24 non-unitary context L-3 16 3 .667 L-2 1 1 .708 L-1 1 2 .750 R-1 3 1 .875 R-2 1 5 .917 R-3 2 '12 1.0 24 - 24 Right target .083 .125 *| .167 .208 .292 .> 1.0 E = 1.3361 C l .125 .167 .250 -M .292 .500 1.0 E r = 1.4542 °2 unitary context Response freq, Left Right Rating target target L-3 0 0 L-2 10 2 L-1 12 13 R-1 2 7 R-2 0 2 R-3 0 0 24 24 non-unitary context L-3 0 0 L-2 9 0 L-1 12 16 R-1 3 8 R-2 0 0 R-3 0 0 24 24 Subject 8 - ROC Left Right = 1.0026 = .9943 as Subject 9 unitary context non-unitary context Response freq. Left Right Rating target target L-3 8 3 L-2 3 3 L - l 0 2 R-l 0 1 R-2 4 3 R-3 9 12 24 24 L-3 2 3 L-2 3 6 L - l 0 1 R-l 1 1 R-2 6 5 R-3 12 • 8 24 24 ROC ?,« 1 Left Right target target .333 .126 .458 .250 .458 .333 .458 .375 .625 .500 1.0 1.0 o ..083 .126 .208 .375 .208 .417 .250 .458 .500 .667 1.0 1.0 a Subject 10 unitary context Response f req . ROC Left Right Left Right lating target target target target L-3 19 5 .792 .208 L-2 0 1 .792 .250 L-1 0 0 .792 .250 R-1 3 3 .917 .375 R-2 0 0. .917 .375 R-3 2 15 1.0 1.0 24 24 non-unitary context L"-3.: 16 1 .667 .042 L-2 1 0 .708 .042 L-1 0- 2 .708 .125 R-1 2 12 .791 .625 R-2 4 1 .958 .667 R-3 1 24 : 8 24 1.0 1.0 3 E_ = 1.4568 C l E = 1.4865 C2 CO 39 H Q : D = 0 a = .05 Test s t a t i s t i c t = D/s-C r i t i c a l region: {{ t 9 " " 2 - 2 6 2 { t y H-a/2, 9 " 2'262 t = -.1477 p(t = -.1477) > .05 H i s not rejected Table IV. t-test for correlated means on E^, Experiment 2 Experiment 3 There i s a body of evidence pointing towards a functional asymmetry of cerebral hemispheres for cognitive processes. This evidence for d i f f e r e n t i a l s p e c i a l i z a t i o n comes from several different areas, and w i l l be p a r t i a l l y reviewed. For the purposes of c l a r i t y , "major hemisphere" w i l l refer to that hemisphere considered to be associated with speech function, andf'minor hemisphere"swill refer to the other. References to " l e f t hemisphere" and "right hemisphere" tend to be ambiguous i n this context since, while about 95% of righ t handers have a l e f t major hemi-sphere, this holds for only about 65% of l e f t handers and ambidexters (Witelson 132). Neuroanatomical asymmetries of hemispheres have been reported i n 40 neonates (Geschwind 38, Witelson 132) and adults'(Witelson 132), and such asymmetries have been related to hemispheric s p e c i a l i z a t i o n (Hyde et a l . 63, Geschwind and Levitsky 37, LeMay and Culebras 83). The effects of a u n i l a t e r a l brain lesion vary with the side, not only the s i t e within a hemisphere; the t r a d i t i o n a l view that the major hemisphere was dominant for a l l complex cognitive processes was refuted by the results of c l i n i c a l studies of patients with w e l l - l a t e r a l i z e d brain lesions (see Milner 95). These patients showed more severe perceptual disorders after right sided lesions than l e f t sided ones (Paterson and Zangwall 104, Kimura 72); s p a t i a l d e f i c i t s i n tactual tasks r e s u l t from right posterior damage (Milner 95). A long series of studies of cerebral commissurectomy patients has been done, where a direct comparison can be made of each hemisphere on each task (see Sperry 117, 118; Sperry et a l . 119; Gazzaniga 35, 36). These increasingly support the notion of a basic s p e c i a l i z a t i o n i n hemi-spheric organization (Sperry 117). The disconnected minor hemisphere i s found to be superior for the recognition of faces and nondescript figures as whole patterns, for dealing with s p a t i a l and part-whole relationships, for non-verbal thinking.and direct perceptual transformation. The major hemisphere i s found to have a functional advantage for symbolic trans-formation, verbal communication, l i n g u i s t i c and numeric processing, sequential and analytic thinking, conceptual symbol recoding and the direction of motor a c t i v i t y i n general (Sperry 117). Research with normal subjects has provided another set of presump-ti v e evidence for hemispheric s p e c i a l i z a t i o n . In audition, dichotic l i s t e n i n g studies reveal a consistent right ear advantage for verbal 41 material (Kimura 74) and a l e f t ear superiority for the recognition of non-verbal sounds (Knox and Kimura 76, also see Milner 95). In v i s i o n , a right v i s u a l f i e l d (RVF) superiority has been found for the perception of alphabetic material (Kimura 75, H i l l i a r d 52), while the l e f t v i s u a l f i e l d (LVF) i s reported superior for recognition of faces ( R i z z o l a t t i et a l . 108, H i l l i a r d 52), location of dots (Kimura 75), discrimination of l i n e slopes (Durnford and Kimura 25), and stereoscopic depth percep-tion (Kimura 74). Cohen (18), reporting v i s u a l f i e l d differences i n reaction times, concludes that for alphabetic material, the l e f t (major) hemisphere processes predominantly i n s e r i a l , the other i n p a r a l l e l . Inferences from an observed v i s u a l f i e l d difference to a functional hemispheric dominance can be problematical (see White 129, 130). These studies, however, and others (e.g., Dimond and Beaumont 20, 22, 23; Levy 84; Levy et a l . 85; Bakan 3, 4; Bakan and Streyer 5) lend support to the general notion that while the major hemisphere i s specialized for verbal and other symbolic functions, the minor hemisphere i s associated with s p a t i a l form perception and h o l i s t i c processing i n general. Given t h i s h e u r i s t i c of l o c a l i z a t i o n of cognitive mode—and p a r t i c -u l a r l y the reports of Durnford and Kimura (25) and Kimura '('74) pertaining to l i n e slope discrimination and depth perception—the t h i r d experiment was designed to replicate i n each v i s u a l f i e l d the finding of Experiment 1. The hypothesis tested i s that the difference i n the d e t e c t a b i l i t y of a l i n e segment under the two conditions (unitary three-dimensional and non-unitary) w i l l be greater i n the LVF than i n the RVF. On the basis of the above discussion, a related f a l s i f i a b l e prediction i s that the LVF w i l l show superiority i n performance under both unitary and non-unitary 42 conditions, as compared with the RVF. The attempt was made, therefore, to see i f the contextual effect on feature detection found i n Experiment 1 could be related to a hemispheric l a t e r a l i t y difference. A signal detection task was again employed, from which four indices E^ could be derived—one each for unitary and non-unitary context i n each v i s u a l f i e l d . The target feature was a l i n e segment oriented at 45 degrees and, as before, the s t i m u l i were designed to be equivalent i n the number of constituent l i n e segments, area, and position and orientation i n the v i s u a l space. Cognizance was taken of White's c r i t i c i s m s (129, 130) of l a t e r a l i t y inferences i n v i s u a l perception. In an attempt to avoid the confounding of response determinants with analysis determinants, a verbal response by the s u b j e c t — p u t a t i v e l y associated with the l e f t hemisphere— was ruled out. Instead, a bimanual s i g n a l l i n g system was employed, which was also intended to control as far as possible for contralateral hemi-spheric effects i n the i n i t i a t i o n of motor a c t i v i t y . Because of the impracticality of designing a bimanual s i g n a l l i n g system adequate for c o l l e c t i n g information on confidence l e v e l s , the signal detection task required a binary choice, corresponding to one c r i t e r i o n . The p o s s i b i l i t y that ocular acuity dominance would confound a l a t e r a l i t y difference attributable to hemispheric asymmetry (Hayashi and Bryden 47, Kershner and Gwan-rong Jeng 70) was recognized, and a monocular test of v i s u a l acuity was used. Method Subjects. Subjects were 10 members of the University community, 8 females and 2 males, a l l adults. A l l had normal or corrected v i s i o n , 43 and they were paid $3.00 an hour to take part i n the experiment, which comprised one session of approximately one hour. Apparatus. The s t i m u l i , tachistoscopically presented as i n Exper-iments 1 and 2, are shown i n figure 4. Right Visual F i e l d "1 Unitary l e f t -oriented Unitary r i g h t -oriented Non-unitary l e f t -oriented Non-unitary r i g h t -oriented Left Visual F i e l d Unitary l e f t -oriented "1 Unitary right oriented Non-unitary l e f t -oriented Non-unitary r i g h t -oriented Figure 4. Stimuli for the signal detection task—Experiment 3. For the RVF set, the f i x a t i o n point (see figure 2) was s l i g h t l y to the l e f t of each stimulus pattern, and for the LVF set s l i g h t l y to the r i g h t . When the point i s fixated, material i n each VF goes i n i t i a l l y to the 44 contralateral hemisphere (Sperry 117). The st i m u l i were designed and presented under the same st r i c t u r e s as those of the f i r s t two experiments i n regard to number of constituent l i n e segments, and orientation, area and r e l a t i v e position i n the subject's v i s u a l space; i n addition, the RVF and LVF sets were mirror images within contexts. The monocular v i s u a l acuity test i s shown i n figure 5. It was also tachistoscopically pre-sented. Procedure. Conditions i n general were as those of Experiments 1 and 2, except as noted. After p a r t i a l dark adaptation (10 minutes), each subject was read the following instructions: This i s a task i n v i s u a l perception. You w i l l be presented with a series of st i m u l i consisting of lin e s i n a pattern. A l l these lines w i l l be v e r t i c a l or horizontal except one, which you w i l l try to detect; i t w i l l slope up to the l e f t or up to the r i g h t . Here are some examples.( (Subject i s shown several examples.) On each t r i a l you w i l l see a f i x a -t i o n point, also surrounded by l i n e s . (This i s displayed.) When I say " f i x a t e " look at th i s point, then say "ready" and . continue f i x a t i n g u n t i l you see the stimulus. When you have decided which way the slanted l i n e slopes up, p u l l one of the two response indicators, using both hands—the l e f t one to indicate " l e f t , " the ri g h t one to indicate " r i g h t . " (Subject i s shown thei r use.) You should not try to verbalize your response. I t i s also very important that you keep look-ing at the f i x a t i o n point after you have said i'r.eady." Every 20 t r i a l s or so we'll pause for a rest. Are there any questions? Practice t r i a l s were used to obtain an accuracy rate of approxi-mately 70%, and th i s established a tachistoscopic presentation time for a l l subjects which was within saccadic reaction time (range: 24-120 ms). In a review a r t i c l e , White (129) doubts the a r t i f a c t u a l nature of eye movements as an important component of l a t e r a l i t y differences. In th i s T O Z L E P D P E C F D E DFCZP F E L O PZD 0 E F P O T E C l _ E F O D P C T Z Figure 5. Visual acuity test. 45 experiment, however, an absence of l e f t or right eye movements was con-sidered important, so that I strongly emphasized to each subject the need to keep f i x a t i n g after having said "ready"; subjects seemed to experience no d i f f i c u l t y i n so doing. Each subject indicated his or her judgement on the direc t i o n of the target l i n e orientation by means of one of two manually operated s i g n a l -l i n g systems; these were designed to be awkward i n use except bimanually, * and were placed i n front of the subject, to the l e f t and right respec-t i v e l y and equidistant from the subject. After being shown them, nobody seemed to have d i f f i c u l t y i n using them correctly. There were four experimental conditions i n a 2 x 2 repeated measures design: 2 contexts by 2 v i s u a l f i e l d s . Each subject had 24 t r i a l s on each of the 2 l i n e orientations under each condition, for a t o t a l of 192 t r i a l s i n irre g u l a r order. After completing the main experiment, subjects were run i n the monocular acuity test, for which the following instructions were given: This i s a monocular test of v i s u a l acuity. When you see the card, read the l e t t e r s from l e f t to right at the rate of about one per second, starting at the top.- (Left or right eye i s tested.) Now, as quickly as you can, count backwards by threes from 21. (This i s done.) This time, when you see the card, read the l e t t e r s from right to l e f t , s t arting at the top. (The other eye i s tested.) Each eye was tested separately i n the tachistoscope, the counting task after the f i r s t eye test being intended to eliminate any effect, of short-term memory of the test card for the second eye test. 46 Results Table V shows the data for each subject under both contextual conditions i n each v i s u a l f i e l d . The index of acuity dominance, A, was simply the r a t i o given by the number of correct responses for the right eye divided by the number for the l e f t . The general l i n e a r model was applied with an analysis of covariance on the E^, the eye acuity index being treated as a covariate (Myers 97). The appropriate l i n e a r equation i s : y. . = y + a. + 3, + Y + aB., + ay. + By, + aBY., + b(X. . - X..) + e. . ' i j J k m p j k 'jm M'km 'jkm xi xj where a l l effects are f i x e d , except subjects (random), and a. i s the main v i s u a l f i e l d e f f e c t , J B, i s the main context e f f e c t , k Y i s the subjects e f f e c t , m interactions are as usually defined, b(X^j - X..) i s the error i n y predictable on the basis of knowing X, the covariate, e. . i s the residual error. The analysis i s summarized i n Table VI, together with the adjusted group means. Under the conditions of Experiment 3, neither the v i s u a l f i e l d effect nor the contextual effect i s s i g n i f i c a n t . There i s a trend towards an interaction between v i s u a l f i e l d and context, however. To compare average performance within each v i s u a l f i e l d , two orthogonal contrasts were plan-ned (Myers 97). Table V. Data from N = 10 subjects. Experiment. 3. C. unitary context RVF C unitary context LVF C_ non-unitary context LVF Binary choice L R Response freq. Left Right target target C„ non-unitary context L RVF R L R L R 20 4 24 8 16 24 1 23 24 2 22 24 21 3 24 16. 24 1 23 24 0 24 24 Subject 1 ^ ROC I Left Right' target target .875.-.833 1.0 .333 1.0 .042 1.0 .083 1.0 1.0 .333 1.0 •* .042, 1.0 0 1.0 •<1 .« l . o -.1159 .4173 Covariate R/L eye 2.750 unitary context RVF • non-initary context RVF . unitary context LVF Subject 2 Binary choice L R Response freq, L R L R non-unitary context L LVF R Left target 24 0 24 21 3 24 21 3 24 23 1 24 Right target 1 23 24 3 21 24 4 20 24 0 24 24 ROC Left" target 1.0 1.0 .875 1.0 .875 1.0 .958 1.0 Right target .042 1.0 .125 1.0 .167 1.0 0 1.0 * A- .4 -t ft.'*)" c. 1 6.6369 2.2203 2.0933 6.6366 Covariate R/L eve .9130 • * •& •« |.o MA) co Subject 3 ^ Response freb. ROC  Binary Left Right Left Right A i ^ choice target target target target unitary context L 24 1 1.0 .042 RVF R . 0 _23 1.0 1.0 24 24 non-unitary context L 22 2 .917 .083 A RVF R 2 _22 1L0 1.0 24 24 unitary context L 23 0 .958 0 LVF R 1 _24 1.0 1.0 24 24 i A non-unitary context L 24 1 1.0 .042 LVF R 0 _23_ 1.0 1.0 24 24 Note: this _S, with perfect performance i n C^  RVF and C. LVF, i s assigned one error i n each condition to obviate undefined E.-. C- un i t a r y context RVF Binary choice L R Subject 4 Response f r e q . L e f t Right target target 24 ' 0 24 2 22 24 3 ROC Le f t target 1.0 1.0 Right target ,»j .083 1.0 < C„ non-unitary context RVF L R 21 3 24 16 24 .875 1.0 .333 1.0 C unita r y context LVF L R 17 7 24 3 21 24 .708 1.0 .125. 1.0 C„ non-unitary context LVF 19 5 24 2 22 24 .792 1.0 .083 1.0 •' 4.5920 .1.3023 1.4366 1.9964 Covariate R/L eve 1.0 MM unitary context RVF Binary choice L R non-unitary context RVF L R Subject 5 Response freq. Left target 5 19 24 9 15 24 Right target 6 18 24 4 20 24 ROC Left target .208 1.0 .375 1.0 Right target .250 1.0 3 .167 1.0 * C. l .0972 .4721 Covariate R/L eye .9545 unitary context LVF L R 10 14 24 7 17 24 .417 1.0 .292 -"\ 1.0 • .2581 ••>. A .« non-unitary context L LVF R 13 11 24 7 17 24 .542 1.0 .292 .1.0 .5132 C. unitary context RVF Subject 6 Binary choice L R Response freq. Left Right target target C„ non-unitary context L RVF ' R C. unitary context L LVF R C- non-unitary context L LVF R 15 9 24 15 9 24 15 9 24 15 9 24 15 9 24 11 13 24 14 10 24 7 17 24 ROC Lef-t-target .625 1.0 .625 1.0 .625 1.0 .625 1.0 Right target .625 1.0 I .458 1.0. .583 1.0 .292 1.0 - .• ... . . ^ p C. 1 .3321 .0835 .6949 Covariate R/L eye .9583 u n i t a r y context RVF u n i t a r y context LVF non-unitary context LVF Subject 7 Binary choice L R non-unitary context L RVF R L R L R Response freq L e f t target 21 3 24 11 13 24 23 1 24 19 5 24 Right target 6 18 24 6 18 24 16 24 11 15 24 ROC L e f t target .875 1.0 .458 1.0 .958 1.0 .792 1.0 Right target .250 1.0 s ,250 1.0 .i .333 1 1.0 .458 .< 1.0 1.5883 .4369 1.6897 •t > .» ,7988 Covariate R/L eye 1.1538 CO unitary context RVF Binary choice L R Subject 8  Response freq . Left target 21 3 24 Right target .12 12 24 ROC Left target .875 1.0 Right target .50 1.0 C. l .8660 Covariate R/L eve 1.0435 non-unitary context L RVF R 9 15 24 17 M7 24 .375 1.0 .708 1.0 -.6949 unitary context LVF L R non-unitary context L LVF R 15 9 24 15 9 24 13 11 24 14 10 24 .625 1.0 .625 1.6 .542 l .o • .583 ' 1.0 , .1660 A •• *t,w-° .0835 Subject 9 X C. unitary context RVF Binary choice L R Response freq. Left target 15 9 24 Right target 5 19 24 ROC Left .625 1.0 Right target target .208 .i-l 1.0 C„ non-unitary context L RVF R 18 6 24 5 19 24 .750 1.0 .208 1.0 . .t C. unitary context LVF L R 15 9 24 14 10 24 .625 1.0 .583 1.0 < i C_ non-unitary context L LVF R 16 8 24 12 12 24 .667 1.0 .50 1.0 Covariate R/L eye .7256 .9583 1.2633 .0835 .3364 unitary context RVF C0 non-unitary context RVF C. unitary context LVF Subject 10 Response freq, Binary choice L R L R L R C„ non-unitary context L LVF R Left target 11 13 24 12 12 24 13 11 24 15 9 24 Right target 17 7 24=» 11 13 24 13 11 24 10 14 24 ROC Left target .458 1.0 .50 1.0 .542 1.0 .625 1.0 Right target .708 -H 1.0 .458 1.0 .542 ' 1.0 •+ .417 1.0 •* ••May* * •+ .* .t -.5132 .0817 .4174 Covariate R/L eye .9565 OS Table tfj Summary of analysis of covariance. Experiment 3. sv df ss MS F p Total 39 1 .896073 X 10exp+2 Visual f i e l d 1 1 .320890 X 10exp-l 1 .320890 X 10exp-l .2052 .6633 Context 1 8 .598383 X lOexp-1 8 .598383 X 10exp-l 1.1070 .3215 Vis u a l f i e l d X context 1 8 .129547 X lOexp 0 8 .129547 X lOexp 0 3.7.143 .0836 Subjects 9 1 .353311 X 10exp+2 1 .503678 X 10exp+l - -V i s u a l f i e l d X subjects 9 5 .792831 X lOexp 0 6 .436378 X 10exp-l - -Context X subjects 9 6 .990338 X lOexp 0 7 .767042 X 10exp-l - -V i s u a l f i e l d X context X subjects 9 1 .969821 X 10exp+l 2 .188689 X lOexp 0 - -Adjusted group means RVF LVF 2.031 0.837 1.245 1.853 58 The coefficients are: RVF LVF Gp. 1 Gp. 2 Gp. 3 Gp. 4 ( V < EC2 ) <EC2> W- 1 -1 0 0 W O 0 1 -11 for the contrasts: ^ 2 = y 3 - y 4 We set EG = EW/k = a/k = .05/2 = .025 If the estimate of the variance of the sampling d i s t r i b u t i o n of i s given by •> §& = MS _w2/N. then the probability i s (1 - a) = .95 that the values of both contrasts simultaneously l i e within the confidence intervalss of the form ii a/k r a/k To test HQ : y 1 - p_ = 0, noting that s^ = /(.777 x .20) = .394, 1.194 - .394 x /5.48 < ifj. $ _C_94 + 7394 x /5T48 = .272 «: i|> < 2. 116 Since the confidence i n t e r v a l does not include zero, the hypothesis i s rejected at EW = .05. To test H. : p. - y^5= 0, using s^ = .394, .608 - .394 x /5.48 < \p £ .608 + .394 x /5.48 = -.314 s< ^ „ £.1.530 59 Since the confidence i n t e r v a l includes zero, the hypothesis cannot be rejected at EW = .05. The results of these contrasts show a s i g n i f i c a n t context effect i n the RVF, but not i n the LVF. The hypotheses of a larger d i f f e r e n t i a l i n the LVF under the two contextual conditions, and of an ove r a l l LVF super-i o r i t y i n both conditions are shown to be untenable. Discussion The results of the three experiments w i l l be b r i e f l y recapitulated. Experiment 1 showed a marked effect of context on feature detection, such that embedding a target l i n e segment i n a whole, unitary figure f a c i l i -tated the target's d i s c r i m i n a b i l i t y . Experiment 2 demonstrated that f i g u r a l information leading to a three-dimensional interpretation of the context pattern i s a necessary condition of th i s context effect. In the th i r d experiment the s t i m u l i were presented i n each of the l e f t and right v i s u a l f i e l d s , the result being that the context effect favoring the three-dimensional unitary pattern held for the RVF but not for the LVF. In attempting to explain these findings i t i s necessary b r i e f l y to consider the question of explanation per se, and to derive ( i f only i n broad outline) some set of c r i t e r i a for an explanation of perceptual phenomena (see Appendix A). We may take the question of explaining a perceptual phenomenon as bifurcating into two a n a l y t i c a l problems: Explanaridum^(expl^)—the transduction (or organization, or process-ing) of sensory input (or information) i n the nervous system of the percipient. 60 Explanandum^(expl^)—the a r i s i n g of a conscious state putatively associated with expl^. While t h i s may turn out to be based on l i m i t i n g or false assumptions, i t seems clear both that events i n the nervous system have perceptual consequences, and that these consequences—while not necessarily behav-i o r a l — a r e usually conscious for the percipient. Under th i s analysis, we cannot claim to have explained a perceptual phenomenon u n t i l e x p ^ has been accounted for. In this discussion several current approaches to expl^ w i l l be examined, with p a r t i c u l a r reference to the results reported here; i n Appendix Bla tentative outline to the solution of e-xpl^ w i l l be proposed. For many phenomena, i t i s clear that under our present understanding of explanation, there may be more than one explanatory account (perhaps at different levels of a n a l y s i s — c f . Fodor 34). Hemispheric l a t e r a l i t y , for example, may be said to explain a perceptual phenomenon, i f the l a t t e r can be shown to be a pa r t i c u l a r consequence of some substantive general theory of d i f f e r e n t i a l functioning i n the two cerebral hemispheres. As previously indicated, a number of authors (e.g., Sperry 117, Levy 84, Bakan 4) interpret experimental findings as indicative of just such a duality. To f i t the results of Experiment 3 into such a framework i s not easy, at f i r s t sight. The hypothesis was disproved that an ove r a l l minor hemi-sphere superiority i n s p a t i a l perception would produce a v i s u a l f i e l d effect favoring the LVF. Instead, no difference was found between l e f t and right v i s u a l f i e l d s , and the context effect hald only for the RVF. Pointing to a number of possible sources of confounding—which Experiment 61 3 was designed to control for—White (129) argues that there are no r e l i a b l e v i s u a l f i e l d differences i n the recognition of non-verbal stim-u l i . The serious d i f f i c u l t y here i s determining that a stimulus i s non-verbal; while response verbalization was controlled for i n Experiment 3, i t was not possible to rule out verbal imagery or sub-vocalization of a response ( " l e f t , " ''right") which would normally have a verbal l a b e l . Indeed, some subjects, when asked, reported a tendency to do t h i s . This possible confounding makes i t d i f f i c u l t to explain the results of Experiment 3 under a hemispheric l a t e r a l i t y paradigm. Levy (84) has suggested, however, that while the minor hemisphere abstracts integrated, h o l i s t i c stimulus information, the major hemisphere analyzes stimulus elements, for description i n language. Given, p a r t i c u l a r l y , the physio-l o g i c a l evidence that neurons with receptive f i e l d organizations send axons across the corpus callosum (see Sekuler 112), form perception probably concerns both hemispheres j o i n t l y . Under th i s view, an i n i t i a l stimulus presentation to the minor hemisphere would produce a shunting to the major hemisphere of information concerning stimulus elements, for verbal l a b e l l i n g , while presentation to the major hemisphere would involve transmitting across to the minor information from which s p a t i a l properties could be abstracted. The finding of Experiment 3 would be interpreted as indicating that under the condition of i n i t i a l presentation to the major hemisphere, the unitary f i g u r a l information from the minor hemisphere helps i n the l i n e detection task—as compared with the non-unitary i n f o r -mation—when this information i s combined with the major hemisphere's processing of the target l i n e ; while under the other c o n d i t i o n — i n i t i a l presentation to the minor hemisphere—it does not help. While some such 62 interpretation may be v a l i d , our knowledge of l a t e r a l i t y of function i n hemispheres seems s u f f i c i e n t to provide only a potential explanation for e x p l 1 . The work of Gibson (39, 40) has given r i s e to the mathematical modelling approach of Johansson (65, 66). Rather than inquiring into the processes of perceptual phenomena, Johansson aims at a purely math-ematical analogy i n terms of the geometry of projective r e l a t i o n s . Employing a geometry dealing with relations which remain invariant under perspective transformation, Johansson i s able to model the information transfer from stimulus array to percept, at least i n certain cases (which generally involve perception of motion—see Johansson 66). While such mathematical modelling may be said to provide an explanation of phenomena, this approach seems unable, at present, to account for the results of Experiments 1 and 2; i t i s not clear how the context effect could be predicted using Johansson's model, though one notes that the rule of object constancy i s important to his whole approach. A different mathematical analogy i s advanced by G r i f f i t h (44). Like the model of McCullough and P i t t s (91) and those discussed by R a t l i f f (107), this i s a theory of neural functional organization based on the assumption of certain mathematical properties of neural aggre-gates. Central to the theory i s the concept of a "mode"—a set of neurons in t e r - l i n k e d i n an extensive and excitatory manner. The aim i s to mathematically model psychological processes i n terms of interactive feedback between modes—postulated as sub-cortical masses of neurons— and the cortex, seen as functionally organized i n v e r t i c a l chains of neurons (cf. Hubel and Wiesel 60). G r i f f i t h ' s discussion of v i s u a l 63 perception i s b r i e f , but implies the existence of feature detectors, such that the excitation of c e l l s i n the v i s u a l cortex corresponding to a certain feature leads to the excitation of a mode related to the feature. Like Johansson's, G r i f f i t h ' s model i s not s u f f i c i e n t l y developed to help very much i n explaining the results of Experiments 1 and 2. The concept of the mode as derived, however, seems more rigorous than Hebb's notion of the c e l l assembly (48), and a potential account of expl^ i n the case of the contextual effect would perhaps come from the mathematical speci-f i c a t i o n of a mode, or modes, corresponding to s p a t i a l information, and excitatory/inhibitory interaction with l i n e orientation feature detectors. While G r i f f i t h does not discuss l a t e r a l i t y of hemisphere function, the model evidently has the scope for corresponding modes on the two sides of the b r a i n — t h a t i s , modes containing the same category of c e l l s (see G r i f f i t h 44, p. 9 ) — t o l i n k together, and act as a single, larger.mode (44, pp. 30-31). Two quite different attempts to derive a functional mathematical analysis of psychological events have been b r i e f l y overviewed. To account for the context e f f e c t , Johansson's model i s found inadequate, while G r i f f i t h ' s provides, at best, a potential explanation. Functional anal-ysis by mathematical analogy, however, may be expected to y i e l d increas-ingly precise solutions to the problem of expl^ (and, possibly, of expl.). To provide an explanation of the contextual e f f e c t , i t seems neces-sary to refer to a system of feature detection. On rigorous grounds, Sekuler (112) has shown that a single Hubel and Wiesel type c e l l cannot by i t s e l f uniquely signal any property to which i t responds d i f f e r e n t i a l l y , 64 such as l i n e orientation. The mechanistic assumption—evidently favored by some (e.g., Weisstein 126; see also Sekuler 112)—that i t i s single c e l l s which have the property of s i g n a l l i n g complex information i s not only probably f a l s e , but unnecessary. While the evidence indicates that such c e l l s play a role i n the analysis of such stimulus aspects as orientation (Sekuler 112) ,itcisasf part.of a h i e r a r c h i c a l system of feature detection. The context effect shows that meaning or structure has con-sequences for this system, as suggested by Weisstein and Harris (127) . Expl^ requires an account of a c t i v i t y i n the nervous system associated with such an ef f e c t , an account provided, at least i n outline, by Milner (96). Whether Milner's suggestions about f i g u r a l unity (to be described) represent any advance on the Gestalt notions i s a debatable point (which w i l l not be debated here). While some such theory could provide the micro-analysis to complement the functional analysis of the Gestalt theorists, i t may w e l l turn out that h o l i s t i c phenomena are to be under-stood best i n terms of t h e i r organizational laws—where these are deriv-a ble—as properties of functional wholes. (Si m i l a r l y , the understanding of how neural aggregates behave does not give any sign of being advanced by the discussion of quarks—with or without the property of "colored charm.") Milner suggests that the perception of f i g u r a l unity depends on the contiguity of active c e l l s i n organized parts of the projection pathway. A process of synchronizing interaction i s postulated, such that adjacent c e l l s i n t e n s i f y each other's a c t i v i t y and have the effect of j o i n i n g the elements of a figure together perceptually. Synchronous volleys of 65 impulses re s u l t at subsequent l e v e l s , which f a c i l i t a t e the channelling through the feature detection system of the signals corresponding to f i g u r a l units (Milner 96, p. 526). Milner's use of the notion of a c e l l assembly bears quite a resemblance to G r i f f i t h ' s (44) concept of a mode. Basing his analysis partly on Hebb's (48) notion of central f a c i l i t a t i o n of a c e l l assembly, Milner presents an account of how attentional feed-back could operate to influence input. A signal "corresponding" to a (centrally f a c i l i t a t e d ) c e l l assembly f i r e s the assembly, which feeds back to the input l e v e l v i a a postulated divergent pathway, running p a r a l l e l to the ascending convergent pathway. This feedback i s ascribed the property of enhancing the a c t i v i t y of c e l l s , causing them to dominate the feature detection networks (96, p. 532 f f . ) . Under this theory, some central representation of the three-dimen-sional unitary stimulus (Experiment 1) would provide f a c i l i t a t o r y feedback to enhance the detection of the oriented l i n e segment i n the detection system. The contextual effect i s plausibly retro-dictable i n that, for the unitary figure (as opposed to the non-unitary), synchronous impulses would f a c i l i t a t e the passage of the signals through the feature detection system. This should also occur, however, i n the case of the two-dimen-sional unitary figure of Experiment 2. What the percipient seems to be doing i n the f i r s t experiment (C^) i s postulating the target l i n e as part of a three-dimensional object. To r e t a i n object constancy (cf. Gibson 39, Johansson 66) the l i n e i s seen i n perspective, and perhaps weighted i n some way during processing such that i t i s perceived as being longer than the corresponding l i n e s i n the non-unitary condition, f a c i l i t a t i n g i t s detection as a target. The depth 66 effect on this view, has the consequence of shortening the l i n e for the perceiving organism, who concludes that i t i s longer than i t appears, and compensates for the foreshortening on the basis of past experience with three-dimensional objects. An alternative account along the same li n e s i s that i t i s the manipulated depth cue alone which i s the e f f i -cient cause of the difference i n detection performance between Experiments 1 and 2, the unity or "goodness" of the figures being incidental to i t . This could be tested, i f one could devise a non-unitary figure containing a depth cue. Postulations and use of cues are not readily ascr.ibable to c e l l assemblies, which are unlikely phenomena anyway (see, for example, Neisser 100)—apart from the usual underlying mechanistic assumption that psychological events, insofar as they are r e a l , are to be understood as being a mechanical analogy, such that the operations of the whole are predictable from an analysis of parts which are ( i n princ i p l e ) observ-able . Given the current state of psychology, the foregoing probably repre-sents the best approximation to explaining explanandumi for the experimen-i l t a l results described e a r l i e r (and taking the set of i n i t i a l conditions C^ as being those described i n the experimental r e p o r t — t h e C^ actually represent a whole problem i n themselves, especially for a p r o b a b i l i s t i c explanation [see Kim 71]). I t s somewhat unsatisfactory nature i s due perhaps to reasons outlined i n Appendix A — v i z . the absence of c r i t e r i a for determining what constitutes an explanation of a perceptual phenomenon, *I am indebted to Stan Coren and Larry Ward for t h i s suggestion. and a lack of a theory of consciousness which could be applied to expl . References Alpern M.t Eye movements. In Jameson D. and Hurvich L. M. (eds.): Handbook of sensory physiology - v o l . 4. Springer Verlag (1974). Attneave F.: Psychological Review, 61 (1954). Bakan P.: Nature, 223 (1969). Bakan P.: The eyes have i t . In Psychology Today, A p r i l (1971). Bakan P. and Streyer F.: Perceptual and Motor S k i l l s , 36 (1973). Barnett L.: The universe and Dr. Einstein. Bantam (1968). Beck J . : Surface color perception. Cornell University Press (1972). Bohm D.: Causality and chance i n modern physics. Harper (1961). Braithwaite R. B.: Explanation. In Preece W. E. (ed): Encyclopedia  Britannica - v o l . 8. Benton (1973) . Broad C. D.: The mind and i t s place i n nature. Routledge & Kegan Paul Ltd. (1925). Brown D. and Owen D.: Psychological B u l l e t i n , 68, 4 (1967). • Burgers J. M.: Science, 189 (1975). Burt C: B r i t i s h Journal of Psychology, 59, 1, (1968). Burt C : B u l l e t i n of the B r i t i s h Psychological Society, 22, (1969). Chomsky N.: A review of B. F. Skinner's 'Verbal behavior'. In Language, 35, 1 (1959). Chomsky N.: Cognition, 1, 4 (1972). Corcoran D. W. J.: Pattern recognition. Penguin (1971). Cohen G.:Journal of Experimental Psychology, 97, 3 (1973). De Broglie L.: New perspectives i n physics. Basic Books (1972). Dimond S. and Beaumont J . : Nature, 232 (1971). Dimond S. and Beaumont J . : Acta Psychologica, 36 (1972). Dimond S. and Beaumont J . : B r i t i s h Journal of Psychology, 63, 4 (1972). Dimond S. and Beaumont J . : Cortex, 9 (1973). Dixon N. F.: Subliminal perception: the nature of a controversy. McGraw (1971) . 69 25. Durnford M. and Kimura D.: Nature, 231 (1971). 26. Eccles J. C: The neurophysiological basis of mind. Clarendon Press (1960). 27. Eccles J. C: Brain and conscious experience. Springer (1966). 28. Eccles J. C: The brain and the unity of conscious experience. Cambridge University Press (1965). 29. Egan J. P. and Clarke ?. R.: Psychophysics and signal detection. In Sidowski J. B. (ed.): Experimental methods and instrumentation  i n psychology. McGraw-Hill (1966). 30. Einstein A. and Infeld L.: The evolution of physics. Clarion Books (1938). 31. Ferguson G. A.: S t a t i s t i c a l analysis i n psychology and education. McGraw-Hill (1971). 32. Festinger L., Coren C. and Rivers G.: American Journal of psychology, 83, 2 (1970). 33. F i r s o f f V. A.: L i f e , mind and galaxies. Oliver & Boyd (1967). 34. Fodor J . A.: Psychological explanation. Random House (1968). 35. Gazzaniga M.: The bisected brain. Appleton-Century-Crofts (1970). 36. Gazzaniga M.: American S c i e n t i s t , 60, 3 (1972). 37. Geschwind N. and Levitsky W.: Science, 161, 186 (1968). 38. Geschwind N.: Science, 170 (1970). 39. Gibson J. J.: The senses considered as perceptual systems. Houghton M i f f l i n (1966). 40. Gibson J . J . : Psychological Review, 64, 5 (1957). 41. Glezer V. D., Leushina L. I., Nevskaya A. A. and Prazdnikova N. V.: Vision Research, 14, 7 (1974). 42. Green D. M. and ^wets J. A.: Signal detection theory and psycho-physics . Wiley (1966). 43. Grier J . B.: Psychological B u l l e t i n , 5, 6 (1971)." 44. G r i f f i t h J. S.: A view of the brain. Oxford University Press (1967). 45. Haber R. N. and Hershenson M.: The psychology of v i s u a l perception. Holt, Rinehart and Winston (1973). . 70 46. H a l l C. S. and Lindzey G. (eds.): Theories of personality. Wiley (1957). 47. Hayashi T. and Bryden M. P.: Perceptual and Motor S k i l l s , 25 (1967). 48. Hebb D. 0.: The organization of behavior. Wiley (1949). 49. Helmholtz H. L. F.: The facts of perception. In Warren R. M. (ed.): Helmholtz on perception. Wiley (1968). 50. Helmholtz H. L. F.: Treatise on physiological optics. Dover (1962). 51. Hering E.: Outlines of a theory of the l i g h t sense. Harvard University Press (1964). 52. H i l l i a r d R.: Cortex, 9 (1973). 53. Hochberg J. and McAlister E.: Journal of Experimental Psychology, 46, 5 (1953). 54. Hochberg J . and Beck J . : Journal of Experimental Psychology, 47 (1954). 55. Hochberg J. and Brooks V.: American Journal of Psychology, 73, 3 (1960). 56. Hodos W. A.: Psychological B u l l e t i n , 74 (1970). 57. Hospers J.: Journal of Philosophy, 43, 13 (1946). 58. Hospers J.: An introduction to philosophical analysis. Prentice-H a l l (1967). 59. Hubel D. H. and Wiesel T. N. : Journal of Physiology, 160 (1962). 60. Huhel D. H. and Wiesel T. N.: Journal of Neurophysiology, 28 (1965). 61. Hubel D. H. and Wiesel T. N.: Journal of Physiology, 195 (1968). 62. Hubel D. H. and Wiesel T. N.: Journal of Neurophysiology, 28 (1967). 63. Hyde J . B., Abesson E. J . and Berinstein E: Experientia, 29, 1131 (1973) . 64. James W.: The princ i p l e s of psychology. Dover (1950). 65. Johansson G.: Perception and Psychophysics, 14 (1973). 66. Johansson G.: S c i e n t i f i c American, June (1975). 67. John E. R.: Mechanisms of memory. Adademic Press (1970). 68. Kaplan D.; Philosophy of Science, 28 (1961). 69. Katz D.: Gestalt Psychology. Ronald Press (1950). 70. Kershner J. and Gwan-rong Jeng A.: Neuropsychologia, 10 (1972). 71. Kim J . : Explanation i n science. In Edwards P. (ed.): Encyclopedia of Philosophy, v o l 3. Collier-MacMillan (1947). 72; Kimura D.: Archives of Neurology, 8, 264 (1963). 73. Kimura D.: Cortex, 3 (1967). 74. Kimura D.: S c i e n t i f i c American, 228, 3 (1963). 75. Kimura D.: Neuropsychologia, 4 (1966). 76. Knox C. and Kimura D.: Neuropsychologia, 8 (1970). 77. Koestler A.: The act of creation. Pan (1964). 78. Koestler A.: The ghost i n the machine. Pan (1967). 79. Koffka K.: P r i n c i p l e s of Gestalt psychology. Routledge & Kegan Paul Ltd. (1935) . i 80. Koffka K.: Gestalt psychology. In Johnson A. (ed.): Encyclopedia of the Social Sciences - v o l . 6. MacMillan (1963). 81. Kohler W.: Gestalt psychology. Mentor (1947). 82. Kvasnica J . : Physical f i e l d s . I l i f f e Books (1966). 83. LeMay M. and Culebras A.: New England Journal of Medicine, 287, 168 (1972). 84. Levy J.: Nature, 224 (1969). 85. Levy J . , Trevarthan C , and Sperry R. W. : Brain, 95 (1972). 86. Lewin K.: A dynamic theory of personality. McGraw (1935). 87. Lindsay P. H. and Norman D. A.: Human information processing. Academic Press (1972) . 88. Mach E.: The analysis of sensations. Dover (1959). 89. Marg E. and Adams J. E.: Experientia, 26 (1970). 90. Maxwell J. C: A dynamical theory of the electro-magnetic f i e l d . In Niven W. D. (ed.): The s c i e n t i f i c papers of James Clerk  Maxwell. Dover (1965). 71 McCullough W. S. and P i t t s W. : B u l l e t i n of Mathematical Bio- physics, 5, 115 (1943). Michels K. M. and Zusne L.: Psychological B u l l e t i n , 63, 74 (1965). Miles T. R.: Gestalt theory. In Edwards P. (ed.): Encyclopedia  of Philosophy - v o l . 3. Macmillan (1967) . M i l l e r G. A.: Psychological Review, 63(1956). Milner B.: B r i t i s h Medical B u l l e t i n , 27, 3 (1971). Milner P. M.: Psychological Review, 81, 6 (1974). Myers J . L.: Fundamentals of experimental design. A l l y n & Bacon (1972). Nachmias J.: Signal detection theory and i t s application to problems i n v i s i o n . In Jameson D. and Hurvich L. M. (eds.): Handbook of sensory physiology - v o l . 4. Springer-Verlag (1974) Natsoulas T.: Psychological B u l l e t i n , 81, 10 (1974). Neisser U.: Cognitive psychology. Meridith (1967). Neisser U. : Changing conceptions of imagery. In Sheehan P. W. (ed.) The function and nature of imagery. Academic Press (1972) . Pap A.: An introduction to the philosophy of science. The Free Press (1962). Pastore R. E. and Scheirer C. J .: Psychological B u l l e t i n , 81 12 (1974). Paterson A. and Zangwall 0. L.: Brain, 67, 331 (1944). Popper K. R.: The l o g i c of s c i e n t i f i c discovery. Harper (1959) . Rabbitt P. M. A.: B r i t i s h Medical B u l l e t i n , 27, 3 (1971). R a t l i f f F.: Mach bands: quantitative studies on neural networks  i n the r e t i n a . Holden-Day (1965). R i z z o l a t t i G., Umilta C. A. and Berlucchi G.: Brain 94 (1971). Rothman M. A.: The laws of physics. Fawcett (1963) . Ryle G.: The concept of mind. Hutchinson (1949). Sakitt B.: Nature, 241 (1973). 112. Sekuler R.: Spatial v i s i o n . In Annual Review of Psychology (1974). 113. Selfridge 0. G.: Pandemonium: a paradigm for learning. In: fhe mechanization of thought processes. H. M. Stationery Office, London (1959) . 114. Simpson A. J. and F i t t e r M. J . : Psychological B u l l e t i n , 80, 6 (1973) 115. Skinner B. F.: Science and human behavior. New York (1953). 116. Skinner B. F.: About behaviorism. Knopf (1974). 117. Sperry R. W.: L a t e r a l i z a t i o n of cerebral function i n the s u r g i c a l l y separated hemispheres. In McGuigan F. J. and Schoonover R. A. (eds.): The psychophysiology of thinking. Academic Press (1973). 118. Sperry R. W.: American Psychologist, 23, 10 (1968). 119. Sperry R. W., Gazzaniga M. and Bogen J . : Interhemispheric r e l a t i o n -ships: The neocortical commissures; syndromes of hemisphere deconnection. In Vinken P. J. and Bruyn G. W. (eds.): Handbook of c l i n i c a l neurology - v o l . 4. North Holland (1969). 120. Sperry R. W.: Psychological Review, 76 (1969). 121. Sperry R, W.: Psychological Review, 77 (1970). 122. Theobold D. W. : The cono.ept of energy. Spon (1966). 123. Tolman E. C: Purposive behavior ; i n animals and man. University of C a l i f o r n i a Press (1932). " 124. Ward L. M. and Wexler D. A.: 'Levels of feature analysis i n processing v i s u a l patterns. (Unpublished manuscript). 125. Weinstein G. W., Hobson R. R. and Baker F. H.: Science 171 (1971). 126. Weisstein N,:\ Psychological B u l l e t i n , 72 (1969). 127. Weisstein N. and Harris C. S.: Science, 186 (1974). 128: Weyl H.: Philosophy of mathematics and natural science. Atheneum (1963_~ 129. White M. J . : Psychological B u l l e t i n , 72, 6 (1969). 130. White M. J . : B r i t i s h Journal of Psychology, 63, 4 (1972). 131. Whitehead A. N.: Adventures of ideas. Cambridge University Press v i (1933). ~ 74 132. Witelson S.: Brain, 96 (1973). 133. Zusne L.: Journal of Experimental Psychology, 69, 3 (1965). 134. Zusne L.: V i s u a l perception of form. Academic Press (1961). 75 Appendix A: Explanation A p a r t i a l examination of the l i t e r a t u r e on explanation leads one to the conclusion that the very important question of the l o g i c a l r e l a t i o n -ship between explanans (that which i s explaining), and explanandum (that which i s to be explained) i s not yet s a t i s f a c t o r i l y solved (see Braith-waite 9; Hospers 57, 58; Kaplan 68; Kim 71). This means that, hitherto and perhaps i n d e f i n i t e l y , there exists no set of c r i t e r i a with reference to which a claimed explanation of anything may be judged, or ruled out. Even the derivation of Kaplan i n formal logic (68)—perhaps the most rigorous analysis extant of explanation—can be shown to permit "explan-ations" which are c l e a r l y inadequate (see Kim 71); that i s , results which f a i l to supply an adequate answer to the question "Why i s the event what i t i s ? " Despite this outstanding problem, however, under the covering-law theory s c i e n t i f i c explanations are analyzable i n terms of two models— the deductive and the p r o b a b i l i s t i c . In the deductive model, the explanans comprises two sets of state-ments: (a) a set specifying i n i t i a l conditions, (b) a set of lawful or theoretical statements, T.. I The explanandumiis deducible—or, equivalently, pre- or retro-d i c table—from the explanans. Under the p r o b a b i l i s t i c model, a high probability of event A given the occurrence of another event B can be said to explain why event A was observed, given the information that event B occurred. A p a r t i c u l a r kind of explanation (or analogy which, despite the l o g i c a l d i f f i c u l t i e s , i s generally held to be explanatory) i s provided 76 by models. Based on an " a s - i f " h e u r i s t i c (Braithwaite 9), these include the important class of mathematical models of phenomena. Any purported explanation of a perceptual phenomenon, on this view, should permit a deductive or inductive inference from an explanans to an explanandum—where the explanans, i t should be added, includes a theory or "law" which meets the c r i t e r i o n of f a l s i f i a b i l i t y (Popper 105); or, i t should provide a powerful, i f i n t u i t i v e , mathematical analogy i n terms of which the phenomenon may be understood. The form of the explanandum—the problem to be solved—may also be specified. While phenomena labelled "perceptual" may occur below the threshold of consciousness (see Dixon 24), most have an experiential aspect. If we also assume that nervous a c t i v i t y i s associated with per-ceptual events, we may conclude that an adequate explanation of a perceptual phenomenon should provide an account of both of the following: Explanandumm~—the transduction of (assumed externally i n i t i a t e d ) stimulus energy i n the nervous system of the percipient. Explanandum.—the existence of a state of awareness apparently associated with explanandum^. While explanandum^ may be soluble with present techniques, explan-andum. i s s u f f i c i e n t l y d i f f i c u l t that, with certain exceptions, the ques-tion has generally been avoided by defining i t as outside the purview of science (Tolman 123), or c a l l i n g i t a "pseudo-problem" or "category error" (e.g., Ryle 110). 77 Appendix B: Toward a Theory of Consciousness The problem of accounting s c i e n t i f i c a l l y for the existence of sub-je c t i v e states of awareness i n what i s generally regarded as a material cosmos—though not universally (for example, Whitehead [131], Burgers [12], F i r s o f f [33])r -remains unsolved. Psychologists have f a i l e d , on the whole, to deal with the issue, evidently because i t did not seem pertinent to the solution of certain kinds of problems (Skinner 116; but see Chomsky 15, 16), because i t was held to be outside the domain of science (Tolman 123), or on the substantive grounds that i t i s d i f f i c u l t . I have argued that an explanation of the kinds of perceptual phenomena with which the experiments reported here have been concerned implies an explanation of their experiential aspect. Sperry (120, 121) has sug-gested that the present s c i e n t i f i c model i s incomplete without an addition that brings q u a l i t a t i v e mental phenomena into i t , and Natsoulas (99) points out that we should either take Sperry's problem seriously, or ". . . defend the claim that as we know them there are no hues and pains, for example" (99, p. 621). While the problem w i l l not be solved here with a full-blown theory, i t i s relevant to outline some suggestions which may f a c i l i t a t e a solution. A discussion of the standard philosophical treatments of the mind-body problem w i l l not be undertaken—they seem to have their standard l o g i c a l refutations (e.g., Hospers 58), except perhaps i n the case of interactionism, the usual "refutation" of which i s based on a misconstru-ing of the law of conservation of energy. On the basis of the evidence for the existence of a feature detec-tion system, and some other considerations, i t may be said that the human 78 nervous system shows a fundamentally hierarchic organization (see Milner 96, Sperry 121, Burt 13, and, for a broad view, Koestler 78). Perceptual a c t i v i t y t y p i c a l l y has a conscious component—indeed, this i s how the word "perceptual" has any meaning as a referent. Our present understand-ing, i n the case of perceiving a form, for example, i s that stimulus energy i s transmitted as information through a hierarchic system of feature detection, to which i s imputed such properties as information processing and extraction. The question of how the f i g u r a l percept i s organized i s more problematical (compare, for example, Koffka 79 and Milner 96), but the process, or series of processes, produces for the percipient a conscious experience of a figure, whatever consciousness may be, i t c l e a r l y has a passive aspect as a r e c e i v e r — o r , we may say, detector—of information. In addition, i t i s argued, i t has an active aspect, self-evidently being associated with the i n i t i a t i o n of voluntary a c t i v i t y , whatever may be the cause of our subjective states of w i l l i n g , or deciding, there can be l i t t l e question that we act upon them, and that they are distinguish-able from the behavior to which they give r i s e . Sperry (120, 121) i s one investigator (among others; see, for example, Eccles 26, 27, 28; James 64; Burt 13, 14) who finds i t necessary to give consciousness a causal role i n brain processes. Sperry's view i s that consciousness i s an emergent property of neural a c t i v i t y , and t r i e s to combine this idea with the requirement that consciousness also be a determinant of that a c t i v i t y (120, 121). The d i f f i c u l t y here i s that emergent phenomena are far from being "generally accepted" i n science as having "causal potency," as Sperry would have us believe (121, p. 586); rather, the theory of 79 emergentism seems to have a quite problematical status (see, for example, the discussion by Pap 102). From Burt's analysis of some purported examples of emergent phenomena (14), one i s led to the conclusion that the la b e l "emergent" when applied to anything represents nothing more than an admission of ignorance of i t s cause. Be that as i t may, there are reasonable grounds for ascribing to consciousness two fundamental aspects—detector and effector. We are dealing with an unobservable which has observable consequences of two k i n d s — t h e p u b l i c l y observable behavior of an organism which i s conscious, and the existence of private subjective states. Like the term "perceptual,'' "observable" implies consciousness, as does "empirical" and i t s deriva-tives . I t i s i n s t r u c t i v e , at t h i s point, to examine some relevant notions of gross neural a c t i v i t y . William James has pointed out (64) that one characteristic of the nervous systems of organisms whose consciousness seems most highly developed i s i n s t a b i l i t y . A nervous system constructed to have i n f a l l i b l e reactions would be able to react to only a small sub-set of environmental changes, and be unadaptive; conversely, a system poten t i a l l y adapted to respond to an i n f i n i t e variety of s t i m u l i would be i n a continuing state of unstable equilibrium, " . . . i t s f a l l i b i l i t y as great as i t s elaboration" (64, p. 138 f f . ) . James goes on to argue that the performance of a highly developed brain i s l i k e " . . . dice thrown forever on a table," and postulates that consciousness i s the agent which "loads the dice," acting to control what would otherwise be the indeterminateness of brain a c t i v i t y . Arguing on neurophysiological grounds, Eccles (26, see also 27, 28) 80 comes to similar conclusions. Eccles proposes that during any process i n which, v o l i t i o n i s involved, a whole f i e l d of neurons i n the active network of the cortex i s i n a state of poised a c t i v i t y . Suggesting that the " c r i t i c a l l y poised neurons" would be the detectors and amplifiers of an "action of w i l l , " Eccles presents a model showing that an influence exerted at only one node of an active c o r t i c a l network could be s u f f i -cient to modify the discharge pattern of hundreds of thousands of neurons (26, pp. 261-286). We abstract from the foregoing the h e u r i s t i c notion of the brain as a f i n e l y balanced s t a t i s t i c a l machine—as Eccles concludes, with an i r o n i c reference to Ryle (110), just the sort of machine a "ghost" could operate. We also note, after William James, that the postulated effector action of consciousness i s what might be expected i n a component " . . . added for the sake of steering a nervous system grown too complex to regulate i t s e l f " (64, p. 144). Adherents to an evolutionary view may derive from this the proposition that the development of consciousness, p a r t i c u l a r l y i n i t s active mode, has adaptive consequences. I f the postulation of consciousness as both detector and effector i s accepted, what remains unsolved i s the c r i t i c a l question of how conscious-ness and nervous system act upon each other. The only t r a d i t i o n a l view under which the foregoing could be subsumed, i n p r i n c i p l e , i s dualism, though whether the suggestion to be outlines here w i l l be regarded as d u a l i s t i c may be no more than a terminological issue. For the sake of c l a r i t y , however, we note that the two usual objections to dualism are: (a) that i t provides no account of how the mental—as non-physical—and the physical could interact; (b) that such an account cannot be provided 81 anyway, because any such interaction would v i o l a t e the law of conserva-t i o n of energy. The f i r s t proposition says, undoubtedly, that there i s a problem for the dualist to solve. The second i s not acceptable. A law i s an empirical generalization, more or less t h e o r e t i c a l l y based, and there-fore subject to counter-example and refutation (cf. Popper 105). More important, perhaps, the conservation law says only that if_ energy leaves a system A, then i t must appear somewhere else, say B, such that A and B form a closed system; i t does not l i m i t the conditions under which energy can be transferred, and i t does not imply that any causal r e l a t i o n necessarily involves such a transfer. This has been pointed out else-where (Broad 10, Hospers 58). Skinner has argued (115) that the s c i e n t i f i c model for psychology should be that of physics. I t i s c u r i o u s — i f only i n c i d e n t a l — t h a t the whole approach of the " r a d i c a l " behaviorist school i s firmly rooted i n the mechanistic thinking which declined spectacularly i n physics around the middle of the 19th century (see, for example, Einstein and Infeld 30, part 2). Whether or not one agrees with Chomsky (16) that the results of Skinner's brand of science are vacuous, however, i t does seem reasonable to take his suggestion seriously and to examine how a more developed science has set about solving a problem which i s analogous to one facing psychology. The problem i s that of an unobservable which has observable consequences, and I suggest that a p o t e n t i a l l y f r u i t f u l approach to the question of consciousness i s to posit i t as a f i e l d effect. I t i s clear from an examination of some relevant l i t e r a t u r e that i 82 the f i e l d i s one of the fundamental constructs of physics (Einstein and Infeld 30, De Broglie 19, Bohm, 8, Weyl 128, Theobold 122, Rothman 109, Barnett 6, Kvasnica 82). Maxwell's postulation of the electro-magnetic f i e l d and quantitative derivation of i t s properties was an attempt to overcome the mechanical d i f f i c u l t i e s presented by the phenomenon of par t i c l e s seemingly acting at a distance (Maxwell 90, p. 527). While other kinds of f i e l d s are now within the domain of physics, each i s expressed as a s p a t i a l d i s t r i b u t i o n of energy defined mathematically as the capacity for action at each point and instant. The introduction of the f i e l d concept has involved a fundamental modification of the concepts of matter and space, implying as i t does that even a space devoid of bodies could s t i l l be the s i t e of continuously varying f i e l d s (see Bohm 8, Kvasnica 82, section 43). The simple idea of material e n t i t i e s as "sources" of the f i e l d was demolished by the finding that both f i e l d s and p a r t i c l e s have a wave/particle duality, involving quantal concepts and expressible only p r o b a b i l i s t i c a l l y (see De Broglie 19, chapter 3; Kvasnica 82). While Einstein seems to have been committed to the view that the whole of nature could be reduced to f i e l d s (see, for example, Bohm 8; also Weyl's related discussion of the f i e l d theory of matter, 128, p. 171 f f . ) , the question of p r i o r i t y between matter and f i e l d seems to be a "chicken-and-egg" o n e — i f one strengthens the metaphor with the addition that the closer one looks, the more the egg looks l i k e the chicken (see Theobold 122; also Bohm's conclusion that " f i e l d s and bodies co-determine each other," 8, p. 42). A potential solution to Sperry's problem (120, 121) of giving a 83 causal role to consciousness—and at the same time overcoming the prob-lematical aspects of emergentism—would be to postulate a f i e l d of consciousness. While we should beware of misleading analogies, our analysis seems to have yielded a notion of consciousness as both deter-mining and determined- by nervous system a c t i v i t y ; apparently associated with a s p a t i a l d i s t r i b u t i o n of neurons (which, indeed, comprise wave/ par t i c l e s consisting i n turn of f i e l d e f f e c t s ) , yet p u b l i c l y unobservable. I t i s also evident s u b j e c t i v e l y — t h a t i s to say, i t i s observable—that what James called the "stream of consciousness" (64) varies i n i n t e n s i t y . A quantification of this property, and, no doubt, others which I am yet unable to specify, could form the basis of a set of f i e l d equations analogous to those of Maxwell (90; see Kvasnica 82 for a "hindsight" derivation). At whatever point the analogy may reach what has been termed i t s "breaking point," any adequate f i e l d theory of consciousness surely entails some such set of equations. These would determine how the conscious f i e l d changes as a function of time, i n terms of the values of the f i e l d quantities and of the a c t i v i t y of the '(neuronal?) bodies with which i t i s associated. In regard to potential t e s t a b i l i t y , I would make the c r i t i c a l point that the empirical consequences of such a theory could very w e l l turn out to be purely subjective; there i s nothing wrong, for example, with the p o s s i b i l i t y that the analog of the "test charge" (whatever i t may be) would have observable consequences only for the person whose brain serves as the experimental apparatus. The question of just what constitutes a s c i e n t i f i c observation w i l l not be entered i n t o — I merely underline the main thesis by noting that "empirical" means having consequences for 84 someone1s experience. On this view, the organization of neural a c t i v i t y i s determined, i n part at lea s t , by a conscious f i e l d with i t s own dynamic properties. To refer back, b r i e f l y , to the experiments reported e a r l i e r , the nervous system's hierarchic f a c i l i t y of information processing i s regarded as open-ended at the top (cf. Koestler 78), perhaps the normal operating space of the f i e l d i n i t s passive mode. We may say that i n i t s active mode, the f i e l d can feed forward to affect neural response p r o b a b i l i t i e s further down the hierarchy, producing such phenomena as the context effect. Speculating further, i t seems l i k e l y that the f i e l d operates d i f f e r e n t i a l l y over the major and minor hemispheres, and possibly only i n the major—Eccles has suggested (28) that minor hemisphere a c t i v i t y becomes associated with consciousness only when information i s transferred from i t into the major hemisphere. Under the present paradigm, approaches to the question of consciou-ness (e.g., Hebb 48, John 67) are based on the i m p l i c i t assumption that subjective phenomena are to be understood only i n terms of the operation of neural elements which are observable, i n p r i n c i p l e , and which operate as machine-like analogs; consciousness i s t y p i c a l l y described as "a r i s i n g from" t h i s nervous a c t i v i t y , i n a way which no one seems to have any notion of how to specify. If we take the model of physics seriously, such mechanistic assumptions seem both r e s t r i c t i v e and unnecessary, and I argue that the derivation of a f i e l d theory could deal s c i e n t i f i c a l l y with that aspect of the universe which i s conscious, and which has not hitherto been brought into the s c i e n t i f i c domain. Even to make the attempt and find the analogy inappropriate seems worthwhile. 85 There i s no claim to uniqueness i n the preceding discussion, which i s intended as h e u r i s t i c suggestion, or at best a thematic sketch for a symphony which i s barely s t i r r i n g i n the composer's mind; i t s debt to some of Burt's l a t e r writings (13, 14) i s obvious, and some of Koffka's ideas (79) are germane. F i e l d t h e o r i e s — o r what one may only loosely term such—have been proposed before i n psychology, notably by the Gestalt school (Kohler 81, Koffka 79) and by Lewin (86; see also H a l l and Lindsey 46, chapter 6). Like some other such approaches, however (see the discussion by Zusne 134, p. 155 f f . ) , they may be characterized as providing, at best, approximate ways of dealing with the c o l l e c t i v e properties of observable elements. I conclude on the suggestive note of a p a r a l l e l between such views i n psychology and some of the early conceptualizations i n physics, before Maxwell's postulation of the f i e l d as e x i s t i n g i n i t s own r i g h t (cf. Bohm 8, chapter 2). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100081/manifest

Comment

Related Items