Science in Christian Perspective

Is A Mechanistic View of Creativity
A Contradiction?
JAMES L. ROGERS
Department of Psychology
Wheaton College
Wheaton, IL 60187

From: JASA 36 (June 1984): 105-112.

Although views on determinism and freedom are often found in this journal and other forums for Christian thought, much less frequently are the implications of such views brought to bear on specific issues. In the present paper dews of determinacy by Daniel Dennet and Donald MacKay are used to address a current controversy concerning the underlying nature of human creativity.

The present paper examines the relevance of arguments by Donald MacKay (1974) and Daniel Dennett (1978) to a current controversy concerning the underlying nature of human creativity. The starting point of this paper is the author's observation that theories and descriptions of human creativity generally can be categorized under one of two conceptual frameworks. Borrowing terms from MacKay, these viewpoints will be called the personalistic and mechanistic.

An example of the personalistic view of creativity is that held by many Christians and voiced in Dorothy Sayers'(1941) book, The Mind of the Maker. Sayers' thesis is that man is made in the image of God who creates: thus man is creative. Sayers goes so far as to relate by analogy the persons of the Trinity to the trifold processes she sees in the creative mind (idea, energy and power). She is emphatically antimechanistic and writes:

From our brief study of the human maker's way of creation, it should be fairly clear that the creator does not set out from a set of data, and proceed, like a crossword solver or a student of elementary algebra, to deduce from them a result which shall be final, predictable, complete and the only one possible. The concept of 'problem and solution' is as meaningless, applied to the act of creation, as it is when applied to the act of procreation. (p. 186)

Sayers joins her friend, C. S. Lewis, in the tradition of freedom and dignity, believing that the creative mind is indeterminable and its acts are volition-based.

The personalistic view of creativity has been criticized by some social scientists, particularly those psychologists critical of the humanistic school. Addressing the personalistic view, Albert (1975) writes: "Seen from these vantage points, creative people are heroic, mysterious, and inexplicable. But they are also not the stuff of science or, often, of this world." (p, 141) The mechanistic view holds that human beings are material and exhibit adaptive, not creative, behavior. In the very sense that Sayers detests, they are merely problem solvers and quite predictable.

The point of friction lies in the seeming incompatibility between the two views. Indeed, viewing man as a machine, regardless of the complexity, precludes "creativity" altogether, does it not? The present paper addresses this question.

Because the mechanistic view of creativity assumes some familiarity with a number of technical papers, a selected review of these will first be presented. In doing so, some strengths as well as uncertainties of the mechanistic view will surface.

To date the supporting data for this position are correlational and therefore prove nothing in the strict sense. As will be seen, support for the first assertion comes in the form of a repeated failure to partial indices of creativity (either tests of divergent thinking or peer ratings) from intelligence scores. A critical experimental test that satisfactorily separates the two may yet emerge. A redefinition and/or better indices of creativity also may develop. Second, it goes without saying that the debate over the genetic hypothesis of intelligence is not nearly settled. Nevertheless, some recent studies supporting the genetic hypothesis are of particular importance to our consideration of the creative process.

Several researchers have attempted to show that IQ and creativity are fundamentally different mental abilities. One widely cited study is that of Getzels and Jackson (1962). These investigators compared two groups of students on academic achievement where one group consisted of 26 students in the top 20 percent on five creativity measures but below the top 20 percent on IQ, while the reverse situation was true for a second group (i.e., high IQ, lower creativity). When mean standard achievement was compared, little difference emerged between the groups. On the surface these data indicate that creativity is equally important to academic achievement as IQ. Closer examination reveals a critical flaw in the statistical design. if IQ and creativity are equally correlated with academic achievement (which would be true if there is no fundamental, underlying difference between them), then selecting two groups, each high on either IQ or creativity but not both due to, say, measurement error, assures that both groups will be equally high on academic achievement. This, of course, is what Getzels and Jackson found. It is, then, only on the a priori assumption that creativity and IQ are not different names for the same underlying process that Getzels and Jackson's argument supports the independence hypothesis. But this a priori assumption rests upon the very issue needing proof, thus the argument is not germane to the question.

McNemar (1964) offers an unusually comprehensive review of early studies purporting to have shown a quantifiable distinction between IQ and creativity. Although McNemar's critique was published in 1964, the situation has not changed. The details of McNemar's critique are left to the interested reader, but his two major conclusions are worth reviewing, as they are often reiterated by the proponents of a mechanistic view. First, McNemar observes that at the outset of many investigations the assumption is made that creativity and intelligence are distinguishable. Consequently, adequate investigation into the question is not carried out. For example, Getzels and Jackson (1962) did not report the correlation between IQ and creativity in their study despite their claim that the two were unrelated. By indirect computation from information that is published a correlation of 0.40 is indicated. This figure, as McNemar argues, would be attenuated due to the usual measurement error, a highly selective subject sample (mean IQ of 132) and the use of several different instruments across subjects to measure IQ. in light of these conditions, a correlation of .40 may reflect a relatively high degree of association between IQ and creativity. As McNemar observes, Verrion (1964) found that correlations between

The argument is clear. It is that a sizeable number of individuals has not been found who exhibit a high degree of creativity but are not highly intelligent.

tests of creativity and IQ were as high as correlations among different measures of creativity. After correcting for attenuation, Shaycraft, Dailey, Neyman and Sherman (Note 1) found a correlation of .80 between IQ and creativity in 7,648 school children.nt is crucial to the mechanistic view. He notes that in those studies that have measured both IQ and creativity but have found a low correlation, samples under study were restricted to the most gifted, these groups in turn displaying a narrow range of IQ scores well above average. By significantly eliminating variability in IQ measurements, correlations between IQ and creativity were predictably low. For instance, McNemar finds Holland's (1964) conclusion that there is little relationship between intelligence and creativity unfounded because subjects were in the upper one percent of intellectual ability. Homogeneity of this degree, of course, does invalidate correlational analysis.

Other approaches to the independence hypothesis for IQ and creativity have generally coincided with the conclusion that McNemar draws-that supporting evidence for the independence hypothesis is lacking. Cox (1926) carried out biographical research on 300 persons recognized by society as highly creative. For those with sufficient data available, Cox estimated the average IQ to be above 140. MacKinnon's (1972) more recent study supports this conclusion. Wechsler (WAIS) Full Scale IQ tests were collected for 185 noted architects, mathematicians, scientists and engineers cited as among the most creative contributors to society by other professionals in their fields. After reviewing the work of MacKinnon, Jensen's (1980) concluding remarks so cogently express the sentiments of the mechanistic camp that they have been reproduced in full.

In this highly select group, the judged ranking in creativity correlated only +11 with WAIS IQ. But the more important fact, which is often neglected in popular accounts, shows the threshold relationship of IQ to creativity: the total IQ variance in this group of creative persons is less than one-fourth of the Q variance in the general population. The entire creative sample ranges between the 70th and 99.9th percentiles of the population norms in IQ (i.e., IQ's from 107 to 151), with the group's mean at the 98th percentile (IQ 131). To the extent that these groups are typical of persons whom society regards as creative, it can be said that some 75 to 80 percent of the general population would be excluded from the creative category on the basis of IQ alone. Despite claims that IQ has little or nothing to do with creativity, no one has ever published the crucial evidence that would necessarily follow from this position: a distribution of IQs among persons of generally acknowledged creativity that differs nonsignificantly from the distribution of IQs in the population. It would be absolutely astounding if such evidence could ever be found. (p. 355)

The argument is clear. It is that a sizeable number of individuals has not been found who exhibit a high degree of creativity but are not highly intelligent. When such individuals on occasion do surface, it is more reasonable, say these investigators, to appeal to measurement error than to distinct mental abilities as the explanation. In essence, the first link of the mechanistic argument states that after repeatedly failing to produce evidence disavowing the null hypothesis (i.e., that IQ and creativity are the same), it is time to accept it. This position may not be palatable to humanists or, for that matter, to statistical purists, but it is being, if not openly advocated, at least implied.

One line of support for a mechanistic explanation of intelligence comes from twin studies. These widely publicized studies have attempted to determine the relative contribution of environmental and genetic background to intelligence. Usually such studies compare MZ (monozygotic or identical) and DZ (dizygotic or fraternal) twins reared together on mental abilities, with the finding that MZ twins are considerably more homogeneous on cognitive ability scores than are DZ twins (Vandenberg, 1962; Blewett, 1954). It is reasoned that because both types of twins experience

similar environments while only MZ twins have identical genetic backgrounds, the greater similarity in intellectual ability between MZ twins is due to a common genetic background.

A criticism of studies using twins reared together is that the environments of MZ twins are more identical than the environments of DZ twins (Kamin, 1974). Because MZ twins look alike, they more often are treated alike (Scarr, 1968). Thus, it is argued that greater similarity in IQ among MZ twins could be due to their greater commonality in environment rather than genetic background. At least two recent studies suggest that although environmental factors are more similar in MZ twins, this greater similarity does not influence IQ variance (i.e., does not bias the results of studies using twins reared together). Loehlin and Nichols (1976) found that MZ twins who experienced dissimilar parental treatment and life circumstances exhibited about the same variability in intellectual ability as MZ twins experiencing very similar treatment. Scarr and Carter-Saltzman (1979) have replicated the essence of this finding in a unique way. They note that from 18% to 40% of MZ twins believe they are DZ and vice versa. Capitalizing on these false beliefs, Scarr and CarterSaltzman have compared across twins who accurately and inaccurately perceive their zygosity. The degree of accuracy of perception by self or others has little relationship to the similarity between MZ twins in intellectual ability.

A second type of twin research presents an even stronger case for the role of a genetic mechanism in determining intellectual ability. Presently underway at the University of Minnesota is a study by Bouchard and associates (Holden, 1980; Bouchard, Heston, Eckert, Keyes and Resnick, 1981; Eckert, Heston and Bouchard, 1981) that utilizes twins reared apart from birth. The study appears to be meticulously conceived, avoiding the methodological flaws of the few other existing small sample studies of this type. These investigators are carefully documenting the degree of dissimilarity in environment between twins as well as any correlation that might exist in cognitive and personality variables. Although the full report of this investigation presently is not available, the preliminary findings overwhelmingly support the genetic hypothesis. It appears that 70% of the variation in intelligence among individuals can be attributed to genetic factors and 30% to environmental factors and errors of measurement

James L. Rogers received his Ph.D. in psychology from Northwestern University, Evanston, Illinois. He presently holds appointments as associate professor at Wheaton College, Wheaton, IL and research associate at the Northwestern University Medical School. Dr. Rogers has published a number of papers in the area of man-machine performance and human learning, and recently completed a sabbatical leave at the University of Louisville Medical School where he collaborated on an NIMH funded study concerning the impact of hemodialysis on schizophrenia. Additional interests are the effects of aerobic exercise in older aged women and the psychological aspects of decision making in commodity/ futures trading.

(Bouchard, in press). Unless future analyses significantly qualify the preliminary findings, these results constitute some of the strongest support to date in favor of the genetic hypothesis.

Another line of investigation (Jensen, 1979, 1981; Jensen and Munro, 1979; Jensen, Schafer and Crinella, 1981) is thought by some to tend even stronger support to the mechanistic view (see Eysenck's response to Kamin in Eysenck and Kamin, 1981). Jensen (1979) required subjects to move their finger from a start button to one of several possible goal buttons. A given goal button was selected as quickly as possible whenever a light over it was turned on by the experimenter. The time taken to lift one's finger from the start button was called reaction time while the time to move between the two buttons was called movement time. Task difficulty was varied so that each subject would have to process either zero, one, two or three bits of information when selecting the appropriate goal button. Jensen demonstrated that a measure of general intelligence increased as reaction time improved, as the variability in reaction time decreased across different levels of information that needed to be processed, and as the within subject variability decreased across task trials at each level of required information processing. This result emerged regardless of age or social economic status.

Apparently reaction time can be used to predict general intelligence as defined by Jensen (i.e., the first principle factor on a large battery of mental tests). Jensen suggests that there is a common neurophysiological process underlying reaction time and mental test performance. He speculates that the refractory time of the neurons that fire during the selection of the appropriate light may be involved. Although there is no proof at the moment, many scientists find it difficult to imagine something so basic as reaction time, measured in milliseconds, as not being fundamentally tied to an individual's genes. Twin studies that directly examine the heritability of RT will no doubt soon appear.

As a final point, it should be emphasized that even if intelligence were shown not to be genetically determined in any sense and/or not to be important to creativity, a mechanistic explanation of creativity still would not necessarily be ruled out, The strong assumption of many social scientists would be that environmental factors, in an equally mechanistic fashion, would explain the creative process. B. F. Skinner's views on environmental determinants of behavior and personality constitute an example of this position.

A more striking example of this line of reasoning is found in recent work by Gartner and Baunack (1981). These investigators compared MZ twin mice with isogenic: DZ twin mice. Isogenicity in DZ twins was accomplished by repeated brother-sister mating over more than 150 generations and was verified biochemically. MZ twins were derived by splitting 8-cell embryos into two 4-cell entities and then transferring them to the uterus. Although the statistical analysis in the original paper was flawed, it appears that the main result stands (Jinks, et al, 1982). Monozygotic: twins had a greater degree of similarity on genetically loaded parameters (body weight, time to eye opening and appearance of hair) than dizygotic twins. Note that both MZ and isogenic DZ twins were genetically similar, but MZ twins shared exactly the same position, location and events within the uterus before the third cell division (i.e., before two 4-cell entities were created) while DZ twins did not. That MZ twins are found to be more similar than isogenic DZ twins, argue these investigators, apparently is due to an identical early intrauterus

. ..There might not be a unified view available to finite man that would capture the fullness of the creative process, however desirable that might be, and thus both the personalistic and mechanistic views Of creativity would be necessary.

environment that has directed, at least to some extent, subsequent biological occurrences. The important point to the present discussion is that here "environment" is not being thought of as something easily manipulated by the subject or others, although in the strict sense some form of at least indirect, purposeful manipulation during the 31/2 days between coitum and the third cell division cannot be ruled out. Undoubtedly the environmental influence that Gartner and Baunack have in mind is a mechanistic event that presumably can be identified through future research.

From this brief review, the direction of thought for a significant portion of the scientific community seems clear. increasingly, the working assumption is that a satisfactory mechanistic explanation of intelligence (and therefore creativity) will emerge given adequate research and time.

If man retains something of the image of God, then man's creativeness reflects to some degree that image. But it is dif f icult to think of God in terms of genes and environment! Thus it is difficult to think of man that way. Is it possible to reconcile an empirically based mechanistic view with the heartfelt personalistic view?

In what follows I rely upon Daniel Dennett (1978) and Donald MacKay (1970, 1974, 1979), using their arguments and analogies to hoist the antireductionistic thesis that the limitations of man, as well as the richness of the stimuli that impinge upon him, might require that explanations be from multiple perspectives. That is, I will argue that there might not be a unified view available to finite man that would capture the fullness of the creative process, however desirable that might be, and thus both the personalistic and mechanistic views of creativity would be necessary.

The physical stance, Dennett writes, "is generally reserved for instances of breakdown, where the condition preventing normal operation is generalized and easily locatable, e.g., 'Nothing will happen when you type in your question, because it isn't plugged in.' " (p. 237). This stance is used by the repairman or the physicist designing the new microchip to replace the old one. The design stance, on the other hand, is more global and uses higher-order units of analysis. Here we might think of many logic gates wired together to form operations that in turn make up any number of computer programs. The sine qua non of the intentional stance is the assumption of rationality in the system. Dennett emphasizes that perfect rationality is not required, but to qualify as intentional, a system must be of sufficient complexity that its "behavior can be predicted, and most efficiently predicted, by adopting the intentional stance toward ... [it]." (p. 238) The final stance posited by Dennett, the personal stance, can be thought of as the intentional stance plus a moral commitment. It would not, for example, be such a terrible thing to dismantle a computer to see what was inside, but the same could not be said about a man because a moral commitment is involved.

MacKay, in a similar vein, distinguishes between the personal aspect" and the "mechanistic aspect" of systems. Respectively, these correspond approximately to Dennett 's personal plus intentional stance where the former presupposes the latter, and Dennett's physical or design stance where MacKay's term includes both. As noted earlier, it is from MacKay that I have borrowed the terms "personalistic" and "mechanistic" used throughout this paper.

Both Dennett and MacKay argue that the personal and mechanistic stances can coexist as equally valid forms of explanation, neither stance reducible to the other. If this position is accurate, then the personalistic and mechanistic views of creativity can be viewed as complementary. In what follows, I will present three arguments for complementary explanations, two by Dennet and one by MacKay.

Dennett (1978) begins his first argument by quoting MacIntyre:

Behavior is rational-in this arbitrarily, defined sense-if, and only if, it can be influenced, or inhibited by the adducing of some logically relevant consideration.... But this means that if a man's behavior is rational it cannot be determined by the state of his glands or any other antecedent causal factor. For if giving a man more or better information, or suggesting a new argument to him, is both a necessary and sufficient condition for, as we say, changing his mind, then we exclude, for this occasion at least, the possibility of other sufficient conditions.... Thus to show that behavior is rational is enough to show that it is not causally determined in the sense of being the effect of a set of sufficient conditions operating independently of the agent's deliberation or possibility of deliberation. So the discoveries of the physiologist and psychologist may indefinitely increase our knowledge of why men behave irrationally but they could never show that rational behavior in this sense was causally determined. (Note 2)

If MacIntyre is correct, explanations from the personal (intentional) stance, because they assume rationality, cannot coexist with mechanistic explanations. Dennett counters this line of reasoning by considering the tropistic, behavior exhibited by the sphex wasp. The sphex manifests certain ritualistic patterns of behavior when preparing to lay her eggs that will be repeated from the beginning upon interruption, even though it would be more "logical" to continue on from the point of interruption. Suppose, Dennett reasons, that we were to build a better wasp, taking into account literally millions of possible interruptions that might occur with internal instructions that would enable the sphex to continue rather than start over. If the wasp builder were clever enough, an observer might never guess that the wasp's behavior was tropistic, even though it would be. Because of the observer's limitations as a perceiver and data processor, he would be forced to view the wasp from an intentional stance. Might not the behavior of the "super wasp" be analogous to that of a man who creates, say, a painting? Dennett's response is delicate and therefore will be quoted verbatim.

The grain of truth in MacIntyre's contention is that any system that can be explained mechanistically-at whatever lengthmust be in an extended sense tropistic, and this can enhance the illusion that mechanistic and intentional explanations cannot coexist. But the only implication that could be drawn from the general thesis of man's ultimately mechanistic organization would be that man must, then, be imperfectly rational, in the sense that he cannot be so designed as to ensure rational responses to all contingencies, hardly an alarming or counterintuitive finding; and from any particular mechanistic explanation of a bit of behavior, it would not follow that that particular bit of behavior was or was not a rational response to the environmental conditions at the time, for the mere fact that the response had to follow, given its causal antecedents, casts no more doubt on its rationality than the fact that the calculator had to answer "108" casts doubt on the arithmetical correctness of its answer. (p. 246)

Clearly, Dennett is arguing that both types of explanationpersonal and mechanistic-are valid. In MacKay's language, it would be incorrect to reduce our wasp to "nothing but" a mechanistic gadget.

The second two arguments come in the form of an answer to the question: if a man's behavior can be mechanistically explained, then in what sense is he responsible for his actions? It is from the personal stance that acts are to be understood in terms of responsibility, thus upholding responsibility in the presence of a mechanistic explanation presupposes the validity of the personal stance. Translated to terms relevant for the creativity issue, responsibility denotes the ability to choose, and the ability to choose is fundamental to the personalistic view of creativity.

Consider the possible outcomes of an electrode implant in Tom's brain that would, if it were possible, make Tom believe he has a brother in Cleveland which in fact he does not. This would be similar to rewiring our wasp. If we can make Tom believe anything we want by adjusting his mental machinery, then is not an understanding of the machinery the only

Translated to terms relevant for the creativity issue, responsibility denotes the ability to choose, and the ability to choose is fundamental to the personalistic view Of creativity.

explanation of Tom's behavior that is needed or even possible? Dennett's answer comes via an analysis of the extremes of what might happen to Tom. First, someone might ask Tom if he is an only child, to which he might reply, "I have a brother in Cleveland." Tom might then realize that what he has said does not agree with everything else he knows and change his mind. In that case, it could hardly be said that manipulating Tom's brain relieves Tom of his rationality or his responsibility to tell the truth. Second, suppose the surgeon implants Tom with a set of new false beliefs, and also destroys true beliefs. Or, suppose the surgeon revamped Tom's sense of logic so that he would say bizarre things like "I am an only child and have a brother in Cleveland." Dennett argues that if Tom were stripped of so many of his original thoughts that it would be unreasonable for us to expect him to reject "new thoughts" by comparing them with old, or if Tom were deprived of his understanding of logical relationships, then of course we would conclude that Tom was not rational, but rather insane. This would place Tom in a dif f erent arena than ordinary men, and we would be forced to seek explanations of Tom from the physical and design stance ("something has gone wrong chemically!"). It would also be true, however, that Tom, no longer viewable from the intentional stance, could not be conceptualized in terms of responsibility. Tom, the original "object" of our examination, has been so radically changed that for all practical purposes we are looking at a different object. Dennett summarizes as follows:

One cannot directly and simply cause or implant a belief, for a belief is essentially something that has been endorsed (by commission or omission) by the agent on the basis of its conformity with the rest of his beliefs. One may well be able to produce a zombie, either surgically or by brainwashing, and one might even be able to induce a large network of false beliefs in a man, but if so, their persistence as beliefs will depend, not on the strength of any sutures, but on their capacity to win contests against conflicting claims in evidential showdowns. (p. 252)

The final argument for complementary explanations is due to MacKay. (For further discussion of MacKay's argument by evangelicals, see Cramer and MacKay, 1976; Hasker, 1978a, 1978b; MacKay, 1978 and MacIntyre, 1981.) The question MacKay seeks to answer is, Can a man be denied responsibility for his actions if his brain is "as mechanical as a clock?" To address this question, MacKay postulates the existence of a super-scientist who understands Joe's brain and is able to correctly predict that it will move from state A to A'. We would be incorrect, MacKay tells us, to claim that Joe is obliged to believe our super-seientist's prediction if only he knew it, because the act of telling Joe the prediction would upset the basis of the prediction (i.e., Joe's brain would be in a state other than A). Further, if the super-scientist were to predict a brain state that would become accurate only if Joe were told the prediction and believed it, then until Joe in fact "believed," he would not be in error to disbelieve the prediction, because at the moment there would be no basis for the prediction. MacKay (1970) summarizes the impact of this peculiar state of af fairs as follows:

To test the strength of the argument, let us take the most favourable case imaginable by the determinist. Suppose that from our observations of Joe's brain and its environment we can write down, well in advance, a whole series of predictions that we keep secret until after the events, and then triumphantly produce to Joe as a proof of our success. By showing him a cine film of the calculating process, we convince him beyond doubt that our mechanistic theory of his brain is correct. Would not this show, at least retrospectively that he was mistaken in believing that he was facing genuinely open possibilities? It would not. It would show, of course, that the outcome was predictable by us. What it would not show is that it was inevitable for him. It cannot do so, for it cannot produce a specification of the outcome that Joe would have been unconditionally correct to accept before he made up his mind. In this sense, no matter how many detached observers could predict the outcome, Joe-and you and I-are free in choosing. We cannot logically escape (or be denied) responsibility for our choices on the grounds of their predictability by non-participant observers. (p. 25)

MacKay (1979) is claiming that our future "has no fully determined detailed specification for [us] to discover, with an unconditional claim to [our] assent." (p. 53)

Conclusion

I have reviewed some of the supporting material for the argument that intelligence and creativity are the same mental ability. Further, attention has been called to some recent studies that advocate more strongly than ever that intelligence is primarily a function of one's genetic inheritance. If these data and the implications drawn from them by many, are valid, then it is tempting to speculate that creativity someday will be equated with a well-defined genetic mechanism, Would the existence of a material explanation of creativity supplant a personalistic explanation? Using arguments by Dermett and MacKay, I have presented a case for the compatibility of parallel explanations in the mechanistic and personalistic domains.

Through examination of the creative process from complementary perspectives (the mechanistic and personalistic), a more complete understanding of the complexity of the creative act is provided than would be possible using only one. If Dennett and MacKay are correct, these two views need not produce endless, unproductive attempts to replace one with the other.

An objection raised concerning the complementarian approach applies to the present discussion of creativity. Would not a unified, holistic perspective be better? Is it not premature to give up the search for a single framework from which the creative process might be understood? To this

Through examination of the creative process from complementary perspectives (the mechanistic and personalistic), a more complete understanding of the complexity of the creative act is provided than would be possible using only one.

We see the implications of this possibility in the theories of some other psychologists as well. For example, Paivio (1975) has performed experiments concerning the way humans encode incoming stimuli. His "dual coding" hypothesis suggests that there are two independent (but interconnected) mechanisms of memory: an imagery system and a verbal system. While discrete semantic stimuli are stored in verbal memory, Paivio believes that something qualitatively different is stored in the imagery system. Here, representations of the external world that are analogue and holistic in nature reside. For example, it can be demonstrated that large perceptual memory differences are detected faster than small differences-e.g., subjects are slower to report that a mouse is smaller than a squirrel than to report that a mouse is smaller than an elephant (Moyer, 1973). In a discussion of encoding theories and research, Kintsch (1977) has made the following comment:

We conclude with Paivio ... that the evidence for an imagery system, independent but interrelated with the verbal system and analogue in nature, is overwhelming. Nevertheless, it appears that his dual coding theory may still be unsatisfactory. First of all, Paivio concentrates on two modalities, imaginal and verbal. What about acoustic memory codes, and other sensory modalities? Most importantly, what about abstract, conceptual memory traces that are neither words nor images?.... It seems that what is needed is not a dual-coding theory, but a multiplecoding theory! (p. 243)

The present purpose is not to describe the details of the various encoding mechanisms to which Kintscb alludes, but rather to note that the presence of different kinds of memories, each encoding a unique aspect of the outside world, suggests that parallel, complementary explanations of observed phenomena (e.g., the creative process) may be a function of the physiology of our brain-an innate characteristic not likely to change any time soon.

As a concluding remark, I would emphasize that a person is more than the language be or she uses to communicate observations. It does not follow from anything said here that it would be correct to reduce a person to two or more materialbased encoding mechanisms! Rather, the point is that the whole spectrum of observable human activities, including creative acts, is of such complexity that our ability to explain at an integrated level our own activities is inadequate. We are awed by our greatness, but humbled by our inability to comprehend and explain the same. The actual experience of creativity in ourselves, or the vicarious experience of creativity in another, might well transcend an explanation per se from any given perspective. Perhaps all perspectives are needed, but all perspectives together remain insufficient. To simultaneously and/or successively experience multiple perspectives may be, in the language of the Gestalt psychologists, the whole that is more than the sum of the parts, and might well leave us feeling that surely an integrated understanding ought to be possible. But at the level of information flow from one individual to another, or within the cognitive system of a given individual, descriptions from multiple perspectives may be the best that are possible.

It is interesting to speculate about the present thesis in light of the fall. Could not the inability to holistically comprehend and describe what we observe be one manifestation of the fragmentation caused by sin? And would not a continual search for reinstatement of a unified perspective, and even a strong belief that such ought to be possible, underscore the present creation's yearning for a new, perfect creation order?

²Maclntyre, A. (Unreferenced citation by D. Dennett, 1978).

REFERENCES

Albert, R. Toward a behavioral definition of genius. American Psychologist, 1975,30,140-151.

Blewett, D. An experimental study of the inheritance of intelligence. journal of Mental Science, 1954, 100, 922-933.

Bouchard, T. Twins-natures twice-told tale. Encyclopedia Britanica 1983 Yearbook of Science and the Future (in press).

Bouchard, T., Heston, L., Eckert, E., Keyes, M. and Resnick, S. The Minnesota study of twins reared apart: project description and sample results in the developmental domain. In L. Gedda, P. Parisi and W. Nance (Eds.), Twin Research 3: Intelligence, Personality, and Development. New York: Alan R. Liss, 1981.

Cox, C. Genetic Studies of Genius: The Early Mental Traits of Three Hundred Geniuses (Vol. 2). Stanford: Stanford University Press, 1926.

Cramer, J. and MacKay, D. The clockwork image controversy. Journal of the American Scientific Affiliation, 1976, 28,123-127.

Eckert, E., Heston, L. and Bouchard, T. MZ twins reared apart: preliminary findings of psychiatric disturbances and traits. In L. Gedda, P. Parisi and W. Nance (Eds.), Twin Research, 3: Intelligence, Personality, and Development. New York: Alan R. Liss, 1981.

Eysenck, H. and Kamin, L. The Intelligence Controversy: H. J. Eysenck vs. Leo Kamin. New York: Wiley, 1981.

Garrett, S. Putting our whole brain to use: a fresh look at the creative process. Journal of Creative Behavior, 1974, 8, 239-249.

Girtner, K. and Baunack, E. Is the similarity of monozygotic twins due to genetic factors alone? Nature, 1981, 292, 646-647.

Getzels, J. and Jackson, P. Creativity and Intelligence. New York: Wiley, 1962.

Hasker, W. MacKay on being a responsible mechanism: freedom in a clockwork universe. Christian Scholars Review, 1978, 8,130-140.

Hasker, W. Reply to Donald M. MacKay. Christian Scholars Review, 1978, 149-152.

Holland, J. L. Creative and academic performance among talented adolescents. Journal of Educational Research, 1964, 6, 163-169.

Jensen, A. g: Outmoded theory or unconquered frontier? Creative Science and Technology, 1979,2,16-29.

Jensen, A. Reaction time and intelligence. In M. Friedman, J. Das and N. O'Connor (Eds.), Intelligence and Learning. New York: Plenum, 1981.

Jensen, A. and Munro, E. Reaction time, movement time, and intelligence. Intelligence, 1979,3,121-126.

Jensen, A., Schafer, E. and Crinella, F. Reaction time, evoked brain potentials and psychometric g in the severely retarded. Intelligence, 1981, 5, 179-197.

Jinks, I., Haley, C., Last, K., G5rtner, K. and Baunack, E. Are monozygotic twins similar due to genetic factors alone? Nature, 1982, 295, 172.

Kintsch, W. Memory and Cognition. New York: Wiley, 1977.

Loehlin, J. and Nichols, R. Heredity, Environment and Personality. Austin: University of Texas Press, 1976.

MacKay, D. The bankruptcy of determinism. The New Scientist, 1970, 24-26.

MacKay, D. The Clock Work Image. Downers Grove: InterVarsity Press, 1974.

MacKay, D. Responsible mechanism or responsible agent? a response to William Hasker. Christian Scholars Review, 1978, 8, 141-148.

MacKay, D. Human Science and Human Dignity. Downers Grove: InterVarsity Press, 1979.

MacKinnon, D. and Hall, W. Intelligence and creativity. Proceedings of the Seventeenth International Congress of Applied Psychology, 1972, 2, 1883-1888,

MacIntyre, J. A pictorial representation of MacKay's argument for human freedom in a mechanistic universe. Journal of the American Scientific Affiliation, 1981,33,169-171.

McNemar, Q. Lost: our intelligence? why? American Psychologist, 1964, 19, 8714~82.

Moyer, R. S. Comparing objects in memory: evidence suggesting an internal psychophysics. Perception and Psychophysics, 1973, 13, 180-184.

Ornstein, R. The Psychology of Consciousness. New York: Harcourt Brace Joyanovich, 1977.

Paivio, A. Perceptual comparisons through the mind's eye. Memory and Cognition, 1975,3, 635-647.

Scarr, S. and Carter-Saltzman, L. Twin method: defense of a critical assumption. Behavior Genetics, 1979, 9, 527-543.

Vandenberg, S. The hereditary abilities study: hereditary components in a psychological test battery. American Journal of Human Genetics, 1962, 14, 220-Z37.

Vernon, P. E. Creativity and intelligence. Educational Research, 1964, 6, 163-169.