We are coming to conclusions. I need, therefore, to state, for you but for myself as well, what I have said and why I am asking you to believe it.
I have tried to "the" an "I." I am trying to make it possible to talk rigorously, even, in a sense, impersonally, about the inner, experiencing being whom you or I call "I." To do so I am suggesting a threefold concept of identity. I ARC. "I" is an agency, a representation, and a consequence.
As an agency, the I tries out on the world the guesses, hypotheses, strategies, or behaviors by which we create perceptions that we find satisfying. As a consequence, the I is what results when those hypotheses and strategies and behaviors feed back as perceptions onto--into--the self. The I as consequence is the sum or history of all that has happened to the I, everything the I as agent has shaped into a coherent self in response to the "feel" of those feedings back.
That coherent self is also, however, somebody's representation of a somebody. An I is not just something "there," inside us, which our descriptions come more or less close to. We do not "have" identities. Rather, identity is the way somebody represents an I, usually in words. Identity as I propose it is a relational term. Identity is between people, and because identity consists of words between people, identity presupposes a cultural realm of shared ideas and shared ways to represent them--language.
Specifically, by thinking of identity as a theme and variations (and the history of that theme and variations), we acquire a way to represent in words the wholeness and historicity of a person. A theme and variations can represent the dialectic of sameness and difference, the mingling of continuity and discontinuity, the dialogue of change and constancy which is a human life.
If we combine Heinz Lichtenstein's idea of identity as a theme and variations with identity as the agency and consequence of feedbacks, we can picture our minds as a chain of higher and lower networks, each lower loop answering to the standards of a higher, with a theme and variations identity setting levels throughout the whole network. With such a picture, we can interrelate either the small details of a life, one analytic hour during Freud's treatment of the Rat Man, for example, or cataclysmic changes, like the brainwashing of Dr. Vincent or Anna S.'s breaking out of her painful cycles of drunkenness and prostitution. We can picture how we share symbols with other humans, yet each use those symbols with a personal style. We can understand how we are able to perceive the world in ways common to other people who live in our culture or who have the same biological equipment as we, yet also show a distinctly personal style in our perceptions.
We can propose an answer to the ancient problem of the one and the many: How can one human individual, single and unique, come out of experiences all or many humans have? Why does each of us use our common human experience of being born, mothered, fed, cleaned, clothed, loved, frustrated, punished, challenged, and limited to become an individual with a unique history? Because we style what happens to us. Each of us shapes the experiences we share with other humans through an identity theme, its variations, and its unique history. Hence each of us experiences these events in our own personal way.
With this theme and variations concept of identity, we can add to Freud's monumental discovery of infantile stages like orality the dialectic by which they lead not only to human types but to human uniqueness. We can trace that dialectic throughout the individual's history from infancy to old age, the whole trajectory of a human life.
Experimental psychologists, however, will question the kind of explanation I am offering. To be sure, one can use identity to account for an astonishing range of human experiences, but are the explanations scientific? Or is all this no better than astrology? And does that matter? some humanists would reply. Since all ways of construing the world are fictions--human inventions--isn't one fiction as good as another?
No. I do not believe that all fictions are equal, and it does matter to me what claim I can make on you to believe what I have written. It matters terribly to me. Behind The I there is a real I, no fiction he, and his needs for certainty and consistency and confirmation are fierce indeed. This theory of I is very much a function of
I am a student of literature. If I want to analyze the I who is writing The I, I turn to his books, as I did with Scott Fitzgerald and Herbert Graf. I think I could arrive at a fairly clear picture, although I hesitate at this point to inflict any more of this I's prose on you. Here is a paragraph from early in The I (quoted from my journal of 1951):
I am glad to have kept this book, even as sketchily as I have. Someday I shall look back, and when I do I daresay the then-I will wonder what the now-I was like, just as the now-I wonders about the then-I. . . .The passage proceeds by means of two sharply divided pairs, the now-I and the then-I, the then-I looking back to the past and the now-I looking forward into the future. At the same time, the prose doubles back on itself, as if to say, these seem different, but they aren't. The writer wants something that will resolve the dichotomy, an I or an end to wondering that will make the seeming two really one. The two mysteries he probes are time and the self, and the tactic that may provide an answer is looking at "this book."
I wrote my first published book out of my excitement at the then critical mode of looking "objectively" at texts not for some other reality they supposedly recorded but simply as words-on-the-page with a unity. In The First Modern Comedies (1959), I analyzed eleven plays of the late seventeenth century, using the close reading that was yielding so many rich interpretations of literature in the late fifties. I developed a unity for each play by showing how its several plots, characters, and figures of speech could be made mutually significant around a central theme (in much the same way that I have "made sense" of the Rat Man, Dr. Vincent, or Scott Fitzgerald in this book). I went on to put these themes themselves into a still larger unity and claim that Restoration comedy in general dealt with one basic theme, the tension between social surfaces and inner nature and the mastering of that tension--another dichotomy, between inside and outside. I though I had discovered a truth.
Shortly thereafter, I began a television series teaching Shakespeare in the same straightforward, words-on-the-page, unpsychological way as, in an earlier series, I had analyzed current movies. I was continuing to focus objectively on surfaces and appearances, as I had with Restoration comedy. I was treating Shakespeare as a surface, as simply the words on the page, a verbal artifact to be analyzed as one would analyze a painting or a piece of music, or as a film makes people and things into a surface on a screen. I was interpreting the plays in what I thought was a rigorous, quasi-scientific way, arriving at the meaning from a surface that was undeniably there.
In the series and the book it gave rise to, The Shakespearean Imagination (1964), I prized "objective" and scientific truths, the safe ones, and I contemned what I perceived as fuzzy terms, vague boundaries, or unverifiable intuitions and values. I wanted either to see just a clear, daylit surface that we could all agree on or to make all of whatever I was observing into such a surface (as, in a strange, almost deconstructive way, by this very act of self-exhibition, I am still trying to do). I had a powerful curiosity about the self as it related to literature but an equally strong inhibition against entangling myself with that self. Hence, I engaged in paradoxes, like confining myself to the analysis of literature as a verbal surface but searching the depths by using psychoanalysis to study literature.
I had always been interested in laughter, perhaps because it, too, is something undeniably there, a reassuring translation of something deeper and perhaps more dangerous to a behavioral surface. In writing and teaching about comedy, I had read many of the hundreds of theories of why we laugh. Of them all, only Freud's, I felt, encouraged me to look in detail at the surface text of a joke, which I, thriving on the critical methods of the 50s and 60s, thought the only valid procedure. I began to wonder whether I could adapt Freud's theory of amusement to make a general theory for all of literary response. The more psychoanalysis I read, the more wondering became belief. One could, I was sure, understand--master, really--our responses to literature by a psychoanalytic analysis of the text one was responding to.
In 1966 I published Psychoanalysis and Shakespeare. Mostly the book continued my study of texts as objective, isolable words-on-the-page, but the last chapter dug under the firm boundaries with which critics of those days marked off author from text from reader. Rather, I urged, the literary transaction forms a continuum. Characters become real because we endow them with our own psychic reality. Readers give literary events a part of the readers' own psychic functioning. Hence the critic needs to accept his own role in the literary experience he creates, and for "the" critic, read "I"--this I.
Partly The Dynamics of Literary Response (1968) treated texts as the words-on-the-page with built-in conscious and unconscious meanings to be discovered. Partly the book acknowledged that the reader--this I--contributes to the experience of the text. We put something of ourselves into the "out there" of Romeo and Juliet. Yet I grounded my argument on the "scientific" Heider-Simmel experiment, which showed that people interpret all kinds of experiences, even those of little triangles, as acts of human beings. I was still on the side of a self that could be seen, proven, made certain and scientific, and so mastered.
When I began to test the theories of Dynamics with actual readers, I found that these real, live readers edited what they read as much as any Maxwell Perkins. They interpreted, remembered, and even perceived according to their own lights. I became more and more convinced intellectually that the personal style in which we transact literature is also a style that permeates every aspect of our being. Poems in Persons (1973), 5 Readers Reading (1975), and Laughing (1982) all rest on the careful observation and interpretation of actual readers or laughers. All deal with surface behaviors: words-on-the-page or at least on the tape. All begin with at least an aspiration to the "scientific." Yet all led to the same conclusion: that we transact the world by means of our identities, that therefore interpretation, either of books or people, indeed the very isolating of words-on-the-page, is a function of somebody's identity. Certainty led to uncertainty.
Here are some concluding paragraphs from the last of these books before The I:
Laughing involves its theorists in their whole philosophy or in my case (yes, "case") my whole psychology. And I use that word too in a double sense: the psychology of me and the psychology I believe in. How, according to identity theory, could it be otherwise? In explaining, we slowly and strugglingly re-create our identities.Lo and behold, we have come back to the style of that early journal. Again we can find the dichotomies, from the double sense of "case" to the contrast of "know" with "think I know." We can find the same doubling back, the play on "case," the balancing off of "slowly and strugglingly" against "suddenly and playfully," and the return (as in the "now-I" and the "then-I" and this very sentence) to the self of long ago. We can find now an explanation (identity) that resolves the dichotomies in an overriding unity, but seeing and understanding and knowing certainties remain paramount values. We can find the same wonderment at the passage of time ("twenty-five years ago," "a moment of ourselves"), and it too is rendered in a dichotomy.
In laughing, we suddenly and playfully re-create our identi ties. & quot;S lowly, " " strugg lingly ," "suddenly," "playfully," and "identities" themselves, however, are functions of our identity, and our identity in turn is a function of the identity of the person (even a moment of ourselves) construing that identity.
Identity, then, offers but one small step in a dialectical understanding of many laughings. Nevertheless, I "know"--or at least I think I know--more than I did when I began my long inquiry into laughing twenty-five years ago.
My I of 1984 can read continuities between my I of 1951 and my I in 1982. I still think the careful observation and analysis of verbal surfaces is the best way to understand the personal depths. I still use that divided way of thinking about the world, surface and depth, although the very identity theory I advocate renders it obsolete.
That is, identity theory says we take in the other--literature, people, society, politics, culture, even our own genders and selves--through our identities, which are themselves representations. That is the thesis of this book, and I believe it is true. Emotionally, however, I find that a very hard position to hold. I still want a firm and reassuring division between the I and what the I looks at, what I once would have called reality and now would call otherness. My prose sometimes, even with editing, lapses into that older mode of an I definitively and dichotomously different from a not-I.
I find identity theory emotionally hard to believe because of these paradoxes. That which is most our own to have or to be, our essential wholeness, is not our own, because the minute we review that wholeness, either we must divide ourselves into an observing self and a self observed and so cease being whole, or our supposed wholeness is something someone else must provide us.
No philosopher, I cannot resolve these paradoxes. I find them deconstructive, provocative, hard to take, but intellectually necessary. Hence, when I use these ideas of identity and knowledge and development and objectivity, they give me control and mastery but they also control me. That is, I feel helpless as I try to work with the between-ness of a self I once thought autonomous. I feel weakened by intrinsic limitations to science and human knowledge. I feel split if one me is really two, a self I feel immediately and a represented self. I can almost taste my need for a more traditional kind of certainty.
Identity, however, calls into question the basic polarities around which we have up to now defined certainty and the "scientific": self and other, "in here" and "out there," the invented and the discovered. These are the very polarities I characteristically favor as a base for making a reassuring sense of the human transactions around me. Identity says that we each transcend those seeming opposites, however, even when we are doing science. Identity throws everything into between-ness, leaving us no mooring outside or beyond the sea of human relations. Identity offers only itself as the ultimate reality, the kind of thing a self, a cogito, was, but identity is, precisely, relationship and therefore requires other realities for it to be between.
Hence, I think this theory of "the I" has shifted the usual debate. No longer is it enough to ask, Is psychoanalysis "scientific"? Is identity "scientific"? And then I would parry with, What do you mean by "scientific"? The question has become, Is physics "scientific"? Is biology? Is anything? If minds construe the world in individual ways, if this is the way it is and nothing is going to change that personal style of perception, then what do we mean by
As a non- or ex- or semiscientist, I carry around in my head a notion of science as rigorously confirmed, numerically exact, predictive, "objective," and unblemished by the jealousies, ambitions, and institutional politics that afflict less disciplined occupations. I suspect that most nonscientists share this idealistic picture. Scientists themselves, however, have no such illusions.
Theirs is a world of Murphy's Law and the Fudge Factor. They are willing to race ahead of the evidence and the rules drawn up by philosophical or psychological bystanders. They have to be--if we are to have discoveries. Nor do scientists accept or reject ideas for purely, "objectively" rational reasons. As Max Planck lamented, "A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it." (1949, pp. 33-34). A reading of The Double Helix will quickly dispel the illusion that scientists live by impersonal rigor, as will the occasional scandal of forged and fraudulent results among medical researchers.
Psychologists and others who rattle off criteria for something's being "scientific" often speak of prediction as a necessary condition for science. That may be true for the statics and dynamics customarily studied in freshman physics, eighteenth-century science, really. Prediction does not come about as easily in more modern fields, such as quantum physics, fluid mechanics, strength of materials, or, most notably, twentieth-century biology. In general, says the Nobel-winning biologist Jacques Monod, "The biosphere does not contain a predictable class of objects or of events but constitutes a particular occurrence, compatible indeed with first principles, but not deducible from those principles and therefore essentially unpredictable" (1970, p. 43).
The world of living organisms is unpredictable because the base from which we understand it, evolution, is unpredictable. Natural selection involves two intrinsically random and unpredictable processes: mutation and survival. Indeed, much of the scientific (as contrasted to religious) resistance to Darwin's theory in the nineteenth century came from the demand that a theory, to be scientific, yield predictable results. Much of twentieth-century biology, of course, like much of our physics, deals with the random and the probabilistic, making prediction impossible (Mayr, 1978, pp. 48-51).
As a layperson, I tend to think of scientists as proudly possessing a broad consensus about "the way things really are." Here again, this is not a view shared by scientists. Another Nobel laureate, Percy Bridgman, remarks, "I believe that in society as at present constituted the possibility of consensus, except with respect to the simplest situations and as a first approximation, is a mirage. There is no such thing as true consensus, and any ostensible reality supposed to be revealed by the consensus does not exist" (1959, p. 246).
I have to rely on quotations like that from Bridgman's book because I am not a scientist and I do not know how things look and feel from inside, so to speak, the society of science. Obviously, however, I do not mean to suggest by these quotations that science has lost its awesome coherence, only that what we of the laity tend uncritically to believe about science is not what scientists them-selves believe. In particular, the real world of working scientists is not governed by a few simple rules that can be easily set down.
Jacques Lacan, who in this respect is as laic as I, writes, "Science is an ideology of the suppression of the subject" (1970, p. 89). Again, this idea is something that we laypeople carry around in our heads. Scientists themselves know that science rests on several personal, "subjective" elements. Science to scientists is simply not an "impersonal" or "objective" activity.
Mathematician Jacob Bronowski develops this human element in science by drawing on the psychology of perception: "We are not able even to receive visual impressions except by a process of indirect inference. Inferences are, therefore, at the root of all our mental processes, even those exercised directly through the senses" (1967, p. 22). Brain physiologist Ragnar Granit warns, "We must not underestimate what the interpreting brain itself adds to make the seen world more intelligible" (1977, p. 128). And philosopher Stanislav Andreski spells out the necessity of personal perceptions to the scientist still more bluntly:
If you ask a physicist to tell you how he tested a hypothesis he will say: 'I did this, I did that; I saw this and that. . . . If you disbelieve him and he invites you to take part in experimenting you will say: 'Ah, now I see . . . this moves here and that moves there . . . now I see such a colour or line or what have you.' Thus you cannot give an account of the evidential foundations of physics without hearing and uttering 'I.' And what kind of meaning can you attach to this word without . . . introspection; and without postulating the existence of other minds within which processes are taking place which are similar to those which you alone can observe?In other words, you cannot express physics in the language of physics alone because its theories rest on the evidence of the senses. Scientific research takes place within the frame of human perception, dependent on an I (1972, pp. 21-22) and also on a theory of the I.
Just as important as perception are our processes of belief and especially commitment to our beliefs. Scientific laws are made as much as they are discovered, because laws rest on human beliefs and human self-knowledge. Bronowski develops the position this way:
If I write a paper and it goes to China and Czechoslovakia and South America and Los Angeles everybody in all these places who reads it believes that I am telling the truth as I see it. Nobody assumes that what I am saying is true. It is not given to us to know what is true in that sense. But everybody knows that I write the scientific paper on an implicit, unwritten understanding among scientists that it can be absolutely believed to be what I believe.It is from this trust between persons that science proceeds, and science has been so successful only "because it is based on perfect trust in the truth of statements." "You cannot take the simplest statement in science without having to believe a lot of people." Hence, the constant increase of knowl-edge in science rests on an ethical base (1967, pp. 125, 131, 130, 129).
The biologist Monod makes the same point:
In an objective system . . . any mingling of knowledge with values is unlawful, forbidden. But [the] . . . "first commandment" which ensures the foundation of objective knowledge, is not itself objective. It cannot be objective: it is an ethical guideline, a rule for conduct. True knowledge is ignorant of values, but it cannot be grounded elsewhere than upon a value judgment . . . (1970, p. 176).In other words, the "is" of science rests on the "ought" of scientists.
The ethical imperative "Report only what you believe to be true" is itself absolute, but it can by no means be as clear-cut as the actual observations the chemist or the biologist makes. Do you report the room temperature and the barometric pressure (as we were required to do when I was an undergraduate in chemistry lab), the color of the paint on the wall, the Dow Jones Index for that day, or the biorhythm readings for the experimenter? It is simply not possible to report everything about an experiment. One has to select, and what one selects depends upon various prejudgments. What we believe relevant enough to be reported and even what we believe to be true inevitably depend on our own values. Even more important, what we believe to be true depends on what we understand to be our beliefs. The report of scientific observations depends on self-knowledge and introspection.
Depending on introspection, however, does not mean that scientific observations are "merely" subjective. We need to be able to distinguish my observation of clock time from my belief in disarmament or my love for my wife. They are all "subjective" but differently so. How?
The biologist-philosopher Michael Polanyi begins "by rejecting the ideal of scientific detachment. In the exact sciences, this false ideal is perhaps harmless, for it is in fact disregarded there by scientists. But . . . it exercises a destructive influence in biology, psychology and sociology, and falsifies our whole outlook far beyond the domain of science" (1958, p. vii).
Science, says Polanyi (like Bronowski and Monod), rests on commitment. So, probably, does all knowledge. Polanyi speaks of "the personal participation of the knower in all acts of understanding." "Into every act of knowing there enters a passionate contribution of the person knowing what is being known, and . . . this coefficient is no mere imperfection but a vital component of his knowledge" (pp. vii-viii). He titles his book Personal Knowledge.
This personal, passionate contribution does not, however, make our understanding, scientific or otherwise, subjective. It does just the opposite. It distinguishes what is only subjective from what is truly knowledge. It is precisely the act of commitment "that saves personal knowledge from being merely subjective." When I commit myself to a belief or a fact, I do so because I responsibly submit to "the compelling claims of what in good conscience I conceive to be true." Thus, personal commitment is an act of hope and a striving to fulfill an obligation that goes beyond the personal, that indeed has a "universal intent" (p. 65). Such comprehension is not therefore simply subjective. It is "neither an arbitrary act nor a passive experience, but a responsible act claiming universal validity" (pp. vii-viii).
Consider two beliefs, one that astrology is true, the other that everybody goes through a "nomic" phase in childhood. You cannot separate the supposed truth of astrology from the personal, passionate commitment that says it is a truth. If you do not share that commitment, the belief in question will simply seem "subjective" (or, worse, delusional or faked). I believe, however, that you and I went through a nomic stage. Therefore I am puzzled when I meet people who do not share my belief, because when I say that everyone goes through a nomic stage, I also and inevitably am saying my belief that any sensible person, given the evidence I have seen, should come to the same conclusion.
Conversely, it is not possible to believe that some claim about the world is true and think that believing it is true is merely subjective. You can't believe in the truth of astrology or of a nomic phase and not be committed to that belief, because you can't be both committed and not committed at the same time--so runs Polanyi's argument. "Subjective," then, refers to an uncommitted contemplation of somebody else's "personal," that is, committed, believing knowledge (pp. 303-06). Or, as the bishop of Gloucester said, "Orthodoxy is my doxy--heterodoxy is another man's doxy." Hence we observe the politics of science and the painful truth of Planck's complaint that new ideas establish themselves not by virtue of their correctness but by the demise of their opponents.
The philosopher of science Nicholas Rescher offers yet another perspective on this relation between our minds and what we think lawful in a scientific sense. We make, he points out, a distinction between accidental and lawful generalizations. Consider these:
1. All the oak trees on my street are dying.All three are generalizations. All three are factual. That is, they generalize about the world as it is. The phrase "on my street," however, limits the first two to generalization about the world as it is. The third goes further, making a claim about the world as it must be. The first two are accidental generalizations, the third is not limited that way.
2. All the oak trees on my street are deciduous.
3. All oak trees are deciduous.
We will accept the third generalization as an explanation in ways that we will not accept the first two. Rescher suggests that the difference comes from the lawfulness of "All oak trees are deciduous," and he isolates, as the key to that lawfulness, the sense of necessity, of must-ness. A law goes beyond simply saying, "All Xs are Y." It claims "All Xs must be Y." All oak trees must be deciduous. If you ask me, "Why is that oak tree shedding its leaves?" and I answer, "Because all oak trees are deciduous," that is a "scientific" explanation. But, Rescher points out, if I answer, "Because it is on my street and all the oak trees on my street are deciduous," I sound a bit confused. "Because it is on my street and all the oak trees on my street are dying" seems even weaker. We will accept generalization 3 as an explanation, but not 1 or 2.
In distinguishing kinds of generalization, Rescher illuminates one of the difficulties experimental psychologists have in being "scientific." Psychological experiments often depend on a particular questionnaire or a particular group of people being tested. As someone has said, experimental psychology is the scientific study of the college sophomore. Because the conclusions from psychological experiments are so closely tied to the methods and conditions of the particular experiment, results look more like type 1 or 2 generalizations than type 3.
Rescher points to another property of lawfulness. A law describes not only the world as it is and as it must be, but the world as it might be. "If all Xs are Y, then if Z (which is not an X) were an X, it would be Y." "If this pine were an oak tree, then it would be deciduous." "If there were a Santa Claus and if he were in orbit round the equator, he would obey the laws of planetary motion." That is why, says Rescher, laws are laws, because they apply not only to Mars or Venus but to Santa Claus also, even though he does not exist--indeed especially since he does not exist.
We do not extend ordinary generalizations to "contrafactuals" this way. "If this pine tree were on my street, it would be deciduous." "If this pine tree were on my street, it would be dying." The difference is that we believe a law applies to situations beyond those we have observed and beyond even those we could conceivably observe, but we have no such belief about the accidental generalization.
For me, who want to judge psychoanalysis and this theory of the I as sciences, the key point Rescher makes is that lawfulness can never be simply or wholly based on observation. Laws generalize about the world as it must be and as it might be. Hence, lawfulness rests on imputation, a mental act, an act by an I.
To be sure, we want that imputation to be well-founded, and Rescher gives two key factors for well-foundedness. One is correspondence to fact. The other is the "systematic coherence" that fits the generalization into a fabric of other generalizations to build a rational, integrated body of knowledge which makes up "science."
Beyond those familiar tests, however, "laws are therefore in significant respects not discovered, but made. A law, unlike a simple assertion of regularity, involves claims . . . that are mind-dependent and cannot be rested simply upon objective matters of observed fact" (Rescher, 1970a, pp. 178-183, 195;, see also 1970b). I would add that if laws depend on mind--on belief, really--they depend on I's.
In short, the scientist is an I, and science rests on the ethics of an I, the perceptions of an I, the customs, the modes of inquiry, the procedures, the methods of interpretation, the sense of relevance, and the audience of that I, in short, on a lot of people's mental "set." Science rests on the ethical and other commitments of the I's who do science. Science even, arguments like Polanyi's and Rescher's show, relies on a belief --a faith, really--that I can extend my generalizations from the observed beyond the observable to that which I could not possibly observe. To claim that science is only objective is to befuddle, because science builds not only on disciplined observation but also on personal faith and personal imagination.
To admit that is to show how much our ideas of science have changed since the time of Freud's great discoveries, based on his own Helmholtzian faith in a world of "objective" forces and energies. Today, the "hard" sciences are still quaking (or quarking) from the advent of quantum physics and relativity. These new disciplines, notes Nobelist Percy Bridgman, brought into the subject matter of physics, "the problem of the role of the observer to which quantum theory has devoted so much attention and regards as so fundamental" (1959, p. 6). "It was not possible to formulate the laws of quantum mechanics," writes Eugene Wigner (another Nobelist), "in a fully consistent way without reference to the consciousness. . . . Even though the dividing line between the observer, whose consciousness is being affected, and the observed physical object can be shifted towards the one or the other to a considerable degree, it cannot be eliminated" (1967, p. 172).
Hence Bridgman, developing a theory of science as operations, can insist:
When I make a statement, even as cold and impersonal a statement as a proposition of Euclid, it is I that am making the statement, and the fact that it is I that am making the statement is part of the picture of the activity. In the same way, when you quote a proposition of Euclid the fact that it is you who quote it is part of the picture which is not to be discarded (1959, p. 4).He is saying that science and scientist are inseparable, much as his predecessor Werner Heisenberg did: "The laws of nature which we formulate mathe-matically in quantum theory deal no longer with the particles themselves but with our knowledge of the elementary particles" (1958, pp. 99-100). "The familiar classification of the world into subject and object, inner and outer world, body and soul, somehow no longer quite applies, and indeed leads to difficulties. In science, also, the object of research is no longer nature in itself but rather nature exposed to man's questioning, and to this extent man here also meets himself" (pp. 104-05).
Philosopher of science Karl-Otto Apel points out that before there can be data, there has to be an understanding to describe the data and know what it is. Do we include the Dow-Jones Index in our daily lab report? This preunderstanding of data rests on our understanding of human language and the forms of our life (1972, p. 19). As the novelist E. M. Forster is said to have said, "How can I tell what I think until I see what I say?"
In this vein, another distinguished scientist, Edwin Land of Polaroid fame, speaks of the "polar partnership" between the self and the world, between mind and matter. If you ask about the existence of a tree,
In many ways the tree certainly does not exist in the physical sense without the observer. The tree does not exist for radio waves of a certain wavelength [longer than the tree is tall], nor does it exist for neutrinos [to which matter is transparent]. The tree exists as part and parcel of the interaction between that part of the cosmos and our part of the cosmos, namely the "We" that has evolved over many centuries to be a partner with the tree.Similarly, if you ask about redness or a red object,
In fact there is no exterior red object with a tremendous mind linked to it by only a ray of light. The red object is a composite product of matter and a mechanism evolved in permanent association with a most elaborate interlock--so that there is no tremor in what we call the "outside world" that is not locked by a thousand chains and gossamers to inner structures that vibrate and move with it and are part of it.
Not only is there a unity between matter and being, internally, but the end product of the process is, so to speak, the process itself.
In the gradual acceptance of the hypothesis that the processes involved in exercising the polar partnership are themselves reality, I find it helpful to think of a symphony in which the opening theme asks a question and the closing theme states that the question is itself the answer (1978, pp. 25-26).
Land's image of the symphony reminds me of how close the scientist's idea of the way we humans understand our world can come to the artist's picture of that understanding. The literary critic Alfred Kazin notes, "Just as philosophers discovered the nature of thinking by realizing that we do not really 'see' anything that does not resemble the innate forms in which we think, so achievement in any art lies in the ability to recreate the 'world' into something that the mind feels totally at home with, that it ultimately welcomes as a further aspect of itself" (1980, p. 56). Similarly the classicist Norman O. Brown rephrases Alfred North Whitehead's philosophy this way: "Whitehead says the reality is unification: reality is events (not things), which are prehensive unifications; gathering diversities together in a unity; not simply here or there, but a gathering of here and there (subject and object) into a unity" (1966, p. 155; Whitehead, 1925, pp. 66-72).
Land suggests the way that the scientists' answers depend on the scientists' questions. Physicists these days often question their values and theories about reality. Newsweek quotes one well-known physicist who compares the present situation in the "hard" sciences to the game of twenty questions.
"What is so hard," argues physicist John Archibald Wheeler of the University of Texas, "is to give up thinking of nature as a machine that goes on independent of the observer. What we conceive of as reality is a few iron posts of observation with papier-maâché construction between them that is but the elaborate work of our imagination."Scientific laws rest on nature, yes, but also on human belief, human intellect, human imagination, and human ethics. Wheeler and Land show in a more concrete way than Rescher how laws are made as much as they are discovered.
Wheeler has devised an ingenious "thought experiment" to suggest how the observer himself helps determine the reality that he perceives. Imagine, he says, a game of "twenty questions" in which one player leaves the room while the others select a word he is to guess when he returns. While he is gone, the other players decide to alter the rules. They will select no word at all; instead each of them will answer "yes" or "no" as he pleases--provided he has a word in mind that fits both his own reply and all the previous replies. The outsider returns and, unsuspecting, begins asking questions. At last he makes a guess: "Is the word 'clouds'?" Yes, comes the answer, and the players explain the game.
When the questioner begins, he assumes a word already exists, just as physicists beginning an experiment think reality already exists. Yet the word comes into being through the questions raised, and the physical world emerges from the observations made. If the player asks different questions, he finds a different word, and if scientists perform different experiments they find different realities. Just as the word does not exist until it emerges from the questions asked, says Wheeler, no phenomenon is a phenomenon until it is observed. "For our picture of the world, this is the most revolutionary thing discovered," says Wheeler. "We still have not come to terms with it" (Begley, 1979, p. 62).
Some philosophers of science have come to the same conclusion that the twenty questions image for physics leads to: that how it is depends on what you ask. Traditionally, philosophers have said that the sciences deal with explanations of the causes of things by means of covering laws: if this, then that. If you drop something, then it will accelerate at a constant rate. If you frustrate someone, then they will show aggression.
The humanities deal with meanings and intentions: understanding, as opposed to explanation. I understand the logic of Rescher's argument. I understand Hamlet. Explanation being different from understanding, the humanities and the sciences are supposed to be two separate cultures.
In fact, they are not, philosopher Apel argues, particularly when the science is trying (as psychology and psychoanalysis do) to look at human beings. There, understanding is crucial, and you cannot test your understanding of a person's reasons by the methods that would work with an if-then law. Suppose you are trying to understand why a certain group of farmers left a certain stretch of land. There is no way that observing the operation of if-then, covering laws in this case and others like it will decisively answer the question. At most that kind of testing, writes Apel,
can give hints, on the assumption that these hints can be integrated into the very attempt of understanding. For it is not just observation of data . . . but communication by language with so-called objects as co-subjects which would provide the best test of one's having understood somebody's reasons for action (1972, p. 17)
It is not only the psychologist or the historian who necessarily mixes the sciences and the humanities, however. This involvement of the understanding of reasons with the explanation of events holds true for all the sciences. The reason is that the scientist's idea of an "event" or "data" has profoundly and radically changed.
The older, positivistic idea was that data preceded all understanding or explaining. The more modern, "twenty questions" view has to be that whenever a scientist calls something "data," he has already engaged in an act of understanding. He has already built on his fundamental ideas of what the world is like and how science is conducted. "Data," "event," "the way it was," "what really happened"--all these rest on preexisting assumptions of what counts as an event or data or "really."
Where did these assumptions preexist? Among the community of chemists, biologists, psychologists, or literary critics. These communal assumptions are in our minds whenever--as professionals--we frame a hypothesis, decide something is a "fact," or agree that some idea has been "confirmed." They are in our minds when we judge how much of the personal we will allow in a "scientific" procedure. In short (as we saw in part II), each of us, when we DEFT a new fact, DEFT it by means of the guidelines of our particular profession and the customs of our culture and, at an even more basic level, the capacities of our physical senses. I come back to the model of chapter 6. An I governs cultural and physiological feedbacks.
The twenty questions game would take us that far. Apel adds another chain to the argument, however, drawing on a well-known thesis of the later Wittgenstein. One person alone cannot follow a rule (because a rule requires language, and language requires someone spoken to as well as a speaker). Thus one person alone cannot practice science (despite all the "mad scientist" horror movies). "From this [it] follows that understanding and interpretation as means of communication fulfil a complementary function to describing or explaining" (1972, pp. 17, 19, 22).
That is, we can know an event by objectifying it. We can also know it by relating ourselves as human beings to the human elements in the event--or in our explanation of the event. A scientist or historian tests his "objective" explanation against the responses of his culture, other scientists or historians. These two ways of knowing, explanation or describing and interpretation or understanding, add to each other and exclude each other at the same time so as to give us a more total vision, much the way our eyes work as a pair. Each eye sees differently, but each is equally valid. To see in depth, we need both and we need their difference. Two eyes that both saw the same thing would not give us perspective. Neither would one eye alone.
So with the understanding of reasons between I's and explanations through if-then, covering laws. Apel's argument suggests that to set up understanding through identity as an alternative to explanation through if-then laws would be as false as to set up the right eye as an alternative to the left. The challenge is to see the world with both and to understand how both work together. Then, when we combine the two, we achieve a new richness. Explanation and understanding complement each other, as do description and interpretation.
In short, the scientist is an I, perceiving and DEFTing and building laws from a personal understanding and evaluation of those perceptions. As Bridgman concludes:
If one is reconciled to the inevitability of describing the world from himself as center, a unity is thereby automatically restored to the world, the unity conferred by the necessity of seeing everything from a single origin. This is not the illusory unity which we formerly thought we had, but is the only unity we can use, the only unity we need, and the only unity possible in the light of the way things are (1959, p. 248).
This quotation, like the others, is this nonscientist's appeal to a scientist's authority. I believe that these practicing scientists, unlike some nonscientists or social scientists, frankly acknowledge that science rests on the ethical and other commitments of the very human I's who do science. Because it rests on human mental processes, scientific knowledge is no simple matter of passively observing the world and drawing the inevitable logical conclusions.
Let me return, then, to the question with which I began this
section, How "scientific" is science? What do we mean by
"scientific"? Whatever science means in a positive sense, I
am content to leave to the philosophers of science. In a
negative sense, however, one thing is clear. Science is
not impersonal, objective, independent of its social
setting or of the I's who create science and are its audience.
Science is, like the I, between. There are no doubt many
possible models for science with a capital S, but I would claim
that the picture I gave in chapter 6 of an I using cultural and
physical hypotheses will serve for one. It provides an idea of
science that one can carry about in one's head, against which
to read the various claims and statements that people make, in
particular, those of humans scientifically studying humans, the
psychologists and the psychoanalysts.
14 / The State of the Arts
"Hard" scientists recognize that no matter how objective a science's aims, it rests on the beliefs, ethics, customs, imagination, and commitment of some I. Can there be a science of the I then? Can there be a science that studies the very origin of science or is that necessarily an art? In other words, can there be such a thing as a scientific psychology?
Despite the problems and paradoxes built into psychology, the staunchest statements of a nineteenth-century scientific position nowadays come from
Seymour Fisher and Roger Greenberg, for example, have written a masterful survey of experimental tests of psychoanalysis. They considered "scientific" only those observations "secured through procedures that are repeatable and involve techniques that make it possible to check on the objectivity of the reporting observer" (1977, p. 276). Theirs is the kind of remark that every experimental psychologist I know or have read makes. Their easy use of terms like "scientific" or "objectivity," however, contrasts strikingly with the recognition by a Bronowski or a Land of the interrelation between the observer and the observed.
Of course, they did not consider a report of what happened in the clinic or on the couch as allowing a check on the "objectivity" of the observer. Not unreasonably. Often "scientific" psychologists accuse therapists and clinicians of letting their wishes and fears color their case reports. "Research has . . . demonstrated," say Fisher and Greenberg, "that when an observer enters into transaction with the object of his observations, as in psychotherapy, he is likely to create the behavior he is looking for in a manner analogous to a self-fulfilling prophecy" (p. 276). Joseph Masling and Murray Schwartz would agree: "Such psychoanalytic descriptions as 'libido,' 'oral needs,' and 'castration fears' are constructs that transform their subjects in the act of defining them" (1979, pp. 264-65).
Yes, the phenomenon is well known, but it happens with experimenters as well as clinicians. Psychologists have done a great deal of research on it as the "Rosenthal" effect or (so Masling suggests) the "screw you" effect--referring perhaps to some hypothetical sophomore who gets tired of pushing a button hundreds of times and decides to do in both the experiment and his professor's hopes of promotion (Rosenthal, 1966; Masling, 1966; Friedman, 1967). The experimenter's needs, his subject's "set," the mythology of the profession, ll color the supposed "objectivity" of experiments. Most experimental psychologists hope to "control for" this interference, because they assume that it is possible to isolate the observer from the observed. The physicists, as we have seen, do not. They say that you have to rely on the experimenter's (and the subject's?) ethical commitment to truth or intellectual premises. It seems to me that that must be even more true in psychological than physical experiments.
One can create the illusion of an impersonal objectivity by defining a psychological experiment as just the part that some computer or a graduate assistant carries out, but that tactic only moves the experimenter's personal influence to another, less obvious place in the experiment. He sets it up, decides what shall be controlled for, programs the computer or the graduate assistant, and settles what constitutes success and significance.
Other psychologists recognize, as the physicists do, that the person and the experiment intertwine. It was in 1954, for example, that the distinguished psychologist Gardner Murphy wrote, "A dozen years ago, it was becoming clear that the process of perceiving was soon to be brought into relation to the entire personality of the perceiver" (p. xvii). If perception is a function of personality, then an experimental psychologist ought to ask how he or she can take into account the personalities of both the subject S and the experimenter E. Despite the passage of four decades, however, few do. Most experimentalists assume that they can reach objectivity by "controlling for" personality, averaging it out of the picture.
An easy "objectivity" may be hard to believe, but the rest of Fisher and Greenberg's criterion, "procedures that are repeatable," seems reasonable enough--until we lean on the terms a bit. What is "repeatable"? Is having Scott Fitzgerald answer a questionnaire the same as having Dr. Vincent do it? Typically, psychologists assume that one "subject" can be substituted for another in repeating an experiment or in adding up the number of times a certain question is answered a certain way. Yet doesn't that assume a great deal about human beings? Is the assumption that subjects are interchangeable "scientific"? More scientific than a clinician's careful observation of either Scott Fitzgerald or Dr. Vincent?
You cannot step into the same river twice. No events, if we are to be microscopically exact, are "repeatable." Listened to precisely enough, no two beats of a heart or ticks of a clock or throbs of an engine are the same. What constitutes repeating, then, is a question of degree and involves the experimenter's judgment as to what matters. Often, an experimental psychologist will assume that two different people pushing the same buttons counts as repeating, but the same person pushing a different arrangement of the same buttons would not. That seems to me a debatable, even "subjective," decision.
Different subjects do not read the same instructions the same way, write two psychologists about experimental "controls." Tasks "do change, even in the course of a single experimental session." "People can be found responding in different ways at different times within a single problem" (Cole & Means, 1982, pp. 134-41, 175). Subjects may be frightened, distracted, or bored. These background factors are hardly neutral, and I know of no way an experimenter can control them out of the experiment so as to make the task imposed on the subjects independent of the subjects' feelings.
Nevertheless, Fisher and Greenberg ruled out clinical reports because "there is no way to gauge which clinical reporters are telling it the way it really happened and which are wishfully seeing things in a way that supports their favorite theoretical stance." Here again, the decision may have been wise, but the phrasing states an attitude toward science more characteristic of the late nineteenth than the late twentieth century. First, there is the easy assumption that someone can tell it "the way it really happened," quite different from a Monod or a Bridgman. Then there is the contrasted position of those who are wishfully seeing things so as to support their pet theory. Yet surely most experimenters are heavily involved in shoring up their favorite theories. Otherwise why would they be experimenting? On their experiments, their grants, contracts, and promotions depend. Yet Fisher and Greenberg are seeking "objectivity" someplace other than in the ethical commitment of which Bronowski and Monod speak.
Most striking to me is their assumption that the writers of the papers they are examining are either one way or the other, either wishfully seeing or "telling it the way it really happened." Surely psychologists as sophisticated as Fisher and Greenberg want to be able to take into account the probability that both are happening, both a commitment to truth and a belief in one's theory. Their well-meant "objectivity" is what Polanyi spoke of as "the destructive influence" of the ideal of scientific detachment.
In an earlier, also excellent survey of "scientific" tests of
psychoanalysis, Paul Kline gives these three criteria:
1. Observations - which must be under controlled conditions (that is, eliminate the role of extreneous variables)
2. Constructs - which must be operational (that is, have clearly specified and identifiable empirical referents)
3. Hypotheses - which must be testable (that is, clearly disconfirmable (1972, p. 1).
Again, this all sounds reasonable enough until (as Bridgman prescribes) we press on the words a little.
The very idea of "extraneous" variables implies that Kline already knows what is and is not relevant. Moreoever, what a psychological experimenter typically excludes as extraneous are precisely the human factors, the differences between a Scott Fitzgerald and a Dr. Vincent. amagine a college age Dr. Vincent as a resentful "volunteer" in some psychology department's pool, learning nonsense words for the experiment his professor is hoping to publish. Imagine Scott Fitzgerald as an amorous Princeton sophomore learning the same nonsense words for a beautiful research assistant on whom he has a hopeless crush. How does one "control for" such human factors?
Kline asks for "identifiable empirical referents," things you can see or hear and then count. I have heard attributed to Edward L. Thorndike, one of the founders of modern experimental psychology, this remark: "If a thing exists at all, it exists in some amount. If it exists in some amount, it can be measured." Again, that sounds commonsensical enough and, to me, with my own deep wish that we could understand human beings just by their surfaces, appealing. Yet what a colossal assumption it is. Suppose we substitute for Thorndike's "a thing" honesty or love. "If honesty exists at all, it exists in some amount. If it exists in some amount, it can be measured." Suppose we were to substitute some of the terms that currently exercise philosophers of mind: intention, understanding, or empathy. The seemingly simple assumption that a mental quality can be rendered as "a thing" (or as an "identifiable empirical referent") leads to a drastic curtailing of the psychologist's subject matter to exclude the feelings and experiences that most matter to us. Or it commits us to the Lear fallacy, that unhappy king's tragic effort to quantify his daughters' love.
Kline's third requirement restates the familiar idea that, to be "scientific," a theory or hypothesis must be falsifiable by some observation of fact. Yet this business of confirmation and disconfirmation has many wrinkles, as philosophers of science like Wesley Salmon point out (1973). Suppose we were testing the statement, "All crows are black." Should we check the color of herring? Well, perhaps we should, since a silver herring fits the statement about crows being black--because it confirms the converse, "If it is not black, it is not a crow." Yet, curiously, finding feathers on a crow is not evidence for the proposition, "All birds have feathers"--because that would be reasoning from the consequent to the premises, and the logic has to go the other way: (1) All birds have feathers. (2) This crow is a bird. (3) This crow has feathers. Salmon recalls a "facetious characterization of logic texts as books that are divided into two parts: in the first part [on deduction] the fallacies are explained and in the second part [on induction or scientific method] they are committed."
Surely the scientist is right to ignore the silver of herring when considering the blackness of crows and to generalize from the feathers of some birds to the principle that all birds have feathers. Finally, what counts as a legitimate confirmation and what doesn't depend partly on logic and reality, to be sure, but also on what the community of scientists have agreed. More exactly, as Thomas Kuhn has shown, confirmation depends upon the way the confirmations to which we have committed ourselves determine what we consider logical or real (1962).
Biologists of the 1980s accept all kinds of electrical procedures that biologists of the 1880s would have greeted with skeptical laughter. When I was an engineering student, we thought we were doing pretty well if we got within 10 percent answer we were supposed to get. Psychologists have agreed that confirmation will be a 5 percent disconfirmation of the null hypothesis. That is, let's assume that a boy's grades in school have no relation to the side of the street he lives on, a null hypothesis. Then we compare the grades of the boys who live on my street. If we find that the boys on my side of the street have, by and large, higher grades than the boys on the other side and if--here the mathematics get complicated--the difference is such that there is only a 5 percent probability of its happening by chance, then (so psychologists assume) we are entitled to assume that the null hypothesis is wrong, that there is some connection between the boys' grades and the side of the street they live on. Maybe the boys on the west side study together. Maybe they have daylight longer. Maybe they absorb smartness from their smart neighbor. Who knows?--but the psychological community would agree that the absence of connection had been disproved.
Does 5 percent disprove a hypothesis? Suppose the "hypothesis" is that I have put into a drawer 95 white socks and 5 black ones. In the darkness of early morning I reach in and pull out a black sock. Should the 5 percent probability, the 19-to-1 odds against my doing that, make me doubt that I put in 95 white and 5 black socks? Even if I pulled out black socks twice in a row, odds 361-to-1 against doing so by chance (p < 0.0025), I would still be sure how many socks I had put in the drawer, and no mere statistics would shake my certainty about the number and color of the socks I own. On the other hand, I am less sure about the relation between boys' grades and where they live, and the statistical test, I agree, would make it seem possible that there is some connection. The disproof (and therefore proof) depend not just on the numbers but at least in part on my feelings about the hypotheses before and quite apart from the result of the experiment.
One basic problem for any psychologist is that his supposedly objective discoveries about mind themselves depend on mind, his own, his subjects', and his fellow experimenters'. A biophysicist, writing for a popular psychology magazine, arrives at the paradox by surveying the disciplines: physics, chemistry, biology, and psychology. First, psychologists believe that we can explain the human mind physiologically. Second, biologists believe that we can explain physiology by means of atomic physics, as so many movements of atoms of carbon, nitrogen, oxygen, or whatever. Third (as we have seen in chapter 13), physicists believe that atomic physics requires a mind as a basic part of the system (Morowitz, 1980, p. 16). More simply, it takes a mind to explain a mind. Either way, psychology--any psychology--involves a fundamental paradox.
Another basic problem for the psychological experimenter as for the biologist, but not the physicist, is that the subjects of the experiment vary considerably. Each rat's running of a maze will differ somewhat from every other rat's run. How can the psychologist deal with the variation and arrive at laws that apply uniformly? Some methods are, to be sure, taboo. One day, when I was walking through the psychology building of a large university, an agitated student came running down the hall. He scrambled from side to side of the corridor, stooping, grabbing, grabbing again, running some more. One of the gerbils in his experiment had escaped. After a few minutes, he stopped, stood up, shrugged, and walked off saying, "Oh well, that one didn't fit the curve anyway."
Perhaps we should invoke in that anonymous student's name what Nigel Calder calls "the Harvard law of animal behaviour: 'When the same stimulus is given repeatedly under carefully controlled conditions the animal will behave as it damned well pleases.'" The same thing is true of humans: "Even if one were only attempting to control the minds of a homogeneous group of psychiatric patients [sic] with a drug with which one had had considerable experience, the desired effect would not be produced in all patients, and one would not be able to plan specifically that any particular effect would be produced in a particular patient" (Calder, 1970, pp. 22, 76).
Statistics represent a valiant effort to deal with that variability. While the statistical laws themselves are pure and incontrovertible products of logic, the way they are used and the qualities they purport to measure are--inevitably, I think--problematic. One of the problems is whether smoothing off--cancelling out, really--the variability is the only or even the best strategy for dealing with it. In effect, when the student reports that nineteen gerbils traversed the maze or the professor of psychology writes that twenty sophomores (Ss) turned the knob to the right, whatever individual differences there were among the gerbils or the sophomores are gone.
Reporting the deviation in numerical terms acknowledges variation but loses the detail of it, precisely the detail that is essential to a concept like identity. If--and I realize that is a very large "if"--if identity governing lower-level feedbacks is a telling model, if it truly offers a way of talking about the way shared or "given" structures and individual choices coact to become behavior, then we need to be able to talk not only about countable behaviors but interpretable identity. Statistical methods, however, at least as we know them today, make it impossible to address the identity part of the model. Statistics can describe with great precision how a ship moves, but deliberately make it impossible to say anything about the captain who is steering her.
This is the main reason [writes one of the discoverers of the helical form of DNA, Francis Crick] pure psychology is, by the standards of hard science, rather unsuccessful. . . . The basic difficulty is that psychology attempts to treat the brain as a black box. The experimenter studies the inputs and outputs and tries from the results to deduce the structure and operation of the inside of the box. . . .It is at this point, says Crick, that the brain physiologist wants to get inside the black box (1979, pp. 221-22). The assumptions of psychology, however, make it difficult, perhaps impossible, to combine physiological information with the usual methods of psychological experimentation (disconfirming a null hypothesis, for example, or correlating behavioral inputs and outputs). The difference in methods cuts the natural connection between psychologists and brain physiologists.
The difficulty with the black-box approach is that unless the box is inherently very simple a stage is soon reached where several rival theories all explain the observed results equally well. Attempts to decide among them often prove unsuccessful because as more experiments are done more complexities are revealed.
Psychology's reliance on if-then principles and statistical methods also causes trouble when psychology meets psychoanalysis. Textbooks in psychology customarily convert psychoanalysis into a set of testable general laws or principles. I am thinking in particular of Calvin S. Hall's A Primer of Psychoanalysis, widely used in American psychology departments and a principal reason why American psychologists have such odd ideas about psychoanalysis. It is rather as though one knew Shakespeare only in Chinese translation.
The problem is that psychoanalysis, a largely holistic discipline, cannot be rendered in if-then principles. Trying to do so produces either principles that can be tested but unrealistically rigidify psychoanalysis or principles that cannot be tested (leading to smugness among antipsychoanalytic psychologists).
I think many psychologists share my uneasiness about the methods of experimental psychology, at least unconsciously, to judge from the constant claims to being "scientific" that percolate the psychological journals. Paradoxically, however, all the talk of method and disconfirmation and repeatability is not itself "scientific." Hard scientists--should I call them unsocial scientists?--don't trouble themselves about such basic assumptions. They engage in what Thomas Kuhn has called "normal science," carrying on experiments within the established rules and procedures of the scientific game of today without much questioning or concern about them. It is only when some new discovery challenges this consensus that physical and biological scientists begin the kind of self-questioning that experimental psychologists seem constantly engaged in. As Kuhn says, "It is precisely the abandonment of critical discourse that marks the transition to a science" (1970, pp. 6-7).
Thus, experimental psychology, for all its scientific claims, indeed precisely because of them, shows that it has not yet reached the status of a science. The more psychologists assert that they are being "scientific," the more they tip us off that they aren't yet. "Once a field has made that transition [to a science], critical discourse recurs only at moments of crisis when the bases of the field are in jeopardy" (idem).
That commitment to a tidy, freshman-year science is an old-fashioned virtue in one sense. In another it may be a sacred cow. Noam Chomsky, for example, sees psychology's scientism as a stultifying limitation.
The reader who undertakes the useful exercise of searching the literature will discover, I believe, not only that there is little significant scientific knowledge in this domain, but further that the behavioral sciences have commonly insisted upon certain arbitrary methodological restrictions that make it virtually impossible for scientific knowledge of a nontrivial character to be attained (Chomsky, 1972, p. ix).Niels Bohr is quoted as saying, "There are the trivial truths and the great truths. The opposite of a trivial truth is plainly false. The opposite of a great truth is also true" (Waelder, 1963, p. 18). Experimental psychology as we know it today has cut itself off from the great truths, and that is a pity.
Modern psychology has received immense support from government, industry, and an inquisitive public. Work has been widely abstracted, circulated, and followed up for five sixths of a busy century. Yet I would be hard put to state more than one or two theories of real generality that have wide acceptance in the profession (a notable exception being the theory that perception and cognition are active). The introductions to psychology textbooks, textbook introductions being an index to the basics of any field, present grandmotherly truths or ingenious experimental methods but not general scientific principles.
To a considerable degree, I feel, the "behavioral sciences" are merely mimicking the surface features of the natural sciences; much of their scientific character has been achieved by a restriction of subject matter and a concentration on rather peripheral issues. Such narrowing of focus can be justified if it leads to achievements of real intellectual significance, but in this case, I think it would be very difficult to show that the narrowing of scope has led to deep and significant results.A scientific pride goeth before a scientific fall.
In all but the most elementary cases, what a person does depends in large measure on what he knows, believes, and anticipates. A study of human behavior that is not based on at least a tentative formulation of relevant systems of knowledge and belief is predestined to triviality and irrelevance (1973, pp. xi, ix).
Even though it is unlikely that psychologists will seize upon psychoanalysis, I would like to be able to report that
"Psychoanalysts abstract qualities from the people they observe and then write about these qualities as though they had a life apart from the people in whom they were observed." ("Anality." "The" Oedipus complex. "The" superego.) "Psychologists come along and treat these statements as general laws, devising experiments" to test them. In effect, the psychoanalysts, by making "anality" or "the" oedipus complex or "the" superego into things, have created general laws (as they apply to those things) where perhaps there are none, at least where the psychologists' tests fail to show them (1979, p. 266).
Where, then, are the covering, if-then laws of psychoanalysis?
There may be none. They would have to rest on individual psychoanalyses, and perhaps one cannot generalize from these. In general, writes André Green, internationally renowned psychoanalyst, the analyst and the patient together construct a meaning which has never been created before the analytic relationship began. Therefore, the analyst and the patient are not dealing with truth in its scientific sense (1975, citing Viderman, 1970). To be sure, there are fragments of historical truth in what the patient says, but the aim of therapy is not to recover the truth but to arrive at the best mediation between the various agencies in the mind of the patient, id, ego, and superego (Loch, 1977, p. 228).
This is, however, a very different way of talking about the process than when I entered analysis in 1960. Then, most American psychoanalysts spoke of a psychoanalysis as an automatic, self-correcting process, working out the laws of psychoanalytic psychology. The patient lay on the couch associating away. The analyst from time to time interpreted. The truth or falsity of the analyst's interpretations came out as changes in the patient's free associations. An incorrect interpretation would simply be rejected. A correct interpretation might be resisted but would result in deeper, richer associations.
The content of an analysis would vary from patient to patient, to be sure, but the process of free association was self-correcting, independent of both the patient's and the analyst's personality. It made no difference to the process whether the analyst was a man or a woman, smart or stupid, warm or severe. This is very different from the idea that patient and analyst together create a meaning which depends upon the potential space unique to this patient and this analyst.
To grasp the change, we can think of the science of psychoanalysis as passing through three stages. At least, psychoanalysis has gone through three different ways of considering itself as a science.
At first, Freud set out to achieve a Helmholtzian psychology. It was to have all the rigor of physics or chemistry, and he often spoke about psychoanalysis as if he were creating an experimental science. From particular cases, one would hypothesize general laws, then test them out by particular interpretations to particular patients. In any one analysis, the analyst hypothesizes interpretations based on the general principles of psychoanalysis. If he says them to the patient (one does not, after all, communicate one's every interpretation), then the patient's subsequent free associations would show whether or not the interpretation was correct. The patient's free associations after the interpretation provide the "experiment" that proves or disproves the general law.
In 1959 two of the most distinguished American analysts were simply stating a widely held view when they insisted that psychoanalysis was just like any other science. Psychoanalytic therapy, said Jacob Arlow, "is by no means a perfect experimental tool, but it is, nevertheless, . . . governed by strict methodological considerations and operating within accepted canons of the scientific method." Heinz Hartmann maintained that the aim of psychoanalysis, besides therapy, was "to develop lawlike propositions which then, of course, transcend individual observations." To be sure, observations rely on the mental processes of the analyst, but they are "subject to the constant scrutiny of the analyst," and that is one reason the analyst has to be analyzed himself as part of his training. "Psychoanalysis has discovered potential sources of error and found a way to combat them."
A later view, also widely held, treats psychoanalysis as an observational science. The analyst engages patients in the same interpersonal way an anthropologist does when he moves his bedroll into the tepee or the kraal. He not only observes, he talks to the subject of his science. He actively synthesizes factual evidence like a Darwin comparing finches in the Galápagos. He observes the patterns of his patient's behavior just as the anthropologist observes the customs of a tribe or the biologist the behavior of an orang-utan. The interpretation is not so much an "experiment" as a making known: to the anthropologist, the analyst, the patient, the tribe.
A still later view treats psychoanalysis as a "hermeneutic," a system of interpretation. This is the current view among most analysts outside of the more conservative institutes in the United States. It makes language paramount, the patient's words or the analyst's. Rather than read through language to some inferred behavior or events, the analyst and patient take language itself as the event. The event (in the analysis at least) becomes the language it is reported in. Psychoanalytic principles then become ways of finding the meaning in that language, just as the techniques of literary criticism or biblical interpretation are ways of finding--or, better, assigning--a meaning.
That is, in interpreting one needs to go beyond the old-fashioned "conduit" metaphor of language in which somebody puts a meaning in at one end and somebody else takes it out at the other. Rather, meaning is something people actively create with their minds from a text, just as they actively perceive colors (chapter 5 above) or read symbols (chapter 4).
Most modern psychoanalysts, like Roy Schafer, would assert that the psychoanalyst is no longer committed (as Freud so deeply was) to determinism. For the modern analyst, "Determinism . . . is a way of putting questions to actions" (1976, pp. 228-29). It is a principle of inquiry, a way of insisting, Why?, and again, Why?
The assumption is that actions have reasons. Actions do not occur without meaning. "The analyst's real commitment is not to determinism in a uni-verse of mechanical causes, but to intelligibility in a universe of actions with reasons." "The essence of the psychoanalytic method is the exploration and understanding of personal paradigms," writes Louis Breger (1981, p. 48), by which he may mean something like identities.
Interestingly, Freud demonstrates all three of these modes, the "scientific," the observational, and the hermeneutic, in his well-known comparison of the method of the psychoanalyst to that of the archaeologist, the comparison with which we began this study of holistic method.
Imagine that an explorer arrives in a little-known region where his interest is aroused by an expanse of ruins, with remains of walls, fragments of columns, and tablets with half-effaced and unreadable inscriptions. He may content himself with inspecting what lies exposed to view, with questioning the inhabitants--perhaps semi-barbaric people--who live in the vicinity, about what tradition tells them of the history and meaning of these archaeological remains, and with noting down what they tell him--and he may then proceed on his journey. But he may act differently. He may have brought picks, shovels and spades with him, and he may set the inhabitants to work with these implements. Together with them he may start upon the ruins, clear away the rubbish, and, beginning from the visible remains, uncover what is buried. If his work is crowned with success, the discoveries are self-explanatory: the ruined walls are part of the ramparts of a palace or a treasure-house; the fragments of columns can be filled out into a temple; the numerous inscriptions which, by good luck, may be bilingual, reveal an alphabet and a language, and, when they have been deciphered and translated, yield undreamed-of information about the events of the remote past, to commemorate which the monuments were built (1896c, 3:192; see also 1937d, 23:259-60).
In this fascinating passage, Freud is thinking in all three of the scientific modes of psychoanalysis. There is what the philosopher would call "correspondence" truth: the archaeologist will find out exactly what the rubbish and the ruined walls and the inscriptions consist of, and whatever the stones finally "say" has to correspond to what his picks, shovels, and spades uncover. There is also the participant-observer truth of the anthropologist. Freud's archaeologist has to ask the natives. He has to dig into the ruins. He has to guess at the alphabet. Finally, there is the hermeneutic effort. He has to try to build the fragments into one coherent narrative. He has to arrive at--give them--their "meaning." The archaeologist has to make all three of these scientific modes work together for him. So does the analyst.
These three views of the "science" of psychoanalysis (determinist, observational, interpretive) correspond to three stages in the intellectual develop-ment of psychoanalysis. At first the analyst sought explanations based on the polarity unconscious-conscious, when the analyst tried to dig out what was hidden. Then analysts used explanations based on the polarity ego-nonego, when the analyst worked (like a participant-observer) with the patient's ego in a "therapeutic alliance" toward the best outcome for that ego. Nowadays analysts seek explanations based on the polarity self-nonself, with the analyst acknowledging his own interaction with the patient, as Freud's archaeologist (particularly in 1937) acknowledges his own part in what he finds and how he interprets what he finds (Holland, 1976; see also Appendix, pp. 331-33).
You could state the psychoanalytic combination of strategies another way (as the philosopher Habermas does). The psychoanalyst uses causal generalizations, if-then laws, but in an analysis, the analyst must always be applying them in a given context. Theory gives rise to a narrative of an individual (the history of his identity, if you will), and conclusions about causes always refer to this narrative (1971, p. 273).
It is particularly true of psychoanalysis (although, as we have seen in chapter 13, Polanyi says it is true of all the sciences) that what general principles there are function within boundary conditions that are in principle unpredictable. Freud himself accepted the idea that psychoanalysis would never be able to predict. One could, he noted, trace in great detail an outcome back to its origins. One cannot go the other way, predicting an outcome from the determining origins. You cannot, said Freud, know beforehand which of the determining factors will prove the stronger. A small change in the unconscious forces can end in a much greater change in final behavior, just as a change in the vote of a small faction in an electorate can completely change the policy a whole nation finally adopts (1920a, 18:167-68).
Psychoanalysis in principle does not lead to predictions? From the point of view of the experimental psychologist, this is a damaging, perhaps fatal admission. (See, for example, Kline's third criterion on p. 303.) If you can't predict, then you aren't stating propositions that can be disconfirmed. Or are you?
A single experiment can falsify a determinist scientific principle. If you were to drop a billiard ball from the Leaning Tower of Pisa, and it didn't fall with uniform acceleration, Newton's laws would be in big trouble. They would not be covering the facts they are supposed to cover. How do you falsify an interpretation, though? Not by an experiment but, as we saw with Sherlock Holmes, by new data. If the old interpretation will not deal with the new facts, then a new interpretation is needed.
Isn't this the same way an if-then law is defeated? You show that it does not cover the new data, like the billiard ball that does not accelerate uniformly. I would say the same holds true for experiments as for interpretations. If new facts do not fit the old hypothesis, then the old hypothesis is defeated because a new hypothesis is called for. So far as falsification or disconfirmation is concerned, I think experiment and interpretation do not fundamentally differ. Finally, the old hypothesis is only defeated by a better one--and better is a matter of degree, of the ethics of the scientist, and of the customs and values of the scientific community.
Indeterminacy, narrativity, relationship, the irrelevance of prediction--each looks as though it prevents psychoanalytic therapy and hence the psychoanalytic theory built on it from being "objective" in ways that would suit the experimental psychologist. That is, they do not allow the (to me) questionable assumption that it is possible to isolate the observer from the observed.
If, however, such hard sciences as optics and quantum mechanics are taking the self into account, surely it is time the human sciences free themselves from the posture of what Weston LaBarre irritatedly calls "the invisible man desperately trying not to be seen seeing other men" (1967, p. viii). Georges Devereux, in his cogent book From Anxiety to Method in the Behavioral Sciences (1967), seeks to get his fellow social scientists to "abandon--at least in a naive sense--the notion that the basic operation in behavioral science is the observation of a subject by an observer." "The behavioral scientist cannot ignore the interaction between subject and observer in the hope that, if he but pretends long enough that it does not exist, it will just quietly go away" (pp. 295, xviii). And LaBarre: "A basic datum of all social science . . . is what happens within the observer--in the large sense his own 'countertransference' reactions as a specific human being" (p. ix).
LaBarre's word "countertransference" reminds me that he and Devereux come from the clinical tradition of psychoanalysis. I would like to think, obviously, that psychologists, conscience-struck by LaBarre's and Devereux's strictures, will rush to psychoanalysis, particularly to this theory of identity. It seems unlikely, though. The American Psychological Association has 56,000 members. There are perhaps 100,000 psychologists in and out of the APA. in America alone, probably half of whom are experimentalists. Many have talent, some genius. All, so far as I can tell, are committed to a nineteenth-century model of science that rules the person of the scientist out of the game.
It does not help to hear psychologists proclaim that they are "objective and scientific" (defined in a pre-twentieth-century way) while "psychoanalysis is subjective and humanistic." Weston LaBarre expresses both the intellectual trouble with the position and the animosity that such smugness evokes: "Self-designated 'social sciences,' yearning for the prestige of exact physical sciences from the seventeenth century onward, solemnly continue to pattern themselves on a seventeenth-century mechanistic Newtonian model, quite as if Einstein and Heisenberg had not revolutionized physics in the three-century interim" (1967, pp. vii-viii).
As we have seen from the hard scientists, all science involves both the subjective person and the objective world, both if-then explanation and interpretation or understanding. Indeed all modern science is concerned at one or another level with precisely the relation between them. So is psychoanalysis. So, I think, is psychology.
It would be as much of an error to assume that psychoanalysis is only interpretation as it is to assume that psychology is only "objective" experiments. Indeed the two errors--each common--mesh to create the deadlock that we have today. The psychologist dismisses the psychoanalyst as "unscientific." The psychoanalyst dismisses the psychologist's experiments. "That has nothing to do with analysis." Each claims to be a general psychology that excludes the other. Seeing experimental psychologists and psychoanalysts at such odds is very frustrating for anyone so passionately (and ambivalently) committed to making sense out of human beings as I am.
Crucial to breaking this deadlock and crucial to the status of
psychoanalysis as a science is the problem of generalizing from
Tom, Dick, or Harry to you and me. Where do the covering,
if-then laws, which are so important to the hard sciences, fit
in psychoanalysis? Psychoanalysis always begins with a case
study of some one person in all the uniqueness and
particularity of which humans are capable. The general law, as
we have seen, reaches out by Rescher's "nomic necessity" not
only to all known cases, all possible cases, but even to
impossible cases, contrafactuals like Santa Claus in orbit.
How does one move from a unique and individual case like
Herbert Graf or Ernst Lanzer or Scott Fitzgerald to that kind of
"nomic necessity"? If Santa Claus were in orbit round the
earth, he would obey the laws of planetary motion. Would he
also have an oedipus complex?
15 / An Idea of Psychology
To the achievements of either science or psychoanalysis, the notion that 'Science is objective, psychoanalysis subjective,' is too simple a response. Rather, each involves both that "subjective" person and that "objective" world. It is not enough to ask, 'Which shall we believe?' We need to ask ourselves, 'How can we understand their coming together?' In the same way, I do not think it sophisticated enough to say simply, 'Experimental psychology is objective and scientific, psychoanalysis subjective and humanistic.' Each involves both "subjective" person and "objective" science, and the real question is, How can we use both the personal and the impersonal to understand the coming together of personal and impersonal? How can we combine, rather than oppose, the strategies of the two sciences?
So far, we have seen that experimental psychology tries deliberately to exclude the individual and that psychoanalysis tends not to address the problem of general laws. I would like to find a psychology that will interrelate the individuality that psychoanalysis studies with the general laws which the natural sciences derive and which experimental psychologists seek.
"Anybody who studies personality theory, anthropology, or history," writes D. E. Berlyne, one of the most intense of experimentalists, "is inevitably impressed with the dissimilarities between human beings and human societies that are revealed to him." He might understandably greet any attempt to generalize about all or most human beings with skepticism. "But as soon as any two human beings are compared with, say, a tree, their similarities will appear immense and their differences minor" (1971, p. 29). From this point of view, all that interests the psychologist are general laws that apply to all or just about all human beings.
I believe that psychology can do more, represented, for example, by that maxim of Kluckhohn and Murray which has served us before. All humans are in some ways like all other humans, in some ways like some other humans, and in some ways like no other humans. The idea of identity as a hierarchy of feedbacks responding to positive and negative emotions suggests ways of rendering that idea more tellingly than simply as some, some, and some.
In the two-level picture (p. 145), identity, the reference standard set at the top and working its way down through two--or many--levels, states the way in which one person is like no other person. The combination of theme and variations at any given moment which cumulates the history of that theme and its variations is unique.
In the lower loop, the feedback through reality at the physical level is fixed. Every person shares about the same physiology and the same physical reality--at least as compared with a tree. Further, every person in the same place at the same time as this person would tend to share that same loop: anyone looking through the same microscope, for example. In that sense the lower levels articulate the way in which one person is like every other person.
As for one person's being like some other people, consider this sentence, "The book is ready for the printer." At the general level of physical reality, it is the same for everybody. You see the serifs on b or the downstroke of y just about as I do. In another sense the sentence is quite unique. "Book" brings a picture to my mind of a leatherbound copy of Poems in Persons that I have been wondering where to keep. I am sure "book" brings some entirely different picture to your mind, and I can never create exactly the same picture in my mind. Nevertheless, despite the different identities at the tops of our hierarchies, we communicate through the shared physical reality of the sentence and the reality in our minds of the culture on which it rests.
It is by virtue of that inner cultural set governing our essentially similar physical perceptions that we read or hear that sentence the same way. Yet that cultural set operates in two different modes. In one, it has an almost physiological force. "Book" cannot begin with an h or a t. "The" cannot go after its nouns, "book the" or "printer the." In that sense cultural standards function in the feedback loops to say, This cannot be otherwise. If it is, you will get garbled or zero feedback. You will feel bad: anxious, unfulfilled, or frustrated. If you tried to spell "book" with hieroglyphs or cuneiform, you couldn't. Conversely, if the only writing you knew were ideograms or kana, you couldn't read "book." In that sense some of your inner cultural reality operates like a physical ability or limitation. It is like having absolute pitch or being tone deaf.
In less elementary contexts, however, your cultural set operates much more loosely. When I hear "book" I think of the word as meaning something about like what you are holding now: lots of pages, few pictures, no ads, protective covers (either hard or soft), uniform page formats, one-time publication, just the opposite of a magazine. You probably hear the word much the same way. If you and I were in the advertising business, however, "book" would be the normal word for the upcoming issue of Playboy or Cosmopolitan. On Madison Avenue, "book" can mean "magazine."
The dictionary meaning of a word does not constitute the same kind of "It can't be otherwise" cultural standard as the necessity of beginning "book" with a b. Among the same people who understand that you have to begin "book" with a b, the word can nevertheless have quite different meanings (magazine or, so my dictionaries tell me, a life sentence in prison or a record of bets or a bundle of tobacco leaves or the words of a musical comedy).
The meaning of a word, then, does not have the same limiting power as the spelling. The common answer is that the context "determines" the meaning, but that again oversimplifies. If I am talking to an advertising account executive and I hypothesize that when he says "book" he means magazine, I may be able to make sense of what he says. My hypothesis may yield satisfying feedback. I may feel good. If I am talking to an English teacher and I hypothesize that "book" means a bundle of tobacco leaves, very likely my hypothesis will yield no useful feedback and leave me feeling bad. "Context" means bringing a cultural "set" to bear. One's cultural "set" in this sense does not impose a standard on the comparator of the level below. Rather it means a ready-made hypothesis that I bring to bear on what I hear.
My cultural "set" affects my behavior in at least two quite different ways, then. One is to impose a limitation: this rule cannot be otherwise. The other merely says: this rule might be otherwise, but this is the one to try first. If we use the two-level feedback picture for an idea of my mind, my cultural values can occupy two different places in it. One is as a standard that sets limits on lower levels: being unable to read "book" as other than book. The other is as a store of ready-made hypotheses to try out on the world. "Book" probably means printed pages, bound on one side, between covers, and not published periodically--probably.
All people are in some ways like all other people, in some ways like some other people, and in some ways like no other people. When I talk about identity in the particular I ARC sense that this book develops, I am addressing the unique individual who is like no other person. When people talk about the ways an individual is like all other people, they are usually talking about the physical or physiological human who has two legs, two arms, two ears, and who speaks, laughs, or dies as every human does.
When psychologists show how some individual fits a statistical generalization, they are talking about the ways that person is like some other people. In effect, they are fitting him into a cultural group. It may be a real culture--Southerners behave this way--or it may be an ad hoc culture--the people on whom I tried this experiment behaved this way. In either case, the generalization functions in one of the two ways we read "book." Either it sets an absolute standard on lower level loops or it serves as a ready-made hypothesis that people try out. If it is a standard, then the psychologist is likely to find it among all people in the culture because "the rule could not be otherwise." If it is a hypothesis, then the psychologist will find it only among some people in the culture.
Thus the two-level feedback picture gives us a way to imagine the relation between the conclusions the psychoanalyst draws and the results the experimenter gets. The experimental psychologist tests the human world by projecting hypotheses into it. Then experiments confirm or deny the hypotheses in particulars, as the results of the experiments feed back into the psychologist's perceptions. The subjects behave the way they were supposed to, or they don't. The experimenter feels satisfied or puzzled accordingly.
The lower, the more precise the hypothesis, the closer it is likely to be to sensory reality, the more likely it is to be some process that all humans share, and the more it will lend itself to experimental methods. Experiments that test the way our eyes jump around a scene are more likely to work than experiments on the way we feel toward politicians. Conversely, the larger, the higher the hypothesis, the more likely it will be a function of the subject's identity, the less likely it is to repeat itself in an experiment, and the more an explanation for it will require the interpretations of a psychology like psychoanalysis.
The psychoanalyst is talking, one way or another, about an identity that uses the rules the psychologist describes. Those rules may apply totally, to all people in a given culture (like syntax), or they may apply flexibly and probabilistically (like meanings). Either way, however, in an aphorism of H. J. Home that deserves to be more widely known, "Mind is the meaning of behavior" (1966, p. 46).
If there is an element of interpretation that governs the if-then laws that psychologists seek, could this be a basis for Noam Chomsky's disturbing hypothesis that "some possible sciences lie beyond human grasp, perhaps the science of causation of behavior among them" (1972, p. 157)?
The feedback model can shed some light on the limitations that Chomsky points to. A psychology asks, How do my needs shape what I see and know and believe? A theory of human identity offers an explanation--at the same time that that very theory is shaped by human need and subject to the same sort of explanation as the ideas it would explain! Thus, one can trace in the theories proposed by theorists of personality the influence of their own personalities (Stolorow & Atwood, 1979).
One aspect of the question that Chomsky raises is that the making of a psychological theory is the kind of event that a psychological theory ought to be able to explain. A psychology ought to be able to account for its own coming into being. No mere if-then laws, however, can explain the framing of the if-then laws themselves. Hence a psychology cannot consist simply of if-then laws.
There is something to psychology beyond statistical correlations, then. Admittedly, the feedback model is oversimple. Nevertheless it offers a glimpse of what that something beyond is: identity or, in the broadest sense, some intention or purpose that has to be interpreted rather than reduced to an if-then, cause-effect, ground and consequence, or independent and dependent variable. "Mind is the meaning of behav- ior."
If we can model the mind as a hierarchy of feedbacks, then the psychologist faces two quite different operations. Going up that hierarchy requires an act of interpretation, but going down it one can discover if-then rules. Going down the hierarchy of mind one can discover rules of the like-all-people or like-some-people type. If, however, one wants to discover how those rules are applied or not applied, if one wants to discover why the experiment did not yield a 100 percent, uniform result, then one has to look upward, and this is true of both like-all-other-people and like-some-other-people principles. In effect one has to look upward to find what Polanyi identifies as "boundary conditions" (1968) and Herbert Simon calls "the thin interface between the natural laws within it [the artifact] and the natural laws without" (1969, p. 57).
For example, "All humans have two ears," a like-all-other principle. Once it is established, you learn about the application of such a principle even more by considering the cases where it does not apply--injuries, abnormal births, diseases--than by simply accumulating cases where it does. Another like-all-other principle: "All humans pass through an oedipus complex" (including Santa Claus). Again, having once established the principle, you learn about the oedipus complex by seeing it in action. You learn about the limits of the oedipus complex (its boundary conditions, so to speak) by looking at the cases where the generalization does not hold: autistic children, psychotics, or one-year-olds.
The same holds true for like-some-other principles. "Most human beings are between five and seven feet tall." You would learn one kind of thing by studying the distribution of heights. You would learn another kind, boundary conditions perhaps, by studying children, pygmies, Watusi, the circus's people, and professional basketball players. "Men like aggressive jokes more than women." Again, the exceptions are as interesting as the laughers who justify the rule. One could, in fact, learn relatively little about male aggression by studying only the men who like aggressive jokes compared to studying both those who do and those who don't.
To study the negatives and limits to the type of principle that regular psychologists propose, one needs to look up the hierarchy and find uniqueness and like-no-other behaviors. One has to interpret holistically, and this requires a different kind of intellectual maneuver from the experiments devoted to if-then, covering laws that psychologists conventionally rely upon. One cannot check a holistic interpretation by experiment, only by rating it against another interpretation.
One person, some people, everybody--these three levels of psychological principle claim different kinds of truth. Everybody principles are often intuitively obvious, at least the biological ones. Everybody has two legs, two arms, two ears. These are the obvious truths that show how we humans are like humans and not like frogs or trees. Principles about some people claim a different, a scientific truth: the building and scrupulous testing of carefully wrought hypotheses. The one-person principles of identity theory seem to me yet a third kind. They account for (provide narrative accounts of the inner dynamics of) what actual people do in actual situations.
In effect one could imagine the human being that the psychologists
study, homo studiandus, by means of the two-loop diagram:
As Paul Diesing shows, when working with a holistic explanation, you test a certain level of explanation, not as an if-then by itself, but in relation to the levels above and below it. You discover and improve low-level descriptions from theory. You test high-level theories by the adequacy of the low-level statements they yield. You test a whole case description such as my analysis of Shaw's identity "by its success in organizing and explaining its constituent themes." "Only a theme, at the lowest level, is tested by direct empirical criteria, such as specific . . . predictions, frequency counts, and statistical operations" (1971, pp. 229-34).
The low-level themes, the pawns in the game, came in the first place from theory. Having noticed Shaw's preoccupation with the mouth, I looked in Shaw's life for a preoccupation with skin (and noticed his woolly suits) because I knew "orality" includes stimulation of the skin.
If case descriptions, taken together, yield a theory of types of cases ("nomic" or "intrusive" character-types, for example), you can test that typology only "by the adequacy of the case descriptions it makes possible." If the typology yields a yet more general theory, such as the psychoanalytic theory of child development, still one tests the "general theory only indirectly, by testing the usefulness of its associated typology" (Diesing, 1971, pp. 229-34).
I would sum these testings up and down the levels of generality by saying that these levels are in a dialectical or feedback relation to each other. The general defines the specific and the specific the general. Interpretation of meaning and mind situates the general law, and the general law guides interpretation. If I say, "Intrusive personality traits give rise to symptoms of hysteria," that is partly a way of asking, What was this man like before the hysteria? How did he walk and move and act on the physical world? How are the symptoms from which he suffers related to breaking into a physical or interpersonal world?
In effect, the general statement guides particular inquiries by which I can refine the general statement and learn just what "phallic" or "hysteric" entails. In studying particular cases, then, it is not enough simply to generalize cases into an overall law like, "Intrusive personality traits give rise to symptoms of hysteria." One uses the generality to discover data and data to modify the generality.
"One uses. . . . " There has to be a "one"--somebody who does it. This dialectic of testing and exploration demands human participation at every step.
An older way of thinking about science assumed that the general law was the end point. You collected data, framing laws of greater and greater generality, hoping to arrive at some scientific Utopia with ;scTHE;es law of laws that covered everything. A more modern view would be that there is no end point. Science is a process that will never stop. Data enable us to propose theories which enable us to discover more data. Conversely, new technology opens up new data for us so that we can make new kinds of theories. A theory is but a way station, for at any moment people may take a new slant on theory and so change the kind of data we would search out.
In particular, so far as psychology is concerned, I think experimentalists should no longer regard the general principle or the correlation between dependent and independent variable as the end product of research. We should reverse the traditional roles. Instead of using the individual case only to make a hypothesis to be tested, we can use the test to provide a guide to inquire into the case--and then use the case to build generalizations which one can test and then apply to the case and so on around and around through a continuing feedback.
Interestingly, in imaging this process, the feedback diagrams suggest that the experimenters' correlations may be functioning in different places. If the correlations are very high, 100 percent or 90 percent, the experimenter may have demonstrated some "real" or hard-wired loop that strictly determines the feedback a given behavioral output will have. If the correlations are low, the experimenter may simply have demonstrated a preference for a certain behavioral output or hypothesis. In short, I am suggesting that we use this feedback picture of an identity controlling perceptions to replace the merely numerical concern with p;ml.05 or p;ml.01 or p;ml.001. Instead, we can try to understand where in the feedback process the generalization formulated by the psychologist fits and functions.
I am suggesting that psychologists seek significance not in abstract, arbitrary statistical criteria but in the kind of actual payoff one gets by applying the hypothesis. What counts is not methodology but what we can learn. Does the experimental result enable me to enlarge my inquiry by setting up a dialectic between the principle and the case? Fundamentally, we are really trying to understand some event and understand it more and more as we study it more and more. Whatever method enables us to do that is the more scientific.
The nineteenth century imagined a science of more and more comprehensive laws culminating in some universal law that applied to everything, at which point the scientists could all close down their laboratories. Today, one might define the aim of science more simply: the understanding of what is not yet understood. In that sense, a covering law can be useful either to deduce facts or to ask questions. One could treat a generalization like "Intrusive personalities are likely to have hysterical symptoms" as the occasion for testing and measuring the likelihood of hysterical symptoms. One could also use it simply to inquire into the next case to be understood. Are the symptoms hysterical? Is this person intrusive? What is the relation between the symptom and the total personality? This kind of science seeks a continuing relationship between case and law rather than finding a law and stopping.
For example (I am particularly indebted to Theodore Mills for this demonstration), a sociological study of the Nixon tapes combined General Inquirer content analyses, measurement of the I/We quotient, verb/adjective quotients, and other experimental or statistical techniques to enrich an already sensitive case study of one group's disintegration. One could use a countable increase in "I" and decrease in "we" to demonstrate that the White House "plumbers' group" was losing its cohesion. That insight, however, enabled Mills's team to turn back to the tapes and look for other, less regularizable signs, metaphors of things coming apart, for example, or changes in the figures of speech applied to the group leader, Nixon. These insights in turn led back to other sociological tests, for instance, leadership indicators. The numbers could show a positive or negative or nominal change, but the precise size of the numbers proved not as useful as the insights they gave rise to in the particular case. In general terms, the "experimental" or "survey" techniques and the interpretive ones stood in a dialectical relation to each other, each adding to the other (1977).
That is, Mills and his group used them dialectically, in an exploratory feedback of the kind that diagrams of the I illustrate. In general, in twentieth-century science an active scientist has replaced the neutral observer of the nineteenth century. As Stephen Toulmin points out, such a conception of science is more, not less, scientific than that traditional one, for "a physical world described in terms of order and information . . . is a world within which human perception, communication, and interaction can find the natural place they lacked, both in Descartes' world of matter and machines, and in Helmholtz's world of energy." Such a science "will be capable of embracing not only the world of nature, as human beings experience and interpret it, but also the activities and experience by which humans perceive, think about, and deal with that world" (1978, pp. 332-33). Psychology--any psychology--ought to be that kind of science, not one that tries to blink away the human element, since a psychology is trying to explain just that, the human element.
The most stringent criteria for science have come from realist philosophers like Karl Popper. They have insisted that the only knowledge worthy of the name scientific or indeed of the name knowledge consists of propositions that have successfully resisted disconfirmation (1972). Popper asks for two things: first, disconfirmable propositions, and second, processes of disconfirmation. "The method of science is the method of bold conjectures and ingenious and severe attempts to refute them." He is describing, in effect, the feedback that I believe defines an I.
To apply Popper's test to identity theory, one must recognize that identity theory does not rest on the proposition, "Identity is a theme and variations," but on the claim, "I can represent so-and-so's identity as a theme and variations." Then there are clear procedures for judging that claim or disconfirming it (as by a more inclusive or more direct interpretation). So understood, a psychology based on I ARC DEFTly meets Popper's criterion.
A difficulty, however, would come about if Popper were to insist that these processes of disconfirmation be independent of the disconfirmer. If one insists on that dissociation, then I cannot reconcile what I am suggesting as a science of I with such a notion of realism or objectivity, nor do I think such a notion truly realistic or objective. It leaves out the human element so demonstrably there. If, however, one accepts the role of the human being in disconfirming, then it seems to me that the kind of psychology I am suggesting in this chapter fits both our strongest criteria for what is scientific and also the realities of human experimentation, a science in the full sense of one that includes the activities by which humans perceive, think about, and deal with a real world.
We can put that another way. If we could not talk about people rigorously, then to maintain rigor in a scientific inquiry we would try to keep the personal element out. Suppose, however, we can talk about individuals rigorously (and I am claiming that any psychology must try, and identity theory enables us to do just that). Then we lose no rigor by including the personal influence. In fact, we would lose rigor by leaving it out. An experimenter who claims objectivity because a computer or a graduate assistant can carry out the experiment only obscures the true state of affairs.
We come back to the basic paradox that limits what we can hope for from psychology. Any psychology that tries to describe human thoughts and actions must explain humans doing psychology. A theory of human identity governing if-then laws offers an explanation and a way of describing and interpreting that I, even if it is an I doing psychology. But that very theory is shaped by my human needs and subject to the same sort of explanation as the ideas it would explain. This is both the limitation and the achievement of identity theory. It does not turn a blind eye to the role of the person in science, and the price it pays for that inclusion is the illusion of objectivity and the hope of finality.
"Objectivity" in a theory built on identity means accepting that knowledge is knowledge by persons. At some point we have to give up hopes of finding absolute laws separate from persons and accept a continuing dialogue among different persons formulating laws. You could read that result as the deconstruction of science. I prefer, having my needs, to see it as a way toward unity and a kind of certainty, this book's
I am proposing a model for the way we act on the world. We push trials out from ourselves into the world, and we take in the results in what one kind of engineer would call an information-processing feedback loop and another a generate and test sequence. These feedback loops embody the if-thens and the correlations that regular psychologists discover which are the principles by which our bodies and our cultural values function.
These principles do not run themselves. Something above and beyond them supplies the standards they seek, and I call that something identity. It initiates the loops. It is what results when, in the history of the individual, the world responds to his trials. It is somebody's representation of that individual as a theme and variations.
In I ARC and DEFT, I am suggesting a picture that we can carry around in our heads with which to think about experiments and psychologies and readings and interpretations and the relation between mind and body. Identity (an agency, a consequence, and a representation as theme and variations) governs the if-thens and correlations of our bodies and our cultural values as they function in feedback loops when we feed actions into the world and it feeds responses back to us. Ultimately what decides what we will accept as truth is how that feedback feels: clear or muddied, congenial or dissatisfying, culturally coherent or bizarre.
I DEFTly ARC provides a very general theory that is in itself neither true nor false but a general organizing idea that tells us what to look for. I would like to think "the I" is that kind of large idea of which Bohr said, its opposite is also true. Given such a picture, we can think with some cogency about the way a person has a certain style.
By means of the I and ARC and DEFT we can understand better why a Scott Fitzgerald writes the way he does or a Charles Vincent changes and unchanges under brainwashing or the way an Anna S. renews but continues herself by being psychoanalyzed. We can understand how identity governs such familiar mental processes as using symbols or perceiving. They answer to a picture in which identity is the agency that sets the standards for symbolizing or perceiving in lower-level feedback loops.
That picture in turn lets us write the classic psychoanalytic story of the way children become adult types into a dialogue. Culture and family pose the infant certain general questions (five, in my thinking) which the infant hears and answers in its own individual body language. We can trace this kind of a dialogue all through a life, from a four-year-old Little Hans up to a sixty-eight-year-old Herbert Graf. In this framework we can even speak of Santa Claus's oedipus complex. Nomic necessity applies to the loops and questions, while the language in which they are heard and answered is unique.
Given such a picture, we can begin to grasp some of the paradoxes that I broached in the preface. My I seems realer than anything else I know, yet it is completely intangible. It is so because it is the reference level for the systems by which I sense. I can no more touch or sense my identity than my house can set its own thermostats. The I is what I sense with. It cannot be sensed itself, yet I experience my I directly and immediately. Then, the minute I think about it, it becomes other, not-me. Why? Because thinking of an identity, especially one's own, cannot include the identity doing the thinking. I can only approximate my I by phrasing it, but never completely and never "correctly," because it was first and foremost a bodily identity, a thing of touch and smell and taste, too deep for words. It came into its ghostly being before I knew words.
The I ARC DEFTly picture of an I offers a thread through humanistic, deconstructive paradoxes like these. It also offers the hope of a psychology which is scientific in a twentieth-century rather than a nineteenth-century sense. I intend two things by that distinction between the centuries. First, in twentieth-century science, facts don't come into our senses naked and innocent from the world around us. They come from hypotheses. The familiar phrases of the psychologists use certain shibboleths--"bias," "distortion," "control the vari- ables," "get something to count," "replicability," "reliability," "validity"--in order to criticize psychoanalysis for not providing raw facts. Yet these shibboleths proceed from hypotheses about what counts as a fact just as much as the psychoanalyst's "orality" or "Oedipus complex" does.
Second, because facts come from hypotheses, facts ultimately depend on people because hypotheses do. Science is not a set of abstract laws, devoid of human intervention. People give rise to science, and science carries its ancestry in its principles.
Hence it is, in this century, no longer possible to draw a neat line between subjective and objective, between situations the person is in and situations the person is out of, between science and humanities. Certainly, numbers do not make such a shibboleth, nor can we say simply that science seeks out "covering laws" while the humanities deal with uniqueness and the individual case.
One simply cannot divide the way we know things into scientific and interpretive, personal and impersonal, objective and subjective. This is the idea on which The I is built, and on it the book must stand or fall.
Rather, this feedback picture suggests, science involves both numbers and ideas, both covering laws and the individual case. Science consists precisely of the dialectic or feedback between them. Numbers improve the idea, and the idea improves the numbers. The covering law helps us better understand the individual case, and our better understanding of the case lets us improve the covering law. Both the humanities and the sciences share this pattern, differing not in kind but in the degree to which each emphasizes covering laws. Both ask us to use our skills both to generalize laws and to interpret persons.
As you must have surmised by now, I have a deep and permeating need to see a unity and consistency in things. Since my education first took me along a technological route, through electrical engineering, and only later to the study of literature, two of the things that I seek to unify are the sciences and the humanities or, I could say, certainty and beauty, fidelity and desire.
Science requires a community of scientists to establish the rules of the game. Hence, science and the humanities are not two opposed cultures. Rather each fulfills and complements the other. Measured data about the world (the traditional material of science) presuppose communication between people (the traditional material of the humanities).
I see psychoanalysis, particularly if it combines identity theory with modern feedback theories of the brain, perception, cognition, and memory, as a science that gives reasons for such a synthesis, making it possible. Such a psychoanalysis means we can talk about "the I" with precision. Hence we can unite the psychological and the physical sciences and both with the humanities. "It is only the modern formulation of the mind-body problem which prevents us from unifying psychology and physics in an all-encompassing world-picture" (Ricoeur, 1978, p. 337).
Instead of "two cultures," then, I see a continuum of intellectual activities. Some stress individuality, others generality. Some accent the feedback we get from reality, some the hypotheses we bring to reality, but all take place within a relationship of feedback between a theme-and-variations self and the world that begins at the ends of our fingertips. All presuppose a theory of the I. It is impossible to make a statement in any science or any of the humanities without making some assumptions about human beings.
This theory of the I codifies a set of such assumptions that I like to think works better than most. Identity theory lets us represent an I with precision, as a theme and variations. Further, we can add to that theme-and-variations representation the concept of identity as feedback system. Identity-as-agent sets the standard for a feedback through our cultural resources, our bodies, and our environment, and the feedback itself sustains and creates identity-as-consequence. The I uses eyes and ears, memory, ego (in a psychoanalytic sense), and gender (in a feminist sense) to create and re-create itself. Such an ARCed I fits not only with psychoanalysis but with late twentieth century thinking about perception, memory, cognition, and the architecture of the brain. It is a theory of the I that any humanist or scientist can use.
We need not, as some distinguished modern philosophers and
psychoanalysts have claimed, dispense with the idea of "mind."
Rather we can think about mind quite precisely, and we can use
that concept to explain behaviors. Mind
is the meaning of behavior. With identity we can find that
meaning in a child's triumphant growing or in the achievements of
a Darwin, a Shakespeare, or a Freud. We need not choose
between sciences and humanities. At the heart of the study of
life, no matter what we call it, there is--am--I.
Back to the cover page?