Introduction by PaoloMigone

In the past months, some discussion lists of psychiatry, psychology and psychotherapy have been heated by lively debate on the scientific status of psychotherapy. Several colleagues were suggesting that a so-called idiographic approach, rather than a nomothetic one, was more suited for the psychotherapeutic enterprise. According to the idiographic approach, they say, the object of study is unique [idios], not amenable to the generalizing laws [nomos] of the nomothetic approach, which is typical of natural sciences. In order to contribute to this debate, in the Psychotherapy section of POL.it here we publish a classic paper on this topic, written by Robert Holt in 1962, which shows how this problem was faced by an exponent of an important line of research in psychology. In fact, the topic is not new, and it is striking that it continues to raise lively debates without an easy solution; actually this is just what makes it interesting. This debate was already active in the United States in the '40s around the issues of personology (the study of personality) and Holt was a first hand witness. At that time Holt was a pupil of Gordon Allport, the famous personality theorist and strong believer in the idiographic approach. In contrast to his teacher, Holt took an opposite position and believed that the idiographic approach had to be simply abandoned and used only for artistic, non scientific purposes. His logical arguments are set out in this paper, which appeared for the first time in the Journal of Personality, 1962, 30, 3: 405-422, and soon became a classic, being republished several times. The Italian translation appeared in the Bollettino di Psicologia Applicata,1963, 57/58: 3-24. The version published here is not the 1962 version, but a new edition published in 1978 when it appeared the two volumes by HoltMethods in Clinical Psychology: Assessment, Prediction and Research(New York: Plenum, 1978); furthermore, the author has added new notes and a few changes, so that the version published here can be considered a new, final edition of this paper, reflecting the more recent development of the author's thinking. Two forewords are published here: the first one has been written in 1998 purposely for this POL.it edition, and the second one, with minor changes, appeared in the aforementioned 1978 edition. We thank the Journal of Personality for the permission to reproduce this paper in the English version, and the Organizzazioni Speciali (O.S.) of Florence for the Italian edition. I translated both forewords and the new parts of the article, of which I also improved the overall translation. This article, here published in both English and Italian languages, appears as the sixth document of the Psychotherapy section of POL.it.

Finally, few words about Robert Holt. He is already know to Italian readers because many of his works have been translated. In Italy he is known mostly for his writings in theoretical research in psychoanalysis, which occupied a later phase of his professional life, when he entered the prestigious research group led by David Rapaport. There, with the scholarly precision which is typical of him, and also under the mentorship of Rapaport (who was one of the most important theorist in psychoanalysis), he studied psychoanalytic metapsychology. After Rapaport's death, when he was viewed by some as his successor as the leader of the group, he came to take amore critical stance towards Freudian metapsychology. Among other things, he made a lot of work in psychological testing (for instance, he edited the second edition of the classic Diagnostic Psychological Testing by D. Rapaport, M. Gill & R. Schafer of 1945-46 [New York: Int. Univ. Press, 1968]). A collection of his writing, from his pioneering critical essays to metapsychology of the '60s to his most recent works, was published in 1989 with the title Freud Reappraised: A Fresh Look at PsychoanalyticTheory (New York: Guilford, 1989).

Foreword by Robert Holt to this edition (1998)

In a book that made a great impression on me, Gerald Holton (1973) noted that "there have coexisted in science, in almost every period since Thales and Pythagoras, sets of two or more antithetical systems or attitudes, for example, one reductionist and the other holistic", adding:

In addition, there has always existed another set of antitheses or polarities, even though, to be sure, one or the other was at a given time more prominent -- namely, between the Galilean (or, more properly, Archimedean) attempt at precision and measurement that purged public, "objective" science of those qualitative elements that interfere with reaching reasonable" objective" agreement among fellow investigators, and, on the other hand, the intuitions, glimpses, daydreams, and a priori commitments that make up half the world of science in the form of a personal, private, "subjective" activity. Science has always been propelled and buffeted by such contrary or antithetical forces (Holton G., Thematic origins of scientific thought: Kepler to Einstein. Cambridge, MA: Harvard University Press, 1973, p. 375).

The objective-subjective dichotomy goes beyond the boundaries of science itself, however. I feel that my own interest in the holistic and subjective approach to psychology has two principal roots: my personal psychoanalysis, and what might be called my aesthetic interests, which go back as far as I can remember. Some of my most vivid early memories preserve experiences of wonder and ecstasy occasioned by a dewy spring morning, by the sight and scent of wild flowers, even by an Art Deco painting in a magazine advertisement. My first publications were poems (in literary magazines at prep school and at college). Some years before, the first reflection I can recall about a possible career was a wavering between being an artist (of what kind, I don't recall) and an astronomer. The latter became real enough for a period in my adolescence when I helped form an astronomy club and worked for months grinding a mirror for a telescope. Then I discovered and was enthralled by other sciences: chemistry, paleontology, biology, and finally psychology. My undergraduate teacher and thesis sponsor, Hadley Cantril, showed me the intellectual excitement of social psychology and helped me discover the endless fascination of psychological research.

Clinical psychology made it possible for me to retain and integrate most of these interests and values. Along the way I was gratified by finding teachers and role models who embodied much of the artist as well as the scientist: Gordon W. Allport, Henry A. Murray, Robert W. White, David Rapaport, and Gardner Murphy. These men seemed easily able, in their own lives, to transcend the split between what C.P. Snow called the "two cultures" of art and the humanities, and of science. I have of course experienced the conflict of which Snow wrote, and have found myself pulled in opposite directions by the effort to ride two such differently-minded horses simultaneously. Insight into the underlying unities has come slowly, though helped by identification with mentors.

A century or more ago, the underlying conflict in outlooks took the form of a contrast between nomothetic and idiographic approaches to personality. Not surprisingly, it continues to be rediscovered and revived by generation after generation of psychologists and kindred workers. Though I wrote the first draft of the accompanying paper almost 40 years ago, it seems tome still worthy of being reprinted if it can help some contemporary scholars, clinicians, or researchers avoid a blind alley of methodological confusion.

Foreword by Robert Holt to the 1978 edition

The paper here reprinted with a new subtitle, was written during the year 1960-61 (and appeared in the Journal of Personality, 1962,30, 3 [September]: 405-422), though I had been brooding about many of the issues it considers for two decades. In a real sense it was a kind of starting point for all my work in clinical and personality psychology; thus, I reprinted it as the first chapter of my book, Methods in Clinical Psychology(Holt, 1978).

Appropriately enough, it appeared cheek by jowl in the Journal of Personality with a paper by AlIport (1962) on the same general topic. It was to study with him that Hadley Cantril sent me to do my graduate studies at Harvard, and though Allport fended off my hero-worshipping efforts to become his close disciple by a kind of pained and embarrassed withdrawal, he was one of my main teachers. Sitting in his seminar on personality theory I became increasingly skeptical about his basic approach, and by the time I had completed my dissertation--with his considerable help--I had on my drawing board the sketch of an attack on his methodological position. Fortunately, the pressure of making a living forced me to put off writing it, for I did not yet have the historical perspective on the issues that only the teaching of personality theory would help me gain. During the respite o fa delightful year in the congenial setting of the Center for Advanced Study in the Behavioral Sciences, I was able to take a fresh look at the problems and write the following pages.

In his accompanying paper, Allport restated a vision he followed with admirable tenacity all his professional life, of a psychology of personality that should devote itself to understanding how the structure of the unique person comes about. In his latter years, after becoming acquainted with general systems theory, he took over some of its terminology and tried to adopt its outlook (Allport, 1960; 1961, Chapter 23). Here (Allport,1962) he called for a morphogenetic approach to the system that a personality constitutes, as opposed to the traditional dimensional tack of analytic reductionism. Yet he failed to see that the systems outlook amounted to a Kuhnian revolution, a fundamental transformation of formative principles in terms of which many of psychology's traditional antinomies became reconciled. It vindicated his faith in the molar and structural approach to personalities, his stubborn insistence that there could and should be a science of individuals, but not his ambivalent polemical rejection of atomism. Part of the beauty and power of the systems approach is that it finds an appropriate place or both molar and molecular observations and laws, for both analytic and synoptic methods, as Weiss (1969) argues so cogently with respect to the life sciences.

As a graduate student, I turned away from Allport partly because his methodology led to so little by way of useful method. Twenty years later, he still had come up with few techniques by which to carry out his program of morphogenetic study of individual persons. He persistently favored matching, despite the fact that research using this method had led nowhere (or at least, to very little), concluding lamely: 'Although the method gives us no insight into causal relationships it is, so far as it goes, a good example of a 100 per cent morphogenic procedure" (1962, p. 416). In the same paper, he goes on to recommend several more useful procedures which actually boil down to two: Baldwin's (1942) personal structure analysis, a simple statistical method of identifying reliably recurring co-occurrences of content themes in a person's productions; and interviewing in order to discover important themes, problems, structural foci, stylistic features, values, or other traits in a person, sometimes followed up by the use of generalized rating scales like those of Kilpatrick and Cantril (1960).

Partly, I believe, Allport's difficulty was his personal lack of clinical training and experience, a circumstance that kept him always at a distance from the personalities he wished to study. In his entire career, he never made an intensive study of a single personality at first hand! The apparent exception (Allport, 1965) was a study of a collection of Letters fromJ enny, carried out after its subject's death; Allport had never met her.

Because he could never see the usefulness of dimensions as more than weak compromises with true individuality, Allport only fitfully addressed himself to the task of trying to find useful dimensions. One obvious difficulty in the realm of personality is the abstractness or non materiality of our subject matter, as compared to that of a biologist. The latter has no difficulty with using a concept like mitochondrion or chromosome, because the things in question are visually recognizable, having a recurrent distinctiveness of form despite their manifestly unique and fluctuating configurations in the individual cell. What is a corresponding element, individually variable, that goes to make up the unique personality? To answer, `the trait' (as Allport generally did) is simply to substitute another generalized term for "element." By and large, the traditional approach has been to fall back on the resources of nonscientific language, as Allport did himself (in collaboration with Odbert, 1936), accepting such a descriptive adjective as punctual or dominant as the functional equivalent of an organelle. Yet such terms are inevitably interactive: they are impressions made on an observer by a person, judgments rather than perceptions, in which the orientation of the judge demonstrably plays a very large role. Some times it is more important than what is in the person under scrutiny, especially when we get into the very extensive areas of personal behavior that are socially valued either positively or negatively-- the halo problem.

Somehow, we need to find a way of taking a fresh look at people, without the blinders of standard trait vocabulary. I believe that a case can be made that concepts like self (a person's reflexive experience and conception of his own personality), wish, fear, value, ability, and temperament are the analogues of the cell's fine structure. The psychology of personality, then, should seek for general laws or generalizations about regular structural relations among these elements, which may hold regardless of the unique content of such terms when applied to individuals. (It seems much less fruitful to look for generalizations on the level of these actual contents--e.g., the targets of people's values, which are given in such large part by culture, the appropriate level on which to seek regularities in the value realm.) Likewise, I am doubtful that it will be helpful to proceed just by making case studies, even though that may be an indispensable ground on which to begin work. We can pursue general goals even while ostensibly bending our efforts to the understanding of unique lives. That at least is the spirit in which I have tried to work, both as a diagnostic tester and as a researcher.

One of the hardiest perennial weeds in psychology's conceptual garden is the notion that there are nomothetic (generalizing) and idiographic (individualizing) branches, types, or emphases of science. Many respected and important contributors to psychology-especially to personology, the psychology of personality-have quoted these terms with respect and have used them as if they contributed something useful to methodology (e.g., Allport, 1937a; Beck, 1953; Bellak, 1956; Bertalanffy, 1951; Colby, 1958; Dymond, 1953; Falk, 1956; Hoffman, 1960; Sarbin, 1944; Stephenson, 1953; the list could be considerably extended). It is the purpose of this essay to examine the historical origins of this cumbersome pair of concepts, their logical implications, the reasons psychologists espouse them, and alternative solutions to the underlying problems. In so doing, I hope no doubt fondly, but none the less ardently to lay this Teutonic ghost which haunts and confounds much of modern psychology.

The principal exponent of the nomothetic-idiographic dichotomy in this country has been Gordon W. Allport (1937a, 1940, 1942, 1946, 1955), a pioneer in academic personology and a man who has brilliantly clarified many important issues in the field. On this particular point, I shall try to show, the artist in him has probably dimmed the vision of the scientist. The underlying problem with which Allport wrestles is vexing enough: the unusual nature of personality as a scientific subject matter. Allport readily concedes that everything in nature is unique, but maintains that natural sciences are not interested in the unique leaf, stone, or river. Only personology, the argument continues, takes as its very subject matter the unique personality as opposed to the generalized human mind or the behavior of organisms at large. The rest of psychology takes care of the general laws of behavior and experience and is thus nomothetic (literally, setting down laws) [Footnote 1]; what is left over is the impressive fact that every personality is different and must be studied in such ways as respect and try to capture this uniqueness in short, by an idiographic science (literally, portraying what is private or peculiar, i.e., individual). With these two curious words adopted from Windelband, then, Allport describes what he sees as two complementary branches of psychology, both of which are necessary for complete coverage.

Footnote 1: This is the generally accepted meaning. Brunswik (l943), however, used it in a different sense, which occasionally causes confusion: as pertaining to a science of exact laws expressible as functions or equations, and opposed to statistical generalizations. Both are within the scope of the nomothetic, as understood here. Rickert used a slightly different term, nomological.

On the other hand, many distinguished contributors to personology, from Freud to Murphy (1947), have found no need for such an approach to the scientific study of individuality, and the sharp voice of Eysenck (1954)has been heard rebutting Beck (1953) and proclaiming that psychology should be nomothetic throughout. Clearly, the issue is controversial.

Historical background: the romantic movement in science [Footnote2]

Footnote 2: In preparing this historical summary, I have relied principally on Roback (1927), Allport (1937a, 1937h), Boring (1929), Parsons (1937), L. Stein (1924), Tapper (1925), Friess (1929), Klüver (1929), and the Encyclopedia of the Social Sciences. I am aware of some oversimplification in speaking about the Romantic movement in science; a variety of figures and currents of thought that could be characterized as romantic may be distinguished in the history of nineteenth-century science, some of them only loosely related to the movement described here.

Kant, writing in the middle and late 1700s and reacting against reductionism, is one of the intellectual ancestors of this issue (which can, of course, be traced back to Plato and Aristotle�like any other problem in psychology; see Popper, 1957). Though he did not himself fall into the dualistic belief that mind and matter were so different that different methods had to be applied to their study, he wrote about these issues on too sophisticated a level for his followers. Thus, the analytic and generalizing methods of natural science were fine for the study of matter, but the mind, according to the post-Kantians, had to be studied also by an additional method, intuition of the whole. Being impressed with the concrete uniqueness and individuality of personality, they did not want to analyze it but to grasp it by a direct empathic act [Footnote 3].

Footnote 3: For clarification of Kant's role in these matters, I am indebted to my friend Abraham Kaplan.

Yet for the next century, no one developed such an intuitive approach to personality into anything; meanwhile, physics and chemistry, and even some branches of biology, grew rapidly and used the developing scientific methods with great success in the realm of matter. Mechanics developed early, and Newton's laws of motion were misunderstood as being foundation stones of mechanism and materialism. As C. Singer (1959) points out, Newton's laws were quite abstract and did not deal with physical bodies at all; but in their early great successes they were applied to the motions of the planets, and thus were thought of as the laws of material masses. It could hardly have been otherwise, because of the prevailing tenor of philosophical and scientific thought. The world was simply not ready for the field-theoretical implications of Newton's theories. Even so great a physicist as Lord Kelvin found "meager and unsatisfactory" any physical knowledge that could not be expressed in a mechanical model.

Though the facts of their own disciplines did not require it, then, natural scientists�helped along by the overgeneralizations of contemporary philosophers�adopted a hard-headed, materialistic, and mechanistic positivism. It was assumed that all reality was orderly, classifiable, and susceptible of mechanistic explanation; to the extent that it seemed not to be, the province of science ended. It was expected that the secrets of life itself would shortly be reduced to physico-chemical formulas. The resulting clash with religion and humanism seemed an inevitable consequence of being a good scientist:

What was not realized was that the success of science was due to the faithfulness of its practice, while its destructiveness[of humanistic, cultural values] arose from the error of its philosophy which saw that practice as though it were the outcome of a world-view with which it was in fact fundamentally incompatible (C. Singer, 1959, p.420).

This was a classic atmosphere, ripe for the romantic revolt that started in poetry at the turn of the nineteenth century and swept through the arts. The humanities are accustomed to see the pendulum swing from classicism to romanticism and back again; from a time of reason, order, control, and clarity to one of passion, ambiguity, free expression, and revolt. To a degree, such movements are felt in the sciences as well, though usually less clearly. In science, we have a temperamental difference between the tough-minded and the tender-minded, as James put it, or in Boring's phrase, the advocates of nothing but against those of something more; in the nineteenth century, it was objectivism and positivism versus subjectivism and intuitionism. The hard-headed positivists had had their way for a long time; near the end of the century, however, there was something of a romantic revolt in science, tipping the balance toward the subjectivists. Independently, in two different parts of Germany, Wilhelm Dilthey in Berlin and the "southwesterners" Windelband and Rickert proclaimed the primacy of understanding (Verstehen) in certain kinds of science over quantification of elements, in the part of the general intellectual current with which we shall be concerned here.

They elaborated the distinction between two kinds of science: theNaturwissenschaften, natural sciences, and Geisteswissenschaften, the German translation of J. S. Mill's "moral sciences." The latter term, often retranslated as "social sciences," meant actually a good deal more, for it included philosophy and the humanities as well as history, jurisprudence, and much else that is often excluded from social science today. In an attempt to develop separate methodologies for the Naturwissenschaftenand Geisteswissenschaften, Windelband and Rickert took up, developed, and popularized a distinction between two types of science that had been proposed by Cournot, the French founder of mathematical economics[Footnote 4]. Cournot, who was also something of a philosopher of science, had a sophisticated concept of chance and examined the role it played in various fields of knowledge in the process of classifying them. In the exact sciences, precise laws were possible, he said, but in history, chance played such a large role that only a probabilistic discipline was possible. As a later philosopher of history, Meyer, put it: any particular event "depends on chance and on the free will of which science knows nothing but with which history dealt" (Quoted by Weber, 1949,p. 115, from Meyer's Zur Theorie und Methodik der Geschichte, Halle,1900.)

Footnote 4: See the article on Geisteswissenschaften in the Encyclopedia of the Social Sciences and Cournot (1851). The reader who is interested in a richly detailed picture of the issues and their background will do well to read Popper(1957) and Chapters 13 and 17 in Parsons (1937). An excellent briefer account is given by Klüver (1929), in which references to the principal relevant works of Windelband and Rickert may be found.

It should be clear by now that we are dealing with not just a pair of isolated terms but a complex set of methodological concepts and viewpoints. The nomothetic-idiographic distinction can no more be understood out of the context of the geisteswissenschaftliche movement than can any isolated culture trait torn from its cultural embeddedness. For the sake of convenience, I shall refer to this complex of ideas as the romantic movement in science. There have been so many major and subtle shifts in our outlook that it is difficult for us to see the issues with the eyes of ca. 1900; recall, however, that vitalism was a live doctrine then, and the ideas of chance and free will [Footnote 5] were closely connected, respectable concepts. Many scholars conceived of history as having been shaped primarily by the acts of great men; as we shall see, the theme of the relation between personality and achievement is a recurrent preoccupation of the romantics.

Footnote 5: At the time I wrote this paper, I shared the prevailing and largely thoughtless rejection of free will, accepting as self-evident the proposition that it was opposed to the scientifically necessary assumption of determinism. I have since mended my ways; see my papers of 1967a and 1972a; and in my book of 1989,see pp. 246-252. Similar arguments may be found in Russell (1929), Chein (1972), M.B. Smith (1974), Weiss (1969), and Rubinstein (1997, pp. 440ff.).

It is factually true that history, biography, and literary criticism are primarily interested in increasing our understanding of particular events, persons, or works, rather than in treating these as incidental to the discovery of general laws. But men like Windelband and Rickert took the jump from this proposition to the sweeping declaration that all of the disciplines concerned with man and his works should not and by their very nature cannot generalize, but must devote themselves to the understanding of each particular, and its integration "as a real causal factor into a real, hence concrete context" (Weber, 1949, p. 135). The repetition of the word "real" in this passage underscores the conception that only the concrete was real, hence abstractions could not be conceived of as causes of particular events. Moreover, abstract analysis of specific events or persons was thought to be fallacious, since it destroyed the unique unity that was the essence of any such particular. This essence was qualitative, not quantitative, and often consisted of verbal meanings (as opposed to objective facts, the subject matter of natural science), which could not be measured but only interpreted. By identifying Cournot's methodological distinction with their own between the knowledge of Being (Sein),obtained in physical science, and the consciousness of and relatedness to norms (Sollen) in the cultural sciences, Windelband and Rickert started the great debate on the role of values in science [See Chapter 10 of myMethods in Clinical Psychology (1978), Vol. 2, "The Problem of Values in Science"].

For psychology, Dilthey was the most important figure in this movement. He was a philosopher, an admirer of Goethe and Schopenhauer, rebelling against Christianity and Hegel, though influenced by the Biblical hermeneutics of Schleiermacher. He wanted to respect the heart's reasons the head will never know, to understand life in its own terms, not to explain it. The anti-intellectual element in such a goal is perceptible, and indeed, he is part of the current in German thought that provided the philosophical background for Nazism. He wanted, not a reduction of data either to physical-material or to idealistic terms, but a direct insight into the vital nature of things as articulated wholes. His approach was empirical, but in a different sense from the atomistic English tradition, stressing the importance and primacy of the unbroken whole, theStrukturzusammenhang. Obviously, he helped prepare the seedbed for Gestalt psychology. He was optimistic, unlike some of his successors (e.g., Spengler), and very influential in Germany.

Basic to the development of the social and cultural sciences, he thought, was the development of a new psychology, which he called verstehendepsychology�a descriptive discipline concerned with the systematic knowledge of the nature of consciousness and of the inner unity of the individual life, and with the understanding of its development. It did not analyze or start with elements, but with experienced relationships. The most important unifying forces in a man were purpose and moral character. He saw the intimate relation of the person to his social setting and insisted that individual human character was an outgrowth of institutions, not vice versa.

These are only fragments from Dilthey's large output of ideas, which lacked system and order; his work was brought together only after his death, by friends. Nevertheless, it stimulated many workers in diverse fields: jurisprudence, economics, sociology, philosophy, genetics, history, and psychology.

Dilthey's most important psychological follower was Spranger, who is known chiefly for his book Lebensformen (Spranger, 1922; translated asTypes of Men). He too distinguished sharply between explanatory and descriptive psychology, favoring the latter, verstehende, type. Verstehen,he says, is the mental activity "that grasps events as fraught with meaning in relation to a totality." He was opposed to the analysis of personality into elements, but wanted to stay on the level of "intelligible wholes." As a focus for the study of individuality, he followed Dilthey again in proposing that the person's values, which determined the direction of his strivings, be considered of primary interest.

Dilthey had propounded three forms of Weltanschauung which underlie and pervade the personalities as well as the doctrines of the philosophers whom he studied. Spranger proposed his famous six ideal types of values, to which actual individual values more or less correspond: the theoretical, economic, aesthetic, social, political, and religious. He did not recognize the possible cultural determination of his choosing just these six, but traced them back to instincts. Each value type has its own ethics (e.g., economic: utilitarianism; aesthetic: harmony), and its own style of life in many other ways. The entire scheme was rather ingeniously worked out.

This theory followed the new ideas in stressing the unity of personality, the way in which many details of behavior become comprehensible when we know such key facts about the total structure as the principal values toward which a man is oriented. To underline the contrast between the prevailing atomistic psychology and his own, Spranger called it Strukturpsychology. As a general theory of personality, it suffers from incompleteness, and its main influence today comes from its having stimulated the production of a widely used paper and pencil test, the Allport-Vernon-Lindzey Study of Values, which is still in active use as a research instrument.

The history of the psychology of Struktur and Verstehen since Spranger is not yet finished. Its influence is still felt in personology, and as a school it still has adherents in Germany. Allport has done the most to bring it to this country; there were a number of lesser figures, but they have not made significant contributions.

William Stern, a man of some influence in psychology, must be at least briefly mentioned even though he began in intelligence testing and his work converged only rather late with the main line of development traced above. The nomothetic-idiographic distinction played no part in his writings, though he was influenced by verstehende psychology. He had been a pioneer and an established figure in child psychology and the psychology of individual differences, when he became convinced that conventional psychology was wrongly conceived. As differential psychologists, he said, we are studying isolated mental functions, the ranges and correlates of their variations, but overlooking the important fact that all such functions are embedded in personal lives. As child psychologists, we talk about the growth of intelligence or the like, forgetting that only persons grow. Reasoning thus, and basing his psychology on his personalistic philosophy, he decided that a radical rebeginning was imperative; psychology had to be rebuilt with the indivisible, individual person as the focus of every psychological investigation. Even Gestalt psychology with its emphasis on totalities and its similar antielementarism was insufficient, for: "Keine Gestalt ohne Gestalter." Stern went into most of psychology's classical problems, such as perception, making the point that there are not separate problems of spatial perception in hearing, vision, touch, etc. there is only one space,personal space, and it is perceived by whatever means is appropriate. Most of the facts that had been established in traditional general psychology were brought in, with this new twist.

Stern's theory of motivation was a complex one, including drives (directional tendencies), instincts (instrumental dispositions), needs, urges, will, pheno-motives and geno-motives, etc., in too subtle and highly elaborated a structure to be recounted here. He did not have a theory of personality as such; rather, the personalistic viewpoint pervaded all of his general psychology. There was a specific theory of character, however, conceived of as the person's total make-up considered from the standpoint of his acts of will, his conscious, purposive striving. Though stratified, character is a unified structure and may be described by a list of traits, but this is only the beginning; much stress was laid on the particular, concrete structure. Particular traits, said Stern, no matter how precisely described, have meaning only when you see what function they play in the structure of the whole personality.

These are the principal psychological figures in the stream of ideas that produced the distinction between nomothetic and idiographicWissenschaften and then applied the latter approach to the problems of psychology. Perhaps the name of Jaspers, in psychopathology, should be added. He helps to establish the continuity between the romantic movement at the turn of the century and the contemporary existentialist-phenomenological movement in psychiatry. The geisteswissenschaftlichepoint of view made even more headway in the social sciences, from which some influence still comes to bear on psychology. Popper (1957) has applied the term historicism to one of the main streams in sociology, history, and economics that developed as part of the romantic reaction against positivist, natural-scientific methodology. Such potent names as Marx, Engels, Spencer, Bergson, Mannheim, and Toynbee are among the historicists, and the movement is by no means dead today, despite the vigor of attacks by logical positivists which have refuted the underlying logic of this position. I will not further consider this important group of theorists, who have been adequately routed (Popper, 1957; cf. also Popper, 1950).

How useful were the new ideas to the group of psychologists discussed above? What they took from the romantic revolt was its emphasis on the permission to study as legitimate objects of inquiry, personality, values, motivation, and the interrelation of such factors with cognition (e.g., ideology, perception). Starting with Dilthey's first disciples and going on through the solid contributions of Spranger and Stern, these men did not adhere to a strict distinction between idiographic and nomothetic approaches, and were disinclined to make any substantial change in their accustomed ways of scientific work. Any follower who wholly gave up general concepts and stuck closely to intuitive contemplation of indivisible Gestalten simply dropped out of the picture. The men who are remembered used the new battle cries to help shift their fields of activity slightly, and to develop new types of concepts, which as concepts were on no different level of abstractness from the ones Dilthey and the southwesterners attacked so vehemently.

Note, for example in the above summaries, the generalizing, abstract nature of the motivational concepts used by Spranger and Stern: both retained values and instincts, which were assumed to be found in all persons.

As soon as they stopped their polemics and got down to work, the men of this romantic revolt strayed off the intuitive reservation and came up with conceptual tools methodologically indistinguishable from those of so-called nomothetic science. In a way, Stern was the most consistent in the attempt of his General Psychology from a Personalistic Standpoint (1938) to reshape all of psychology from bottom to top; but on closer examination, the changes turn out to be largely verbal. It is all very well to talk about personal space, for example, but no idiographically personalistic research methods were developed. One could hardly say that there has been any further development of a personalistic psychology of perception, except in the sense that Stern has helped focus attention on new types of generalized variables derived from a study of individual differences in perceptual behavior (see Klein & Schlesinger, 1949).

Nevertheless, all of this work did represent an important ground swell in the history of ideas, and it had some useful influence on the behavioral sciences. It did not make them idiographic, but it directed their attention to new or neglected problems and novel kinds of variables, as well as to the issue of structure: the ways the variables are organized. Like many rebellions, it revolted against a tradition that was stultifying, only to produce an opposite extreme, which if taken literally would have been equally useless or more so. Fortunately, scientists only occasionally take their concepts quite that literally and with such logical consistency. Especially at a time that old ideas are overthrown, the important content of the new movement is often emotional. Through the drama of overstatement, a prevailing but opposite overemphasis may be overthrown, and in calmer times other men may find a sensible position from which to move forward [Footnote 6].

Footnote 6: As stated here, the text seems to imply that the solution is a compromise, whereas I am now convinced that nothing less than a change of ethos or age, in Ackoff's (1974) phrase, is involved. See footnote 18, below.

Certainly the psychology and social science that held the stage in Germany during the 1880s and 1890s were in many ways inadequate as scientific approaches to important human problems. It was a day when not only value judgments but even an interest in the psychology of values was banned from scientific concern. Fechner and Wundt had started with problems it is easy to dismiss as trivial, minute, or far removed from what the man on the street thinks of as psychology. Experimental psychology had to start that way, and it can now look back on an illustrious, slow development of methods and concepts, which today permit laboratory studies of personality and some of life's more pressing issues. But a century ago, is it any wonder that a person who was interested in man the striver, the sufferer, the spinner of ideologies as Dilthey was�thought that the classical scientific approach itself might be at fault? Surely the world of inner knowledge, of passions and ideals, had been left out, and theverstehende movement was a revolt against this one-sidedness.

The historical role of differential psychology

In psychology, the romantic movement has been felt particularly in personology, the psychology of personality. And one reason that its impact was particularly great there is the fact that personology grew out of differential psychology, the psychology of individual differences.

The first efforts of the "new psychology" of the 1890s were devoted to finding empirical generalizations and abstract laws about such functions as sensation and perception (concepts which themselves were the heritage of faculty psychology). It was what Boring has called the science of the average, healthy, adult (and, one might add, male) mind, a subtly Aristotelian conception that generally relegated the study of women and children, and of abnormal and exceptional behavior, to a subordinate status. Even so, there remained embarrassing observations of exceptions to the general laws even when the subjects were "average, healthy adults." Accordingly, the field of differential psychology was invented as a kind of wastebasket to take care of these annoying anomalies. From the standpoint of the highest type of psychology, which was concerned with laws in a way not expected of differential psychology, the unexplained residual variance continued to be considered error and to be treated as if it were random and unlawful.

The psychologists who were content to work with the miscellany of leavings from all the high-caste tables in psychology were further handicapped by the taint of practical application, for they were principally involved in applying psychology to mundane problems like educating children, treating the disturbed, and selecting employees. Such work called for the prediction of behavior, and it quickly became apparent that the general laws provided by "scientific psychology" left a great deal unpredicted; it was practically imperative to supplement them by some kind of lore that dealt with all the other important determinants.

As time went on, differential psychologists made a radical shift in approach. In the era when individual differences were thought of as error as not lawful, really they were catalogued and measured, and a few attempts were made to parcel out the variance in terms of sex, age, ethnic, group, and other gross demographic categories. During the past couple of decades, however, personologists have increasingly begun to recognize that all the error-terms of standard psychological equations are their own happy hunting grounds. Individual differences in such hallowed perceptual phenomena as time-error, size-estimation, and shape-constancy proved to be not random at all but reliably related to other dimensions of individual differences in cognitive phenomena and in noncognitive realms, too (see Gardner, Holzman, Klein, Linton, & Spence, 1959).

The fallacy involved in treating individual differences as if they were random and unlawful resembles that of the eighteenth-century scientists who concretized Newton's laws as propositions concerning mechanical bodies. In both cases, the grasp of certain principles lagged behind what could have been expected. Objectively viewed, the laws that govern individual variation in the perception of apparent movement are just as abstract as the laws that cover the general case, and seem to have a different methodological status only because of the accident of history that brought about the discovery of the latter first. And, despite the implied promise in Klein and Schlesinger's title (1949), the study of such general principles does not bring the perceiver, the person in Stern's sense, back into perceptual psychology. It is merely a change in the axis of generalization, so to speak, not a way of becoming less abstract about perception.

The logic of the romantic point of view in personology

Let us now consider each of the main propositions that make up the romantic point of view, and state the logical objections to them systematically. In brief, they are an indictment of natural sciences as having no room for a meaningful approach to personality; the argument is thus plausible only to the extent that science has approximated the nomothetic caricature.

1. The Goal of Personology Is Understanding, While That of Nomothetic Science Is Prediction and Control [Footnote7]

Footnote 7: This section on understanding was rewritten (in 1966) to take into account some of the ideas of Polanyi (1958), to whom I am indebted for clarifying these subtle matters; I did not have an adequate grasp of them at the time this paper was first written. The would-be tough-minded misconception of science, which denigrates intuition, empathy, and understanding, is endemic in psychology: most of us have been trained to a kind of automatic obeisance to the ideal of objectivity to the point where we lose sight of its proper sphere of relevance. Therefore, I have little hope that this brief section will convince many who were trained in the tradition of behaviorism and logical positivism; I urge them to study Polanyi (1958), who writes from an inside and expert knowledge of several of the "hard sciences." Kaplan (1964, especially Chapter IX) is another good source on the limitations of prediction as an ultimate criterion in science. His summary is worth quoting: "explanations provide understanding, but we can predict without being able to understand, and we can understand without necessarily being able to predict. It remains true that if we can predict successfully on the basis of a certain explanation we have good reason, and perhaps the best sort of reason, for accepting the explanation" (p. 350). See also Holt (1978), Volume 2, Chapter10.

Here is a particularly subtle and mischievous dichotomy, which has been accepted by all too many, even within psychology. Actually, all of the highly developed sciences aim at prediction and control throughunderstanding; the goal is compound and indivisible. Most scientists, as contrasted with technologists, are themselves more motivated by the need to figure things out, to develop good theories and workable models that make nature intelligible, and less concerned with the ultimate payoff, the applied benefits of prediction and control that understanding makes possible. But it is only in recent years that the dual nature of the scientific enterprise has been made clear by philosophers of science, the fact that it has hypothesis-forming and hypothesis-testing phases; moreover, methodologists have concentrated almost entirely on explicating a reconstructed logic of proof and have neglected to describe the very part of scientific work that is most exciting and personally rewarding. Therefore, it was easy enough to portray "hard science" as uninterested in understanding, and solely guided by the aim of verifying or refuting predictions that were rigorously deduced from theory.

This misconception of what actually goes on in such sciences as chemistry and physics was enthusiastically taken up by many psychologists; indeed, it is central to behaviorism. One should not underestimate the emotional appeal of such an ideal of a completely rigorous, objective, machine-like science, especially for such an intrinsically difficult and ambiguity-ridden field as psychology. It is a close analogue of the obsessive-compulsive ideal of completely rational thought and action; the behaviorist and the obsessional alike hoped to escape from the frightening entanglements of emotional subjectivity by banishing it entirely. Affects can indeed be a source of error. But Freud saw clearly as long ago as 1900 that effective, adaptive thought must reduce the scope of affect not to zero but to that minimum necessary for the signal functions involved in judgment. When emotions are completely ruled out, the neurotic has no sense of what course of possible action is promising and what is not. His defenses protect him from hunches and intuitions, leave him without a sense of what data are to be trusted and which ones are probably artifacts, and allow him no way of knowing what is an important problem rather than a trivial one. Clearly, this kind of "freedom from subjectivity" is crippling for a scientist.

Smaller numbers of psychologists, concentrated largely among clinical practitioners, resisted the successive vogues of behaviorism and logical positivism and embraced the notion of an idiographic discipline. They were happy to have philosophical auspices under which to reject scientific method and control, so they could be free to indulge in irresponsible speculation and undisciplined intuition. To distinguish their ideal from the understanding of natural scientists, let us call it Verstehen. The true scientist's goal is an explanation-a cognitive grasp of the significant inner structure and functioning of a phenomenon, an intelligible model arrived at by disciplined analytic and synthetic processes. In Verstehen, however, the emphasis was not on figuring out how something really worked but on gaining an empathic feeling of it as directly as possible, a sense of knowing it from the inside by non intellective means. As its exponents described it, Verstehen was not so much Kekule's insight into the structure of organic molecules in his famous fantasy of the carbon atoms joining hands in a ring, as it was the feeling of understanding a person more deeply after seeing a portrait of him painted by a master. This latter type of nonexplanatory understanding, therefore, is a subjective effect properly aimed at by artists, not scientists. A vivid, compelling portrait, whether in paint, stone, or words, achieves its communicative effect by judicious selection and artful distortion, not by complete fidelity to reality. Indeed, a scrupulous realism that tries to copy nature exactly is the death of an artistic endeavor, though it is necessary to science.

If personology were to be devoted to word portraits that seek to evoke in the reader the thrill of recognition, the gratifying (if perhaps illusory) feeling of understanding unique individuals, it would become not an idiographic science but an art. There is such an art of personality, of course: literary biography. We can enjoy and profit from it, while recognizing that an artist's quest for "truth" differs from a scientist's in being a striving not for strict verisimilitude but for allusive illumination. The criterion of this kind of understanding is the effect on some audience; the ultimate criterion of scientific understanding may be verified prediction, or-depending on the particular science-an elegant and comprehensive accounting of facts already available, like the Darwinian theory of evolution.

In all of these matters, it is difficult to tread a middle path, and the dichotomous thinking engendered by the contrast of idiographic and nomothetic makes it even more so. To the neobehaviorist, the avowed search for understanding, the interest in major (which means messy and uncontrollable) human problems, the respect for introspective data and subjective phenomena, the use of empathy as a method-all of that makes personology look indistinguishable from armchair speculation and self-deluded mysticism. To the idiographer, the determination to use as much scientific method as possible (including statistics and experimental design), to control unwanted sources of variance, to test hypotheses rigorously instead of simply proclaiming them-all that makes personology look indistinguishable from the alleged sterilities of psychophysics and the irrelevancy of rat learning. Yet the heritage of the older, more developed sciences themselves is exactly this middle way.

The difference between the scientist and the artist is at bottom not so great as dichotomous thinking would make us believe. Gardner Murphy points out (in a personal communication) that there are many cases, "all the way from Leonardo da Vinci to John J. Audubon and John Muir, of scientist-artists, in whom it is not conceptually very feasible to make the roles distinct." Some personologists too (notably Freud, Murphy, Allport, and Murray) have had much of the artist in them as well as the scientist and have been masters of prose writing; small wonder, therefore, that at times the artistic side of their identities has come uppermost. If Allport had been less aesthetically sensitive, he might not have failed to distinguish between artistic and scientific goals. Often, too, poor scientists are at the same time poor writers, and an inferior case study may be poor either because its facts are wrong and its interpretations undiscerning, or because it is poorly put together and lacks the literary touch that can put the breath of life into even a routine case report. The more art a scientist possesses-so long as he does not let it run away with him-the more effective a scientist he can be, because he must use his aesthetic sense in constructing theory as well as in communicating his findings and ideas to others.

2. The proper methods of personology are intuition and empathy, which have no place in natural science

As has been indicated above, intuition and empathy were used by the romantics as ways of gaining direct and definitive understanding, and were considered to be complete scientific methods. The contemporary personologist has no quarrel with their use in the practical arts of clinical psychology and psychoanalysis, nor as ways of making discoveries and formulating hypotheses. Indeed, the more secure scientists are in their methodological position, the more respect they usually have for intuition (and in psychology for the closely related methods of empathy andrecipathy [Footnote 8]). Thus, the claim that these operations have no place in natural science is false; they are used by all scientists in the most exciting and creative phase of scientific work: when they decide what to study, what variables to control, what empirical strategies to use, and when they make discoveries within the structure of empirical data. As to their sufficiency, I need only remind the reader that the methodology of verification, the hypothesis-testing phase of scientific work, involves well-developed rules and consensually established procedures, and that intuition and empathy have no place in it.

Footnote 8: Recipathy is the method of "becoming as open and sensitive as possible" not only to "the subject's movements and words" but to one's own feelings of "how the subject's attitude is affecting him [the observer].. . if he feels that he is being swayed to do something, he imagines Dominance; if he feels anxious or irritated he infers Aggression, and so forth"(Murray, 1938, p. 248).

3. Personology is a subjective discipline as contrasted to objective branches of psychology, being concerned with values and meanings, which cannot be subjected to quantification

Elsewhere (Holt, 1961; 1978, Volume 2, Chapter 3]), 1 have dealt with the contention that there is a fundamental methodological difference between disciplines that deal with verbal meanings and values, and those that deal with objective facts. Briefly, the argument is the familiar one that objectivity of method is intersubjectivity, and that meanings (including values) may be perceived and dealt with in essentially the same ways as the data of natural science, which must be discriminated and recognized also. Moreover, a logical analysis of the operations carried out in disciplines such as literature, concerned with the understanding of individual works and little (if at all) with generalization, shows that these workers outside of science use many of the same methods of analyzing texts as the quantitative content-analysts of social psychology, with their exclusive concern with generalization. Their work has shown that meanings may be quantified and in other ways treated as objectively as any other facts of nature. Other objections to quantification grow out of antipathy to abstract variables of analysis and will be considered in the following section.

4. The concepts of personology must be individualized, not generalized as are the concepts of natural science

The belief that the concern of personology with unique individuals (see below) contrasts fundamentally with the exclusive concern of nomothetic science with generalities logically implies that the two types of discipline must have different types of concepts. As the chief spokesman for the romantic point of view in psychology, Allport calls for the use of individual traits, which are specific to the person being studied, not common traits, which are assumed to be present to some degree in all persons. But to describe an individual trait, we have to take one of two courses: either we create a unique word (a neologism) for each unique trait, or we use a unique configuration of existing words. The first approach is clearly impossible for communication, let alone science; personology would be a complete Babel. The second solution, however, turns out to be a concealed form of nomothesis, for what is a unique configuration of existing words but a "fallacious attempt to capture something ineffably individual by a complex net of general concepts"? Allport himself has explicitly ruled out this possibility:

...each psychologist tends to think of individuals as combinations of whatever abstractions he favors for psychological analysis. This procedure, common as it is, is wholly unsuitable for the psychology of personality. For one thing, such abstract units are not distinctively personal (Allport, 1937a, p. 239). [Footnote 9]

Footnote 9: Allport wrote these words in the context of rejecting Murray's system of needs (1937a); yet elsewhere (Allport, 1937b) he praises as "strikingly personal" such concepts (or dimensions) of W. Stern (1938) as depth-surface, embeddedness-salience, nearness-remoteness, and expectancy-retrospect!

An idiographic discipline thus must be a dumb or an incomprehensible one, for intelligible words even some of Allport's favorite, literary ones, like Falstaffian, which he does consider "personal" abstract and generalize, proclaiming a general pattern of resemblance between at least two unique individuals, Falstaff and the case being described. Any such trait thus becomes common, not individual.

One of the great methodologists of social science, Max Weber (1949) developed an apposite analysis of scientific concepts and their development in reaction against the romantic movement in his country at the turn of the century (see Parsons, 1937). He had the insight to see that the exponents of Geisteswissenschaft were trying to do the impossible: to capture the full richness of reality. There are three identifiable stages in the scientific study of anything, Weber said. To begin with, one selects from nature the historical individual (or class thereof) one wishes to focus on; for example, the Boston Massacre, the personality of Einstein, the cathedral at Chartres. Even though limited, each of these is infinitely rich in potentially specifiable aspects and configurations. One could study one of these, or even a tiny "flower in a crannied wall," until doomsday and not exhaust everything that could be known about it. Without doing any more abstracting than focusing on a particular topic, one can only contemplate it; and this is where the idiographic approach logically must stop. The method of intuition or Verstehen is essentially a wordless act of identification with the object, or some other attempt to "live in it" without analyzing its Gestalt.

The second stage, that of the ideal type, is a rudimentary attempt to see similarities between historical individuals, while staying as close as possible to their concrete particularity. Ideal types are much used in psychology, especially in diagnosis, for any syndrome such as schizophrenia is a complex of identifiably separate but loosely covarying elements, never encountered in exact textbook form. The lure of ideal types is that they give the brief illusion of getting you close to the individual while still allowing a degree of generality. But this advantage is illusory, the apparent advantage of a compromise that denies satisfaction to either party. Concrete reality (fidelity to the unique individual) is forsworn, and the advantages of truly general concepts are not attained. An ideal type does not fit any particular case exactly, and the failure of fit is different in kind as well as degree from one case to another. For an ideal type

is a conceptual construct which is neither historical reality nor even the 'true' reality. It is even less fitted to serve as a schema under which a real situation or action is to be subsumed as one instance. It has the significance of a purely ideal limiting concept with which the real situation or action is compared and surveyed for the explication of certain of its significant components" (Weber,1949, p.93).

Weber's final stage of scientific development, therefore, is the fractionation of ideal types into their constituent dimensions and elements, which he called abstract analytical variables [Footnote10]. Paradoxically, only a truly abstract concept can give an exact fit to any particular individual! I cannot say exactly how Falstaffian or how schizophrenic or how big any particular subject may be, but I can name a particular value of an abstract analytical variable, height, that fits him as closely as his skin. The example would be less convincing if chosen from psychology because we do not have as well-established, unitary dimensions as the physical ones, and not as simple and unarguable operations for measuring them as the use of the meter stick. The principle, however, is the same.

Footnote 10: In his analytical emphasis and his neglect of total system properties--which most personality traits seem to be--Weber showed his rootedness in nineteenth-century thinking.

The fit is exact, of course, only because an abstract analytical concept does not purport to do more than one thing. If I try to measure the breadth of a person's interests, I make no pretensions to have "captured the essence of his personality." Not having tried, I cannot properly be accused of failing. But I have chosen a variable that can be measured, and thus potentially its relations to other aspects of personality can be discovered and precisely stated. Curiously, Allport attacks general variables on the ground that they "merely approximate the unique cleavages which close scrutiny shows are characteristic of each separate personality" (Allport, 1946; his emphasis). His preferred ad hoc approach may seem less approximate because many of the general variables used in personology are ideal types, lacking true abstract generality. The solution, however, lies in a direction diametrically opposed to the one toward which Allport beckons. And it does not consist in escaping from approximation. Scientific models of reality can never fit perfectly; the attempt to force such identity between concept and referent sacrifices the flexibility and power of abstract concepts in a chimerical quest for the direct grasp of noumena [Footnote 11]. Parenthetically, the recent vogue of existentialism and Zen Buddhism in psychology may be partly attributed to the promise they extend of providing a way of grasping the total richness of reality. Part of the lure of satori or any other mystical ecstasy of a direct contact with the world, unmediated by concepts, may stem from the necessary distance imposed by the scientific necessity to abstract. But despite their confusing jargons, which make them seem superficially quite different from the late nineteenth-century romantic movement we have been considering, both of these fashionable doctrines suffer from the same fallacies. Mystical experience offers nothing to the scientist qua scientist except an interesting phenomenon that may be subjected to scientific study.

Footnote 11: A few decades later, I feel less inclined to reject typological concepts or to consider them only way stations to attaining the analytical goal of a set of abstractions. Following Weiss (1969), I would say now that some types are attempts to delineate recurrent patterns of system organization, which are not wholly reducible to their components. At the least, the issues seem much more complicated today than they did when I wrote this paper, and I have since discovered that typological concepts are widely used in other, "harder" sciences (Holt, 1998).

5. The only kind of analysis allowable in personology is structural, not abstract, while natural science is not concerned with structure

It is true that the scientific psychology of Dilthey's heyday had no place for structural analysis in the sense introduced by the romantics. Psychology dealt with a number of functions, which were treated implicitly or explicitly as quite independent of one another. It had no methods parallel to those of exegetic Biblical scholarship (hermeneutics) or literary criticism, which seek out the internal organization of ideas in a specific text. And the reductionistic enthusiasts for analyzing things were not interested in putting the pieces back together again, nor very clear themselves that analysis need not mean dismemberment. This state of affairs made it easy to think that analysis could be destructive, and that structural relations between the parts of the personality could be studied only in concrete, unique individuals, so that structure [Footnote12]seemed to be an exclusive concern of idiographic disciplines.

Footnote 12: Ironically, in psychology the early adherents of structuralism were among those who carried atomistic, reductionistic analysis to its most absurd extreme: the Titchenerian introspectionists. The Gestalt psychologists, though appalled by the equally atomistic behaviorism and structuralism alike, concentrated their efforts on perceptual patterning, leaving untouched most of the structural problems that concern personology, particularly the enduring invariances of molar behavior. For a recent, searching critique of structuralism see Thelen & Smith, 1994.

There are really two points here: the distrust of analysis, and the emphasis on structure. The first of these has been partly dealt with in the preceding section; it was based on a misunderstanding of the nature of abstract concepts [Footnote 13].

Footnote l3: But also on a valid recognition that analysis, even supplemented by a synthetic effort to put the pieces back together, is not enough: analysis need not kill, but it does fail to say all that needs to be said. Systems must be described on their own level, not just that of their constituent elements. The point has been beautifully developed by Weiss (1969). For another clear exposition of general system theory's resolution of the false antithesis between holistic and analytic approaches, see Koestler (1967).

On the second point, structural concepts and structural analyses are commonplace in science at large today. Such structural disciplines as stereochemistry and circuit design were (at best) in their infancy at the time of the idiographic manifestoes. Today, natural science uses abstract, structural, and dispositional concepts simultaneously with a minimum of confusion. Presumably, the same may be true of personology someday, too.

One merit of the romantic tradition in personology is that it has consistently highlighted the problem of structure. At the time Allport was taking his position on these matters (in the late 1920s and early 1930s), the predominant American conceptions of personality were "and-summative" (it was defined as the sum total of a person's habits, traits, etc.), and the problem of structure was ignored. The early academic personologists who concentrated their efforts on personality inventories, single variables, or factor analyses, all tended to disregard entirely the structuring of these elements or to assume simple, universal answers (e.g., orthogonal factor-structure).

At the same time, however, Freud (1923b) was developing the structural point of view in psychoanalysis, and today psychoanalytic psychology is increasingly concerned with the problem and has developed a variety of variables to deal with it (Rapaport & Gill, 1959; Holt, 1960b; and see the work of G. S. Klein and his associates on cognitive controls as structural variables: Gardner et al., 1959 [but see also Holt, l975b, 1998]). Drawing on this tradition and that of psychopathology generally, psychodiagnosis concerns itself with structural variables and their constellation into a limited number of ideal types (e.g., the obsessive-compulsive type of ego-structure) which, in the best practice, are used not as pigeonholes but as reference-points in terms of which the clinician creates individualized analyses of personality structure.

6. There can be no general laws of personality because of the role of chance and free will in human affairs

There are hardly any contemporary personologists who openly espouse this argument. It played an important part in the development of the romantic point of view, as we have seen, and persists in Catholic psychology. Itis generally admitted, however, that scientific work requires the basic assumption of strict determinism throughout [Footnote 14]. Closely examined, chance becomes ignorance; when we discover systematic effects where "error" existed before, the chance (at least in part) disappears. The exact events of an individual life could never be predicted any more than the exact path of a falling raindrop, because they depend so much on intercurrent happenings in altogether different systems; but that does not in any way rule out the possibility of general laws that determine these two kinds of "behavior."

Footnote 14: Until about 1970, I did not see that free will and determinism are not antithetical; indeed, personal freedom would be impossible in a non deterministic world. See footnote 5, above, and see also Weiss (1969) for a refutation of the position I took in the first version of this paper. His hierarchical conception of determinism is an important contribution to the systems outlook on an ancient philosophical problem.

7. General laws are not possible in personology because its subject matter is unique individuals, which have no place in natural science

It is not difficult to dispose of this last, supposedly critical point of difference between Naturwissenschaft and Geisteswissenschaft.

The mechanistic, reductionistic science of Windelband's day contained a curious dictum that has been one of the principal sources of confusion on this whole issue: Scientia non est individuorum�science does not deal with individual cases. This hoary slogan dates back to the days when Aristotle was the last word on matters scientific, and the whole point of view it expresses is outdated in the physical sciences. According to this philosophy, the individual case was not lawful, since laws were conceived of as empirical regularities. This is the point of view (Plato's idealism or what Popper [1957] calls essentialism) that considers an average to be the only fact, and all deviation from it mere error.

As Freud and Lewin argued, all forms of thought and behavior are determined too, and the individual case is completely lawful. It is, however, impossible to know what the laws are from a study of one case, no matter how thorough. We can surmise (or, if you will, intuit) general laws from a single case in the hypothesis-forming phase of scientific endeavor, but we can verify them only by resorting to experimental or statistical inquiry or both.

There is truth in the old adage only in one sense, then: we cannot carry out the complete scientific process by the study of an individual [Footnote 15].

Footnote 15: And also in Weiss's (1969) sense that many aspects of individuals are indeterminable, hence we might as well view them as "chance," even though the individual is thereby playing his part in producing an intelligible and predictable (or lawful) regularity on a higher level of analysis.

It is true that in certain of the disciplines concerned with man, from anatomy to sensory psychology, it has usually been assumed that the phenomena being studied are so universal that they can be located for study in any single person, and so autonomous from entanglement in idiosyncratically variable aspects of individuals that the findings of intensive investigation will have general applicability to people at large. Every so often, however, these assumptions turn out not to be tenable. For example, when Boring repeated Head's study (in one case, himself) of the return of sensation after the experimental section of a sensory nerve in his arm, he did not find the protopathic-epicritic sequential recovery, which had been so uncritically accepted as to be firmly embedded in the literature. No matter how intensively prolonged, objective, and well-controlled the study of a single case, one can never be sure to what extent the lawful regularities found can be generalized to other persons, or in what ways the findings will turn out to be contingent on some fortuitously present characteristic of the subject-until the investigation is repeated on an adequate sample of persons. As excellent a way as it is to make discoveries, the study of an individual cannot be used to establish laws; bills of attainder (that is, laws concerned with single individuals) are as unconstitutional in science as in jurisprudence. Note, however, that law of either kind, when promulgated, is still conceived as holding quite rigorously for the single individual. See also Holt, 1978, vol. 2, Chapter 8.]

Science is defined by its methods, not its subject matter; to maintain the opposite, as Skaggs (1945) did in an attack on Allport, is to perpetuate the confusion, not resolve it, and Allport (1946) was an easy victor in the exchange. There can be and is scientific study of all sorts of individuals. Particular hurricanes are individualized to the extent of being given personal names and are studied by all the scientific means at the meteorologist's command. A great deal of the science of astronomy is given over to the study of a number of unique individuals: the sun, moon, and planets, and even individual stars and nebulae. There may not be another Saturn, with its strange set of rings, in all of creation [Footnote 16], yet it is studied by the most exact, quantitative, and--if you must--nomothetic methods, and it would be ridiculous to suggest that astronomy is for these reasons not a science or that there should be two entirely different astronomical sciences, one to study individual heavenly bodies and the other to seek general laws. Further examples are easily available from geology, physics, and biology. Once we realize that individuals are easily within the realm of orthodox scientific study and that science does not strive for artistic illusions of complete understanding, the issue is easily seen as a pseudoproblem. Psychology as a science remains methodologically the same, whether its focus be on individual cases or on general laws.

Footnote 16: After these words had been written, I was amused to find that Cournot used this same example, and even similar wording, in supporting his position that "it is no longer necessary to accept to the letter the aphorism of the ancients to the effect that the individual and particular have no place in science"(1851, p 443). More recently, rings have been discovered around two other planets in our own solar system.

Granted, then, that individual personalities may and must be studied by the scientific method in personology, with the use of general concepts, what is the role of general laws in such a science? Where does it get us to make scientific studies of personalities, if each is unique, and if that uniqueness is the heart of the matter?

Personalities are in many ways unique, but as Kluckhohn and Murray (1953) point out, every man is also like all men in some ways and like a limited number of others in still other ways, making generalization possible. If every personality structure were as much a law unto itself as Allport implies, it would be impossible to gain useful information in this field; there would be no transfer from one case study to another. As anyone knows who has tried it, there is a great deal.

It is a mistake to focus personology on just those aspects of a person that are unique, as Weber saw clearly half a century ago.

The attempt to understand "Bismarck" - he said for example - by leaving out of account everything which he has in common with other men and keeping what is 'particular' to him would be an instructive and amusing exercise for beginners. One would in that case... preserve, for example, as one of those "finest flowers" [of such an analysis of uniqueness] his "thumbprint", that most specific indication of "individuality" (Weber, 1949).

And some of the most critical points about him for predicting his behavior would have to be excluded because he shared them with other persons. Indeed, in contemporary psychodiagnosis, it is considered most useful to treat as a quantitative variable the degree to which a person's responses resemble those of the group as a whole.

The only kind of law that Allport could conceive for personology was one (like his principle of functional autonomy) that describes how uniqueness comes about. Personology has not been much restrained from seeking general relationships among its variables by this narrow view, however; the journals are full of investigations in which aspects of personality are studied genetically (that is, are related to the abstract variable of age) or are correlated, one with another. Once one treats uniqueness not with awe but with the casual familiarity due any other truistic fact of life, it ceases to pose any difficulty for personology.

Writing intensive case studies (on the genesis and structure of individual personalities) turns out not to be a particularly fruitful method, except for the generation of hypotheses. This is a very important exception, but the point is that personology does not proceed mainly by adding one exhaustive scientific biography to another, looking for generalizations afterwards. The Gestaltist taboo on studying any variable out of its context in the individual life is an overstatement. There is, of course, such a phenomenon as the interaction of variables, but it is not so far-reachingly prevalent as to make impossible any study of two variables at a time. As Falk (1956) has shown, this condition of interactive non-summativeness is found in many other kinds of subject matter besides personality and creates no major difficulties of method or procedure [Footnote17].

Footnote 17: At the same time, we have to accept a low ceiling on the possible size of relationships discoverable in this way, and a lower one the further our level of analysis is from that appropriate to the kind of system under study. At the time I wrote this paragraph, I did not grasp the fact that obtainable information about a system is not exhausted by studying its elements and their interactions.

In summary, in this section we have looked at the major propositions of the romantic point of view as applied to personology and have found that the "basic differences" between this field and natural science are completely illusory. No basis for a separate methodology exists, and the objections to applying the general methodology of science to personalities turn out to be based on misunderstandings or on a narrow conception of natural science that is an anachronism today.

It by no means follows, as Eysenck (1954) puts it, that the science of personality should therefore be considered nomothetic. The nomothetic conception of science must be rejected as a caricature of what any contemporary scientist does. The only way to justify the application of the term nomothetic to the natural science of the present is to change the definition of the term so much that it no longer resembles its original meaning and becomes an unnecessary redundancy. It can only lead to confusion to introduce such (unacknowledged) changes of definition. The nomothetic is as dead a duck today as the idiographic, and neither term adds anything to contemporary philosophy of science [Footnote 18].

Footnote 18: Specifically, the nomothetic conception of science is basically the nineteenth-century tradition of mechanistic reductionism (Holt, 1971a), or what Whitehead (1925, p. 18) called scientific materialism (see also Yankelovich & Barrett, 1970; and Ackoff, 1974, Chapter 1). Since nomothetic and idiographic are antithetical, the conflict cannot be solved by compromise, either that of the moderate middle way or of the "sometimes one, sometimes the other" variety. What is needed is a true synthesis, a decentering shift in point of view or theory to a new level of observation and conceptualization. The systems outlook gives just such a synthesis in terms of which we can see the truth and the error in both prior positions (Bertalanffy, 1968; Laszlo, 1972).

Many psychologists have followed Allport in taking the apparently sensible "middle position" of trying to deal with the objections that have been raised to his extreme idiographic pronouncements by saying, let's have a personology that is both nomothetic and idiographic (e.g., McClelland, 1951; MacKinnon, 1951). Thus, whenever he approaches the realization that the idiographic discipline of which he dreams is unworkable, Allport says, in effect, "I am not an extremist; common traits have their uses, even though they are only approximations, and personology can use both nomothetic and idiographic contributions." In practice, what this amounts to is that whenever attention is focused on individual cases, the inquiry is called idiographic, and otherwise it is considered nomothetic.

My objection to this "solution," this apparently reasonable compromise between antithetical positions, is that it is achieved only by a perversion of the original definitions and that it accomplishes nothing except the preservation of a pair of pedantic words for our jargon. If one really accepts the arguments for an idiographic Geisteswissenschaft he can logically have no truck with nomothetic methods. They exist no longer, anyway, except in the history books; scientific method, as understood and practiced today in natural science and personology alike, is not a combination nor blend of nomothetic and idiographic approaches, but something bigger and better than both of them [Footnote 19]. The original dichotomies were badly formulated and based on misconceptions. The accompanying terminology might best be forgotten along with them.

Footnote 19: Alas, I was too optimistic when I wrote these words, underestimating the extent to which mechanistic reductionism flourishes in contemporary psychology. The fact is that Eysenck, Skinner, and other proponents of what Koestler (1967) tellingly calls "flat-earth psychology" continue to dominate our discipline, with many of their zealous followers occupying prominent roles in personology and clinical psychology.

The last stand of the proponents of the romantic dichotomy is the contention that there are distinct generalizing (nomothetic) and individualizing (idiographic) methods in personology. This is the point of departure for Stephenson (1953) and some others who are enchanted by the mystique of Q. Inflating his ingenious rating technique into a whole so-called methodology, Stephenson has argued that his device of rating on an "ipsative" instead of a "normative" scale creates a specifically idiographic method for personology. When one Q-sorts a group of items for a subject, he makes a set of ratings which are forced into a normal distribution and scaled according to each item's applicability to this particular person (which is ipsative scaling, as opposed to the usual normative ratings where the standard is the distribution in a population of comparable persons). The device is clever and often useful; it enables a judge to give quantitative ratings to a great number of variables for one person without any reference to any sort of standard population; the population is intrapersonal (Block, 1961).

Here is a technique suited to individual cases. Is it therefore idiographic, something fundamentally different from conventional scientific methods of rating personality? Hardly. Q-sorts are typically used in large studies in which the individual case is an anonymous statistic. Moreover, it is a kind of Procrustean bed, imposing a standard pattern of ratings on every personality: all must have the same mean, standard deviation, and distribution (normal, flat, or other). What is even further from the spirit of Allport's crusade for individual traits, the "items" are common traits, applied to everyone with no allowance for their failure to fit certain cases. In summary, then, the Q-sort is quite unacceptable in the traditional meaning of the term idiographic, and the use of that term to signify the fact that it is applied to individuals is simply a grandiloquent pose.

Following Allport (1942), others (e.g., Dymond, 1953; Hoffman, 1960 [and all too many more in the ensuing decades]) have revived the tired old terms either in an attempt to bolster, or in an attack on, the contention that clinical predictions must be superior to statistical predictions, because the clinician uses idiographic methods which alone are appropriate to predictions about individual cases. Here is another badly formulated pseudo-issue [see section on Prediction in Holt, 1978, vol. 2]. Whether a clinician or a formula does better in making a particular kind of prediction is an empirical question and one of little general interest. Clinicians and statisticians have their own proper spheres of activity, which overlap but little, and the difference between their activities has nothing to do with methodological issues. The method of clinical judgment has a great deal in common with the hypothesis-forming and theory-building phases of work in all the sciences [see Holt, 1961].

In the end, we see that there is no need for a special type of science to be applied to individual personalities and that the attempt to promulgate such a science fell into hopeless contradictions and absurdities. Today, Windelband's terms continue to appear in psychological writing but largely as pretentious jargon, mouth-filling polysyllables to awe the uninitiated, but never as essential concepts to make any scientifically vital point. Let us simply drop them from our vocabularies and let them die quietly[Footnote 20].

Footnote 20: Surely a trivial but none the less annoying characteristic of the wordidiographic furnishes a further argument for its consignment to oblivion: the strong tendency of printers to assimilate it to the more familiar but wholly different wordideographic (pertaining to ideographs or picture-writing).For example, Skaggs's paper (1945) contains only the misspelled version [After this paper's initial publication, a friend pointed out tome the ironic fact that the last statement is (almost) true also of Holt and Luborsky (1958)!]

Summary. The conception of two kinds of disciplines, a nomothetic science to study general principles and find abstract laws, and an idiographic science to study individuality, arose as a protest against a narrow conception of science in the nineteenth century. But the romantic movement of which it was a part started from fallacious premises, such as the conception that science is defined by its subject matter rather than its method, and its radical principles were never actually applied in pure form by any of its adherents. The idiographic point of view is an artistic one that strives for a nonscientific goal; the nomothetic, a caricature that bears little resemblance to the best contemporary work in the "hard" sciences. Since no useful purpose is served by retaining these mischievous and difficult terms, they had best disappear from our scientific vocabularies.