ABSTRACT: Different approaches to the classificnion of a knowledge field include empiristic, rationalistic, historicist, and pragmatic methods. This paper demonstrates how these different methods have been applied to the classification of psychology. An etymological approach is insufficient to define the subject matter of psychology, because other terms can be used to describe the same domain. To define the subject matter of psychology from the point of view of its formal establishment as a science and academic discipline (in Leipzig, 1879) is also insufficient because this was done in specific historical circumstances, which narrowed the subject matter to physiologically-related issues. When defining the subject area of a scientific field it is neccssary to consider how different ontological and epistemological views have made their inf1uences. A subject area and the approaches by which this subject area has been studied cannot be separated from each other without tracing their mutual historical interactions. The classification of a subject field is theory-laden and thus cannot be neutral or ahistorical. If classification research can claim to have a method that is more general than the study of the concrete developments in the single knowledge fields the key is to be found in the general epistemological theories. It is shown how basic epistemological assumptions have formed the different approaches to psychology during the 20th century. The progress in the understanding of basic philosophical questions is decisive both for the development of a knowledge field and as the point of departure of classification. The theoretical principles developed in this paper are applied in a brief analysis of some concrete classification systems, including the one used by PsyclNFO / Psychological Abstracts. The role of classification in modern information rctrieval is also briefly discussed. 1. Introduction Classification in Psychology and in Library and Information Science This paper is about the classification of a specific subject domain: Psychology. As such it is meant to be a contribution to both psychology and to library and information science (LIS or just IS).2 As shown by Miksa (1998) there has been a philosophical and interdisciplinary "Movement to Classify Knowledge and the Sciences" beginning in the seventeenth century with persons like Tommaso Campanella and Francis Bacon. During the Nineteenth century this movement became an activity of enormous proportions among a wide number of participants. "I sometimes speak of it as a time when anyone who was anybody in the realm of scholarship wrote a trcatise on the topic" (Miksa 1998, p. 34). Among the persons mentioned by Miksa is also Wilhelm Wundt (1832-1920). Wundt's contribution to the classification of knowledge may be obsolete, but his name is worth mentioning in this article. He was the person who most often gets the credit for the formal establishing of psychology as a science because he established the first psychological laboratory in the world in Leipzig in 1879. According to Miksa this movement to classify knowledge and the sciences died out just after the beginning of the Twentieth century. My own knowledge confirms this. It is extremely rare that articles on the classification of psychology or the place of psychology among other fields are published. Works such as Braun & Baribeau (1984) or the present article are exceptions to the rule. What are the reasons? They might be a combination of the following: (a) that nobody finds this problem of any interest any more. This could be influenced by the fact that (b) the norms of scientific methodology have changed. This kind of approach went out of fashion during the strong positivistic trends in the beginning of the Twentieth cen-tury. (c) Research has become more fragmented and also more applied in its orientation, which make this kind of classification research more difficult to justify. A final reason could be that (d) these problems appear too difficult. Nobody feels they have the necessary background to be able to make a contribution. It is my hope that this article might contribute to the understanding that the problems of classification of knowledge are both important and possible to tackle in interesting and fruitful ways. Recent interdisciplinary changes in the philosophy of science (from more empiristic and rationalistic tendencies towards more historical and holistic tendencies) also give some hope for such a change. But who should do this kind of research? Domain specialists (such as psychologists in this case)? Library and information scientists? Sociologists of science? 4 Philosophers? My answer is YES! They all should. Today all groups feel that this is beyond their competence. They feel that they have to learn too much about something that they are not primarily trained to do. We will all benefit if interdisciplinary research in this area begins to flourish. 2. On Methods of Classification In Hjørland (1998b), I have presented a short outline of my theoretical view regarding the methods of classification. It is my claim that different methods of classification basically reflect different epistemological theories as shown in Figure 1 below. Figure 1 Fundamental methods of Classification Research Objects In practice, however, classifications are often made without any explicit methodology; they are just based on the view or horizon of the persons who are doing the classification. 6 Research libraries and information systems (such as PsycINFO) often employ subject specialists to develop and update their systems or they import important parts of their system from recognized handbooks and other authoritative sources. But this is only to move the problem one step back: How do you know when a given source reflects "cognitive authority"? How do you distinguish between good and bad proposals? In order to evaluate this you must develop a theory about the methodology of classifying. Different methods of classifying are in a very direct way related to different epistemological theories. Insight in epistemology can thus provide us with knowledge about the merits and weaknesses of the different solutions. Progress in the scientific method as well as in classification (which may be seen as part of the scientific methodology) must be based on the historical evidence gained in epistemology and science studies. [Figure 1] Figure 1 shows the relationship between basic epistemological theories and basic methods of classifying. Classification is done in all sciences) including psychology [8]. Like any other science IS has different approaches to classification based on different epistemological views. IS is mainly concerned with principles for classifying documents produced in other disciplines) which imply classification on a second order level. Classification in IS is not restricted to documents but can be applied to all forms of "information" represented in information systems. Different sciences may influence each other. Frame-based systems and semantic networks are examples of classifications developed in Artificial Intelligence (AI) and are also applied in IS. "Facet analysis" is a method of classification developed independently in IS and in psychology. In my opinion there exist a limited number of basic methods of knowledge organization corresponding to basic epistemological views. A psychiatrist can, for example, classify mental illness using empirical methods, or rationalistic methods, or historical methods, or pragmatic methods (or, of course, combinations). In the same way, a psychologist can classify forms of intelligence or mental capacities by using statistical analysis of test scores (empirical method), by using computer models of cognitive processes (rationalistic methods), by studying the social construction of the intelligence concept (historical method), or by choosing a concept which fertilizes his general perspectives and aims (pragmatic method). On another level, information scientists can use the same kinds of methods to organize documents, knowledge, or information. They can use empirical methods such as bibliometric linking and produce maps such as the "atlas of science". They can use rationalistic methods such as developing facets or principles for logical division, they can use historical methods such as revealing the cultural bias in different systems, or they can select classifications which support the aim of their activities. Traditional ideals of classifying (as well as other aspects of scientific methodology) have tended to be empiristic or rationalistic, providing "neutral" or "objective" classifications. Modern epistemology, however, emphases the theory-laden character of observations, as well as the theory-laden character of classifications: They are not neutral discoveries but constructions which favor some kind of activities at the expense of other activities. This important insight is today often associated with "post modernism" (cf. Miksa 1998), but it was already developed by the pragmatic philosophers such as John Dewey in the beginning of the 20th century. It was, however, repressed by more empiristic and rationalistic influences (cf. Hjørland 1997). This insight implies that we need to move from more positivistic approaches in cIassification toward more interpretative and neopragmatic approaches. Conclusion: A classification cannot be neutral regarding approaches or theories about its subject matter. On the contrary: The classification of a subject field requires a conception or view of that particular field. (This does not, however, imply that the problems of classification only belong to the single disciplines and cannot be approached in fields like science studies or IS. There may be general approaches to analyzing the subject domains, and such knowledge is not typically part of the knowledge of the members of specific disciplines. In this paper, I shall try to show how this can be done more concretely.) [9] 3. A Short Outline of the Problems of Defining Psychology, its Elements, Methods and Structure The term "psychology" goes back to about 1400-1500, but it first came into common use about 100 years later because of the works of Christian von Wolff (1679-1754) [10]. "Psychology" is, however, just one term among many which have been used as a label to describe the subject area, which it is meant to represent. Aristotle (384-322 B.C.) is considered to be the first person who has given a systematic description of psychological phenomena in the book "De anima" ("about the soul"). Until the establishment of psychology as an "independent science" in the latter part of the 19th century, the study of psychological phenomena was mainly done in philosophy, but also in theology, in medicine, and in other fields. But what are the psychological phenomena? [13] And what principles define the subject area of psychology? The formal establishment of psychology as a science was done under certain historical conditions, which favored certain views and approaches of the times. Psychology was first and foremost recognized as a science because it applied the experimental method. In American textbooks on the history of psychology this approach, which was founded by Wundt, has often been termed "structuralism" and is said to have died in America with E. B. Titchener (1867-1927). It was replaced by "functionalism" and behaviorism and other schools. This is, however, just one interpretation. Another interpretation says that Wundt (and in particular his predecessors Ernest Weber, 1795-1878 and Gustav Theodor Fechner, 1801-1887) founded psychophysics, which is a strong scientific subject area in psychology even today (even though it was born with very problematic metaphysical assumptions). What remains a fact is that psychology became split [14] by many different approaches which tend to define their own subject matter and classification of psychology. Titchener did not, for example, recognize child psychology or animal psychology as parts of psychology, whereas these areas were very central in the contemporary school "functionalism". It is important to realize that these are not "accidental" properties of those theories, but that each theory, approach or "system" in psychology implicates in a very strong sense the subject matter and classification structure for psychology - even though this is seldom explicated very well. [15] The formal establishment of psychology as a science [16] also raises an important question regarding the relation between "scientific psychology" as understood as that part of the psychological knowledge that is produced inside the borders of the formally established discipline, and the production of psychological knowledge produced outside these borders. A common sense consideration would say that psychological knowledge is that which is produced by psychologists (that is, the subject area and its formally educated workforce define each other in a mutual way). This is, however, a very problematic assumption, especially in the case of psychology. When psychology was established as a formal discipline about 1879 this was done on what must today [17] be considered a very limited approach compared to the huge amount of psychological knowledge produced in philosophy, theology, biology, medicine, and also outside the academic world. In the process of its development psychology not only questioned its own foundation and established new different - and often conflicting - approaches. It also began to absorb (and perhaps to monopolize) other areas. Kurt Danziger exemplifies how the study of motivation became part of academic psychology:
In 1928 the Harvard psychologist, L. T. Troland, published the first general text featuring the word "motivation" in the main title [...] No doubt, the massive popularization of psychoanalysis in the post-war period played a significant role in establishing a link between the subject of psychology and the exploration of individual motives. In the subsequent academic literature on the topic of motivation Freud is always mentioned as a motivational theorist, if only to repudiate his theories as unscientific. Among other things, the construction of the new field of motivation enabled academic psychology to extend its dominion to topics that psychoanalysis had put on the algenda and threatened to monopolize. The new science of motivation began to act like a superior court that would adjudicate the truth claims of other psychologies. (Danziger 1997, 111)
Danziger also describes how educational psychology and many other areas of applied psychology became parts of the overall discipline of psychology:
The story of twenty-century academic Psychology is the story of an ultimately unsuccessful struggle against an ever more obvious fragmentation. Intelligence and its testing provided an early example of the discipline's tendency to annex new alreas without being able to assimibte them. [18] Psychologists had gained an academic foothold by doing experiments on such topics as sensation, perception and memory. For some time, that remained the respectable core of the discipline, but how test intelligence related to this core was far from clear. It was much easier to annex such a field institutionally that to assimilate it intellectually. The situation was to be repeated many times over in the course of the twentieth century. Child study, or paedology, as it was known in some countries, was another example. Originating in joint efforrs by physicians and educationists, it became transformed into child psychology, rapidly in the US, more slowly in Europe. But its links to core areas of the discipline remained tenuous at best. The same could be said of educational psychology, another early branch. In the period between the two world wars the discipline sprouted as many arms as Shiva, the Hindu deity. A psychological social psychology challenged its sociological rival, "personality" and "motivation" emerged as semi-autonomous fields of research and teaching, industrial psychology flourished, clinical psychology became a reality. What link was there between these fields, except that they all claimed to be "psychological"? But did that mean anything beyond a vague sense of common focus that was based on popular images rather than on solid scientific grounds? Grouping these diverse areas together as branches of one discipline undoubtedly had certain practical advantages. It advanced the cause of professionalization by implying that the more practically oriented branches had a respectable link to basic science, and it legitimized the otherwise esoteric interests of the academics by implying that their work had significant practical applications. But, for the most part, such implications were nothing more than promissory notes to be cashed in at some time in the future. Why should anyone accept these notes, and, more importantly, how could psychologists justify such promises to themselves? In the period under consideration here [second to fourth decades of the 20th century], that justification depended to a very large extent on the notion that, ultimately, Psychology was one discipline whose various branches would turn out to be linked by one set of principles. The grouping together of diverse fields of research and practice under one umbrella would then be more than a matter of historical accident and administrative convenience, it would be the logical consequence of deep theoretical links; common scientific "laws" would unify the discipline. As a first step in this direction, the various parts of the discipline would need to be tied together by common categories of discourse. Such common categories would establish the claim that there were indeed phenomena of importance that were common to all fields of Psychology. Then one could study these common phenomena in order to discover rhe principles that unified the discipline. This was the role played by the categories of "behaviour" and "learning" in the history of twentiethcentury American Psychology. [19] Of the two? "behaviour" was more foundational, for it became the category that the discipline used to define its subject matter. [...] The history of "behaviour" is not only intertwisted with the history of "learning", it is also deeply entangled with the history of behaviorism. Unlike the other categories considered here, it has the unique distinction of having given its name to a movement. That leads to certain difficulties. We have to be careful not to confuse the history of the movement with the history of the category. They are far from being the same. (Danziger 1997, 85-86)
The main thesis in Danziger's book (1997, p. 17) is that until the second half of the 20th century there was not one but several disciplinary languages of psychology [including categories of psychology and implying implicit classifications of the discipline], and each of them had its own historical trajectory. In the aftermath of World War II, however, the language of American Psychology was adopted virtually everywhere, a situation that has only begun to change recently. The period between 1910 and 1940 was a time of revolutionary change, not because the theories that explained the phenomena were changed, but because the phenomena themselves changed. They changed because the categories that defined them changed. A striking feature of the discipline in its American incarnation was the impressive degree of uniformity achieved in its discourse - at least for a time. If one were to give a name to this hegemonic form of discourse one would have to call it "behavioral". This does not mean that most American psychologists were behaviorists, a judgment about their explicit theoretical commitments. Whatever those commitments might have been, most of them were quite ready to use the specialized terms of their discipline in a manner that conceded many of the assumptions of behaviorism and made them invisible. In chapter ten Danziger raises the question of whether psychological categories can be said to constitute "natural kinds", whether they mirror the structure of a psychological reality that exists independently of them. After considering the social contextualization of these categories and their referential role, that question is ultimately answered in the negative. [20] What have we learned from the above cited concerning the classification of psychology? First of all, we have learned that it is very important to make a distinction between different kinds of concepts. We must analytically distinguish between the concept of discipline (and subdisciplines), scientific categories, subject areas, and approaches (or schools/systems/paradigms) as units for classification. We have seen how the concept of discipline illuminates the influences of the formal establishment and institutionalizing of a subject area. The division of academic labor in society has an influence on which subject areas are included and also the theoretical and methodological approach towards those subject areas. [2l] We have also learned that disciplines can continue to exist and grow even if the criteria that played the decisive role in their establishment are later abandoned. They seem to follow a principle which Allport (1937) called "functional autonomy". Psychological areas or subdisciplines are, for example, psychological processes within organisms (such as perception, learning, memory, emotions and motivation), developmental psychology, the psychology of personality, social psychology, and applied psychologies (e.g., clinical psychology, educational psychology, industrial- and organizational psychology etc.). Such areas can be seen as the object studied by specialized groups of people in the discipline, and in this respect they exist a priori to different approaches and theories. On the other hand, as science develops, areas and approaches studying these areas become more connected. In the case of psychology many subject areas existed before the science was formally established, and in the history of its development, these preexisting fields influenced the scientific approach just as principles and methods developed in the scientific organization influenced the subject areas. Psychological approaches (or movements, views, schools, paradigms, currents, etc.) are the theories, basic forms of understanding, or ideas (ideologies), which have influenced the development of psychology. They have designed the categories as well as the theories and the concrete knowledge, which is established inside and outside the discipline. Examples are behaviorism, cognitivism, psychoanalysis, humanism, and the historical-cultural approach (see also appendix 3). According to Kuhn's (1970) theory of scientific revolutions scientists organize themselves around such paradigms. Kuhn differentiated between pre-scientific phases or stages, normal or paradigmatic phases, revolutionary phases, new paradigmatic phases and so on. Real science is characterized by normal phases in which there is almost total consensus concerning the discipline's fundamental approach. It is a question of interpretation whether Danziger's demonstration of the generality of a soft behaviorist approach in psychology could be said to represent normal science. I would say no. However, a closer discussion would have to include a much more careful analysis of Kuhn's theory and newer related works in the philosophy of science. Psychological categories are the terms that a discipline (or a given approach) regards as its fundamental concepts. In psychology this is, for example, cognition, emotion, learning, personality, attitude, intelligence, etc. Danziger (1997) shows that around half of the fundamental concepts in psychology afe constructed in the twentieth century. The rest originate from the philosophy of earlier periods. Danziger also shows that a psychology oriented towards the Dat ural sciences tends to conceive psychological phenomena as "natural kinds", whereas modern constructivistic approaches tend to conceive them as historical products, as "human kinds". There is a very complicated interaction between the developments of (sllbJdisciplines, categories, and ap· proaches in the hist01Y of psychology. I have already indicated that each approach in a very fundamental way establishes its own subject matter. In my opinion, the analysis of the systems of the sciences must start by uncovering the most basic philosophical and theoretical assumptions and hence identifying the basic approaches or paradigms at the interdisciplinary level as well as at the disciplinary level. I find the "social constructivistic" method applied by Danziger and others absolutely necessary, but that does not mean that I am not a scientific realist. We should start our analysis with the examination of the fundamental approaches, but it would be relativistic to think that one approach is as good as any other. Reality puts limits to which approaches can survive in the long run and is a determining part in the development of knowledge (maybe to a lesser degree in the social sciences compared to the natural sciences). However, the kind of "realism" found in empiricism where it is supposed that science can uncover reality from observations alone, disregarding history and theories, I find very nai"ve and dangerous. Figure 2 illustrates the possibility of analyzing each subject area in psychology from the point of view of each theoretical approach (and vice versa). In a given field (say developmental psychology) it is possible to trace the influence of behaviorism, of psychoanalysis, of cognitivism, of humanistic psychology, of activity theory, and so on. Sometimes such a mapping will be easy, sometimes difficult. Psychoanalysis concentrates on personality, motivation, and emotions. Psychoanalysis is sometimes said not to be about cognitive processes. But it is possible to write books about the psychoanalytic view on cognition, and it has been done. Each theoretical system favors some subject areas and neglects others. Even if nothing has been said directly about a given connection between an approach and an area, it is possible to generalize some principles and to draw some conclusions about these connections. It is, however, important to realize that Figure 2 is only a purely analytical table. The table gives the impression, that these connections can be analyzed as purely external relationships. In reality (as documented in the history of psychology) there are strong internal connections between a given approach and its subject matter. It is also important to recognize the interactive character of these relationships. Developmental psychology is not only some psychological approach used in the study of the psychology of children. Child psychologists are studying a reality (something that develops), and this tends to influence their thinking and theories in a way which makes it more unlikely to develop "static" theories in this field. This again implicates that knowledge domains tend to be characterized both by the phenomena that are studied and the approach which is used to study those phenomena. This again implicates that the methods of psychology are not primarily an independent subdiscipline, but is primarily part of all subdisciplines [22]. Figure 2 Connections between Approaches and Areas of Knowledge in Psychology Our overall conclusion on this section three is that in defining psychology, its elements, methods and structure the basic theoretical approaches (or paradigms) must be taken as the point of departure. You can of course classify psychology by using purely empirical or rationalistic methods on a given set of psychological data (including bibliographical data or literary data). However, the historical development of psychology demonstrates the limitations of these approaches. Just to argue whether one set of data is useful or sufficient for an empirical or rational analysis presupposes that you usc criteria which must in the end be justified by theoretical criteria of what constitutes psychological knowledge. There is no escape from deep theoretical involvement. The question is what kind of methodological approach can be applied to uncover the basic classificatory structure? And can such a methodological approach be of a certain generality so that it can be used to study the classification of other disciplines as well? My working hypothesis is that an epistemological approach is one necessary element in classification research and that it may fulfill the requirements of being sufficiently general. [23] 4. The Epistemological Basis of Psychological Theories There is a close relationship between psychological approaches such as behaviorism, cognitivism, and psychoanalysis on the one side, and epistemological theories such as empiricism, rationalism or historicism/hermeneutics on the other. This is described in most of the traditional histories of the field. It is, however, often assumed that this link is cut when a science breaks loose from philosophy. The connection between a science and philosophy is itself an epistemological question where more positivistic theories emphasize the independence of the sciences, whereas more hermeneutic theories emphasize the connection between the sciences and their often implicit assumptions in ontology and epistemology. It is my thesis and claim that not only the study of psychological theories should be done from an epistemological point of departure, but also that all classification research should do so. This claim must of course be defended, and the relevant contra-arguments must be considered. However, this is my present approach, and the consequence of rejecting this claim is to give up this approach to a theory of classifying subject domains. If no better approach to classification can be provided, this is a very serious problem for LIS. Therefore this claim must be considered carefully. In Figure 3 is shown a simple table of connections between psychological approaches and their main epistemological assumptions. Figure 3 Connections between Theories in Psychology and Epistemology Behaviorism, cognitivism, psychoanalysis, activity theory, etc., are different approaches or theories to the same phenomenon: the human mind [24]. Each of these theories implies its own subject matter for psychology and hence its own conceptual system and classification. Behaviorism is a psychological approach (much related to empiricism), which implies that the subject matter of psychology is behavior, learning, responses, discrimination, and so on. Cognitivism (related to rationalism) is an approach that implies that the subject matter of psychology is information processing of the mind, short- and long-term memory, attention, topdown and bottom-up perceptual processes. Humanistic psychology (related to hermeneutics) is a nondeterministic psychology emphasizing the understanding of persons by means of humanistic methods. The subject matter of psychology is seen as the analysis of concrete personalities as well as human goals, ambitions and choices. The most extreme form of humanistic psychology represents existential psychology (related to philosophical existentialism) where people are seen as responsible for their own existence. The basic choice for human beings is the choice between suicide or continuing to live. If people choose not to commit suicide, the next choice is whether to choose your own life or just to follow the pattern put forward by external circumstances. Psychoanalysis (related to hermeneutics - at least in some interpretations) is an approach that implies that the subject matter of psychology is dreams, neurosis, unconscious processes, and symbol analysis. Cultural-historical psychology/activity theory (related to pragmatism and scientific realism) implies that the subject matter of psychology is human adaptation to various physical, biological, and cultural conditions, that is how languages and cultures form human psychological capacities, processes, and personalities. Some psychological approaches are easier to analyze from an epistemological point of view than other approaches arc. Behaviorism is an example of a relatively easy approach because it is so clearly related to empiricism and logical positivism. Psychoanalysis is much more difficult. Because my suggested approach depends on whether it is in fact possible to analyze all the most important approaches, I should be able to identify the epistemological basis also of psychoanalysis. Andkjaer Olsen & Køppe (1996) regard psychoanalysis as something new, which is irreducible to both mechanicism and humanism. Schultz (1988) regards Freud as the Hero of psychology, but finds that the problem with psychoanalysis is that it is not built on a realistic theory of knowledge. In appendix 4 I have provided a classification of psychoanalytic approaches based on Andkjær, Olsen & Køppe (1996). I find it useful in itself (it could, for example, be applied in the PsyclNFO-database). But I also think that it confirms the hypothesis about the basic role of epistemology in analyzing psychological theories [25]. Maybe the most problematic aspect of establishing psychology as "an independent science" was the tendency to neglect philosophical studies, to regard them as obsolete and to concentrate too much on cumulating empirical facts. All empirical research depends on the theoretical outlook of the researchers, and in the case of psychology a very broad theoreticaL historical and cultural outlook is necessary. The different approaches in psychology reflect the researchers' outlook, and these outlooks tend to reflect - more or less unconsciously and contradictorily - the theories of knowledge. In my opinion, we can come a long way in the understanding of psychology by analyzing this discipline from theories such as empiricism, rationalism, and historicism, even if this classification is a crude one. Empiricism is a philosophy that favors perception and experiences. It arose, in part, together with rationalism from different ways of drawing epistemological and methodological lessons from the ongoing progress of the scientific revolution inaugurated by Copernicus and consummated by Newton. Where empiricism favors observation, rationalism has the opposite tendency and is thus a philosophy that places less relative emphasis on sensory experience and more on reasoning and a priori theorizing. Together, rationalism and empiricism constitute the two main tendencies of European philosophy in the period after Scholasticism and prior to Kant. Empiricism saw people as born without any knowledge ("tabula rasa"), and all the knowledge an individual obtained came from the senses. Users form simple concepts from simple sense impressions. By the laws of association more complex concepts could be formed in the individual. Experiences must always be fragmentary and private. The 20th century has been dominated by empiricist philosophy, especially by logical empiricism and logical positivism from about 1920 up to 1950. In psychology this view has especially been carried on in behaviorism, which dominated American psychology from 1913 to about 1965. This view has more recently influenced "cognitive science" in theories about neural networks and "connectionism". Rationalism, on the other hand, saw sense experiences as a limited way to obtain knowledge. In order to perceive something, a person must already have a certain psychological makeup, which permits her to interpret the sense data. A person must have some concepts and these concepts cannot come from the senses but must be inborn (or they must develop from some pre-form, which is inborn). In modern terms: The brain must run some programs or follow some rules which determinate the fate of all input and the actions of the individual. With the computer revolution came a new rationalist trend which dominated in the 1970s and 1980s. In psychology it dominated "the cognitive revolution" starting about 1956 with the psychologist Jerome Bruner and the linguist Noam Chomsky dominating psychology from about 1965 and culminating about 1985. It was closely connected to research in "artificial intelligence" and to the interdisciplinary field known as "the cognitive sciences". Today there is a reevaluation and discussion about the status of this interdisciplinary trend (see Johnson & Ernding 1997), and many people find its epistemological assumptions very problematic. Historicism is a philosophy that emphasizes that perception and thinking are always influenced by our language, culture, by our preunderstanding and "horizon", including our scientific theories. Historicism has a strong connection to the humanities where hermeneutics has been dominating for centuries. As a theory of science historicism has especially evolved as scientific realism, which is an evolutionary epistemology developed within American pragmatism (by Charles Sanders Peirce) and within historical materialism (by Friedrich Engels) in the 19th century. Historicism thus agrees with rationalism in the view that our experiences arc determined by our psychological make-up. However, it does not see this make-up as something inborn or common for all human beings, but rather as determined by cultural factors. Cognitivism compares the human mind with a computer and tries to explain logical thinking, the working of the memory and decision making as governed by rules which can be uncovered and used in systems with "artificial intelligence". Historicism, however, understands psychological mechanisms as culture-determined. "Logical thinking" in "developed countries" is opposed to "wild thinking" in "primitive cultures". One explanation is that the development of written language changed the cognitive functions. In cultures with written languages it is possible to compare the formal structure of sentences, whereby formal rules of logical thinking can be formulated and taught. Even members of a literate culture who have not had courses in formal logic will be affected by this new way of thinking (Goody 1987). Such a way of explaining logical thinking is very different from a cognitivist's assumptions. In this way the psychologist Lev Vygotsky (1896-1934) sees higher cognitive functions such as memory as determined by culture. Primitive societies think more in pictures, where developed societies have a more verbal functioning of memory. The memory of small children is working by biological principles, but with the learning of a language memory begins to be working on a new higher level determined by sociocultural factors. This cognitive theory was already developed around 1930, but only today (in the 1990s) it seems to represent a main stream in American and international psychology. In the philosophy of science historicism has been influential in the work of Thomas Kuhn (1970). His theory about scientific "paradigms" reflects how the processing of information by scientific knowledge producers (and users) is determined by more or less conscious assumptions.26 Kuhn's theory bridges the individual and the collective level in cognitive processes. In the 1990'ties, historicism seems to become a dominant epistemology. There are several different schools working under the broad headline of historicism, for example, hermeneutics, pragmatism, social constructivism, semiotics, and activity theory/the cultural-historical school. It is beyond the scope of this article to bring an introduction to each of these. 5. The Pragmatic Understanding of Knowledge Production The sciences did not arise as responses to the practical needs of human beings. Rather the sciences arose in ancient history as activities connected to religious beliefs. Only with the Enlightenment (from about 1680) the belief in progress in society depending on the development of the sciences became a dominant vIew. Also in the development of psychology persons such as Ernest Weber (1795-1878), Gustav Theodor Fechner (1801-1887), and Wilhelm Wundt (1832-1920) founded psychology (or rather psychophysics) on the basis of metaphysical assumptions inspired by religious beliefs, and not by the wish to produce practical knowledge (at least not primarily). Weber, Fechner & Wundt worked under metaphysical assumptions from the dualistic tradition of Descartes, in which the human mind is something totally different from the physical world. In fact, the fathers of psychophysics were much inspired by the hope that their research could prove materialism wrong and provide a basis for religious beliefs. Despite such problematic metaphysical assumptions, they succeeded in discovering some very important methods, laws and principles for psychology. That fact that problematic metaphysical assumptions many times in the history of science have led to important scientific discoveries does not, however, mean that metaphysical assumptions do not matter. Indeed, such assumptions are very important because they are the glasses through which scientists look at the world. Such glasses can be more or less helpful or harmful. One of the very important functions of such religious beliefs has been that they allowed scientists to spend much time and energy on attacking problems without any substantiation for practical relevance. A related function has been the search for beauty, which resulting in, for example, mathematically formulated laws. The pragmatic and functionalistic understanding of human psychology can be traced to evolutionary biology and Darwinism. This understanding is very different from dualism, which views the spirit as something different from the material world. Pragmatism understands human psychological processes and structures as parts of the adaptations of living organisms to life on earth. Perception, memory, emotions, intelligence, motivations and so on are products of three interwoven lines of development: biological development (phylogenesis), cultural development (anthropogenesis) and individual development (ontogenesis). Pragmatism/functionalism tries to understand how human psychological phenomena can be understood and explained as adaptations to the environments in their development. A pragmatic understanding of psychology is not identical with the understanding of psychology as an applied science [27]. Applied science takes its point of departure from some formulated problems, for example, the selection of exceptional students for special education or the effectiveness of different methods of treatment of mental diseases. Basic or fundamental science, however, takes as its point of departure some problems formulated by scientists themselves, and for which there need not exist any practical utility at that moment. Basic science in psychology asks questions such as: what are the nature and cause of different perceptual illusions? How many functional memory systems exist in man (e.g., long-term memory and short-term memory)? To what degree are human mental abilities inborn, and what can twin-studies tell us about this? What determines the development of the individual personality? Psychologists who are oriented towards pragmatic epistemology can thus be either applied scientists or basic scientists or both. The connection between basic and applied science is all but trivial in the development of psychology. In the history and philosophy of psychology there has been some interest in the philosophy of applied science (see among others Danziger) 1990 and 1997; Brocke 1980; Hoffman & Deffenbacher 1994; and Schonpflug 1993). Danziger (1990, p. 120 + 126) writes:
"For the very term "applied psychology" reflected the myth that what psychologists put to use outside universities was based on a genuine science, much as engineers based themselves on physics. [Note 6: "The notion that technological change was due to the application of science was part of the popular rhetoric of science at the time, and psychologists were able to deploy it effectively because it was such a pervasive illusion ... "]. ... p. 126}: "If we distinguish between the research published in the applied journals and that published in the basic journals, it is clear that during the interwar period it was only the latter group that was undergoing something like a revolutionary development. The pattern for applied-research styles had been essentially established at the end of World War I, and in the ensuing years no fundamental changes occurred. This was basically a Galtonian style of research concentrating on the distribution of psychological characteristics in natural or psychometrically constituted populations. Thus, by this time there were two quite divergent styles of psychological research in existence. One worked with data from individual subjects reacting under laboratory conditions, the other with populations surveyed statistically. The one had the weight of tradition and the mystique of the laboratory behind it, the other was buoyed up by apparent practical success and immediate social relevance"
The development of psychological knowledge is not only monodirectional from basic science over applied science to practical applications) but also in the opposite direction. Some approaches to basic research in psychology are more open to applied concerns than other approaches are. Wundt's psychology and the so-called "structuralistic" approach in American psychology were relatively uninterested in applied psychology and were also rather unfruitful for the application of psychological knowledge. With the introduction of pragmatism and functionalism came a much more fruitful exchange between theory and practice. Also psychoanalysis is known for its very close relations between theory and practice) basic and applied research. "Applied psychology" (for example in pastoral care) can be traced far back in the history of mankind, long before the formal establishment of psychology as an independent science. According to Danziger (1997) p. 8), the newly established discipline of psychology began to annex these areas (for example child study and educational psychology) without being able to assimilate them theoretically. Thus, the semantic connection between terms that designate subject areas can be more or less theoretically justified or can just represent kind of disciplinary imperialism. Applied perspectives can contribute with valuable perspectives to the development of a knowledge field. However, applied orientations in knowledge production do also have their great disadvantages. It is a fact in the history of science, that basic science has had tremendous practical consequences. The real and deep understanding of the nature of the mechanisms underlying phenomena is often much more valuable than the more superficial attempts to understand phenomena in the frame of some practical problems. It is) for example) better to understand the growth of the normal cell and the cancer cell than just to try one cure after another. Also the classification of phenomena in the sciences according to deep theoretical principles (as) e.g.) in biological taxonomies) is in the larger perspective much more economical than just to classify according to narrow practical purposes (as) e.g.) in domestic animals) pets) and pests). It is important that science can grow according to scientific needs (how they should be described may be difficult, but that is another matter). If science is too much controlled by external factors, there is a real danger that knowledge will not accumulate) that skepticism will flourish, and that knowledge will become fragmented and disorganized. In this way the problem of the classification of knowledge is connected to questions regarding the working conditions of researchers and the relations between researchers and the rest of society. Therefore a pragmatic perspective should not only understand its phenomena in a broad perspective emphasizing the understanding of the phenomena in their development and their mutual relationship with their environments. A pragmatic understanding should also emphasize the interaction between the phenomena and science itself. Science should be reflective and consider its own history. Understand how different motives and interests in science tend to give priority to certain ways at looking of the phenomena) and to analyze how scientific concepts) theories) methods, institutions and so on can or do represent "social constructions". A pragmatic approach to psychology is thus an approach which emphasizes the development of psychological phenomena as adaptations to the environments of organisms and persons. It also examines the motives behind psychological theories, concepts and approaches. It asks: "What practical difference does it make whether this theory is correct?" It tries to develop knowledge that is at the same time relevant for human activities and represents deep structures of reality. There is no conflict between a pragmatic view and a realistic view. On the contrary: a realistic epistemology must be based on pragmatism (pragmatic realism). 6. The Pragmatic Understanding of Classification Just as a pragmatic view of science (e.g., of psychology) is different from an applied view, so is a pragmatic view of classification different from a view of the applications or concrete purposes of classification. The pragmatic understanding of classification conceives a classification as a tool, and as such more suited to some purposes, goals and interests than to other purposes. A classification is never neutral but reflects - consciously or unconsciously - certain values and views of the thing classified and the use of the classification itself. According to John Dewey a classification is objective in the sense that there exist objective standards for its goodness:
Nevertheless there is a genuine objective standard for the goodness of special classifications. One will further the cabinetmaker in reaching his end while another will hamper him. One classification will assist the botanist in carrying on fruitfully his work of inquiry, and another will retard and confuse him. The teleological theory of classification does not therefore commit us to the notion that classes are purely verbal or purely mental. Organization is no more merely nominal or mental in any art, including the art of inquiry, than it is in a department store or railway system. The necessity of execution supplies objective criteria. Things have to be sorted out and arranged so that their grouping will promote successful action for ends. Convenience, economy and efficiency are the bases of classification, but these things are not restricted to verbal communication with others or to inner consciousness; they concern objective action. They must take effect in the world. (Dewey 1948, pp. 151-154)
Applications of classifications are manifold. Library and information science is especially interested in classifications as tools for information retrieval in databases, for the organization of information in libraries and bibliographies, in short, as a tool for supporting the information seeking activities of staff and users. LIS is the field where most explicit interests in and analyses of the problems of classification of knowledge fields are taking place. Another group of users of classification are publishers of great handbooks in specific disciplines (this is especially a German tradition, see appendix 5). In a more fundamental way, however, the basic organization of knowledge reveals itself in the way disciplines are organized in universities and similar institutions for the production and teaching of knowledge and in the structure of scientific publications (especially journals). The way the scientific journals classify new knowledge in their selection criteria and in their mutual delimitation reflects a basic structure of classification. This structure is, however, not explicit, but can be more or less known by researchers and by information specialists or can be analyzed by empirical studies. It should also be clear, however, that this more or less hidden structure reflects many agents' different needs and is rather a representation of the producers' needs and possibilities than a representation that would satisfy the needs of users and information seekers. It is important to realize that the needs of different agents can be in conflict, and that a classification has to consider what kind of needs it is going to fulfill. Just as applied science tends to make knowledge production less deep, less coherent and less well organized compared to basic science, applied classifications can be marked by some local or accidental factors which blur its deeper structure. Classification research should provide a basic approach to classification, which can then be modified in different specific applications. LIS has been interested in both universal classifications covering all knowledge fields and in specific classifications covering a single field of knowledge. Both kinds of systems do have their justification. However, when users seek information about some specific matter, the disciplinary systems are most suited to their needs because they can display the knowledge field from a specific point of view without making compromises regarding how other disciplines would like their subjects represented. Ovesen (1989) describes how the discipline of anthropology almost disappears in the Danish Dewey System (DK5) because almost all anthropological subjects in his view are placed under subject headings which are names of other disciplines (like sociology). A universal classification always has to make compromises and to decide whose interests should primarily be taken care of. In doing so, it is of interest to know how each discipline would like to represent its field of knowledge. Therefore, I find that disciplinary classifications - the subject of this article - are of primary interest both for subject specialists and for LIS. I also see a need for general classifications, but this is beyond the scope of this· article to discuss. Regarded as a tool for information retrieval classifications have to compete with many other ways of accessing information. Today we have a lot of opportunities such as searching in titles, abstracts, full text, citations and so on. A theory of information retrieval should be able to specify the relative strengths and limitations of each subject point in retrieving information. Thus also to specify the relative strengths and limitations of classification codes in relation to all the other possibilities. Seen in this way, a classification of a subject domain in a bibliographical record is represented by a symbol which can be used in identifying relevant information. Behind this symbol exists a whole classification scheme that structures the subject domain in one specific way. In a way we have more than one classification in a database and represented in each record. Descriptors (from thesauri) are also a kind of symbols from a classification (and combinations also exist). What this article is about is the more traditional kinds of classifications which provide an overall mapping of the knowledge field in a top down fashion dividing the subject domain in a number of classes, which are then subdivided and so on. This is not the most important form for information retrieval, but it is one form among others, and it does have certain useful functions (see also Hjørland 1998a+c). I will conclude this section by stating that from the pragmatic point of view proposals for the classification of psychology should be evaluated by the following criteria: A psychological classification should represent all the most important approaches and subdisciplines in psychology (based on empirical, rationalistic, historical, and pragmatic evidence). A classification is expected to identify, label and systematize the main production of knowledge. A psychological classification should reflect an understanding of the history of the discipline, its different approaches, areas and perspectives. A psychological classification should be explicit about the view on psychology on which it is based. A psychological classification should reflect theoretical views on the connection between psychology and other sciences A psychological classification should reflect theoretical views on the connection between different subclasses/subdisciplines of psychology A psychological classification should avoid the reduction of psychology to either biology or sociology A psychological classification should - if possible - be based on a theory about the object of psychology and about its units,28 Research on the classification of psychology should thus include research in issues such as: Is "General psychology" a subdiscipline of psychology? Are both a sociological and a psychological "social psychology" an empirical reality? Whether or not this is the case, one should further ask: "Should there exist two or more social psychologies?!! (providing arguments pro et can and a conclusion, This conclusion could later be changed by new evidence). What are the relations between ethology and (animal) psychology? What are the relations between psychiatry and clinical psychology? To what degree is research in specific fields (e.g., child development) interdisciplinary? And so on Classification research should analyze concrete domains as well as relationships between domains and similar patterns across different domains. Such issues are difficult and cannot be expected to be answered once and for all. If classification research does take itself seriously as research it tries to illuminate such problems without jumping to too fast conclusions, Practical classifications should be made all the time building on the accumulated knowledge at the time of construction (and should be evaluated on that background). The essential result of research in classification should not as much be seen as concrete classifications as a repertoire of arguments for and against different ways of classifying different subject domains. Classification research should build on a realist epistemology and should never be regarded as finished. Real breakthroughs in the classification of the sciences are rare (e.g., Linne's botanical system, the atomic system of Rutherford and Bohr and recent changes in biological taxonomies) and are connected with theoretical breakthroughs in the sciences. Classifications are, however, not only a result of research in the single sciences. Research in classification can also stimulate scientific development. The specific sciences are not independent of philosophy or of knowledge about and views on classification, 7. Some Concrete Psychological Classifications In the appendixes to this paper are shown some specific classifications for psychology [29] Two dispositions by major German "Handbücher" (Graumann et aI. 1981-; Balmer et aI. 1976-1981) (Appendix 4) The one used in PsycINFO (a database with more than 1.500.000 records by 1998) in 1998 and 1986 (appendix 6a+b) One made by the present author for a union catalog for Danish psychological literature in Denmark (Hjørland 1980; appendix 7) This section will comment on these specific classifications in order to illuminate some general principles or tendencies of the classification of psychology (and about classification in general). These comments are not intended as actual suggestions for revision of those systems. In that case a much more detailed examination and analysis should be made. My first comment is that - as far as I know - designers of such classifications have not made use of research done in LIS (1) The knowledge used is mainly based on actual experience with the material to be classified (which is NOT identical with subject knowledge in, for example, psychology as this is taught at universities, even though such subject knowledge is supposed to be helpful). In addition a very small amount of logical principles may be applied. This raises the important question whether such classifications can benefit from research done III LIS? (And what kind of research that might be helpful). In order to answer this question we should know whether these classifications are fulfilling their aims, or can be criticized. In other words, we should develop some criteria as how to evaluate and improve such classifications. My second comment is that there is - at the most overall level - a similarity in the structures. Such classification almost always starts with the metadisciplinary classes (such as the history and methods of the discipline) then goes to the basic discipline and finally the applied areas. Throughout the classifications there is a tendency to go from the more general to the more specific aspects (we shall return to the important question of what "general" means in relation to psychological phenomena).[30] My third comment is that the disposition of two German handbooks (Graumann et aI. 1981-; Balmer et al. 1976-1981) from the same period illustrates that the same classificatory task can be tackled in very different ways. Of course many differences in specific classifications are always arbitrary, and we cannot draw general conclusions or learn important principles from accidental properties. We have to search for essential characteristics. That means that we have to go from the surface of things (or classifications) to their deeper nature - a principle derived from realist epistemology and in strict opposition to empiricist epistemology. One such accidental property might be that Balmer et a1. (1976-1981) is more populist because it uses the names of the most well-known researchers in psychology as labels for specific volumes and thus also as subject headings. Behind such "accidental" differences between the two German handbooks it is my claim that a more general principle can be demonstrated. Balmer et al. tends to integrate the subject matter of psychology in the different psychological traditions, whereas Graumann et al. tends to focus on knowledge fields in abstraction from theoretic al approaches. In my opinion Balmer's approach is the most organic one, whereas Grauman's is a more positivistic approach to classification. Thus the classification of psychology in these two specific examples can In my opinion confirm the influence of different epistemological theories (related to empiricism and historicism respectively). It confirms that different views of knowledge do influence the way people organize knowledge and may be the only general principle on which to base a theory of classification. My fourth comment concerns the fine-gradedness of the classifications. PsycINFO has 155 different subject headings (omitting those without a verbal description) and is thus one of the most enumerated classifications on this subject (see appendix 6). In 1986 the same database had only 81 classes. The expansion was introduced when the database corrected an old sin of omission not to index monographs. First monographs (including chapters) were indexed in a special bibliography, PsycBOOKS (in print, later only electronic, later again integrated in the PsyclNFO database). We have thus every reason to believe that the current and relative finely graded classification was developed as a way of presenting the records in the printed volumes of PsycBOOKS (which were one-year cumulations opposed to Psychological Abstracts' monthly cumulations). Printed monthly abstract Journals like Psychological Abstracts (1927-) are classified in order to permit the users to scan the table of contents each month. However, when the Abstracts are bound in cumulated volumes in libraries, the organization of the bound volumes remains month by month classified. Retrospective searches for specific subjects can easily be done in the electronic versions, but in my opinion this system does not satisfy the need to browse, the need to have a finely-graded classification of large cumulations. To search information electronically by using descriptors and other forms of access has been characterized as peeking into a room through the key hole. Cumulated bibliographies (as well as other kinds of cumulations, e.g., collected works) can sometimes display a beauty, which stands in contrast to such key­hole feeling. The lack of such cumulated subject bibliographies with finely-graded classifications might add to the feeling of lack of coherence in the discipline. Each researcher loses his impression of contributing specific knowledge to a larger structure. In my opinion psychology needs a much more fine-graded classification than the 155 classes in PsycINFO is this need is to be satisfied. My fifth comment turns this problem the other way round and hypothesizes that this classification is not intended to play a significant role in on-line searching. What specific role does this classification play in electronic retrieval? Even though some records have more than one classification code [31] one characteristic of classifications compared to descriptors is that they tend to provide a structure where each document has one definite place. This allows for another kind of searching behavior which is more like browsing and navigating than keyholing. Classification codes tend to give general knowledge about a concept) whereas descriptor searching tends to identify more, but also much more specific knowledge. In is very surprising that the classification in PsycINFO does not contain any classes for children, adolescents or adults (but one for older people (SH=2860)). This must be seen from the fact that the database some years ago made a retro conversion so that each record was given a new field displaying age groups.[32] I am not sure this was a good solution or whether it does not make navigating unclear. It makes it more difficult to evaluate the classification in its own right. It also gives rise to two different criteria of subdivision in class 2800 and thus to confusion about whether 2820 cognitive development contains cognitive development in older people or not. In my opinion classifications of the kind discussed here do have important functions to play both in online retrieval, in printed cumulative bibliographies and as independent "maps" of a subject domain. However, in information retrieval they have been regarded as the opposite of user-friendly (because the users have to remember the classification codes or to look them up during on-line retrieval, which can be a stressful and complicated situation even without this task). Also classification-based retrieval came to look old-fashioned and ineffective compared to free text searching at the start of modern information retrieval. That may be the main reasons why they are today relatively underdeveloped. My sixth comment will be that the classifications lack the understanding of psychology, which I asked for in section 6 above. I shall give some critical comments based on an analysis of PsycINFO. Approaches to psychology arc almost totally lacking. Only 2140 "History and Systems", 3143 Psychoanalytic theory and a few groups in clinical intervention are explicitly concerned with different theoretical views. However, most people interested in psychology approach the subject with some theoretical favorites or dislikes or views they would like to know more about. A detailed classification of theoretical approaches to psychology is in my opinion mandatory to serve all three purposes plus (especially) all the people who have never thought about how theory-laden all subject domains in psychology actually are! Many of the subject headings in the classification are names of non-psychological disciplines (e.g., statistics, genetics, literature, philosophy, linguistics, and robotics). Many more are names of interdisciplinary areas, in which psychology is only one of many contributors (e.g., gerontology, mass media, marriage & family, and so on). All this adds to a feeling that this is not a classification that reflects a view of psychology, but just a col1ection of loosely related themes without deeper internal connections. This is of course not only - or not primarily - a criticism of the designers of the PsycINFO-classification, but of mainstream psychology as such (as analyzed so brilliantly by Danziger 1997). Even then PsycINFO should make their classification reflect an understanding of psychology to a higher degree. [33] One of the most important things is to analyze the relation between subject areas and approaches. What is what, and what provides the basic rule for placement? In my opinion "2390 Parapsychology" should not be regarded as a subject area in psychology, but as an extreme non-materialistic (or dualistic) assumption, for which reason it should be relocated to "2140 Systems". In the same way different ways of testing in psychology (e.g., neuropsychological testing and educational measurement) should be relocated to the respective areas of psychology because methods should reflect their object, not the other way round. This is again about going from the surface of things to their deeper nature. As written above the main structure is Metadisciplines, Basic Science and Applied Science. Below this level, there is in the Basic Science part a (hidden) substructure going from the general to the specific. The first heading is "2100 (General psychology)"[34]. But what is “General Psychology"? and what is general in psychology? At one time in the history of psychology (or rather by some systems of psychology) it was assumed that basic psychological processes like perception, memory, learning, and emotions were based on physiological principles which were common to all human beings, and therefore general or universal.[35] This is in my opinion an untenable and obsolete theory, but it is the background for the term "general psychology" as covering experimental psychology. So from this position the classifications of psychology are per tradition arranged beginning with the more biologically oriented fields towards the more socio-cultural fields.36 Much more could of course be said of these concrete classifications. One subdiscipline in psychology is "Differential psychology”. This is not represented in PsycINFO, and an important discussion of the nature of this discipline could be done (see Asendorph 1991 for one analysis). Each such analysis teaches us important lessons about psychology with important consequences for the understanding of its cognitive organization. 8. Conclusion I have tried to give a broad outline of psychology and the problems of its classification. My intention has been both to say something important about psychology and to contribute to the methods of classification research as part of information science. This has been done by formulating as many explicit principles and theses as possible. My main assumptions have been that classifications are not neutral tools but reflect a view of the subject domain to be classified. Different views, paradigms or approaches exist in every subject domain, and these views have at the deepest level a strong connection to basic theories in ontology and epistemology. Therefore basic epistemological theories like empiricism, rationalism, historicism, and pragmatism can provide a basis for the classification of knowledge fields. The overall picture of traditional mainstream psychology shows a discipline with immense worldly success, but at the same time a very fragmented discipline without a satisfactory theoretical framework. The main problem for psychology is its tendency to be reduced to either biology or sociology. The most promising theory for a united psychology seems to be "activity theory". This theory has the potential of transforming the psychological system of subdisciplines and the relations between psychology and other disciplines. Research on the classification of knowledge fields must itself be based on ontological and epistemological theories. Here it is my claim that empiristic and rationalistic theories have so far been very dominating, but that the broad family of historically and culturally oriented epistemologies has much to contribute. It is evident that other researchers may continue this work. Very many dissertations may be written about epistemological and psychological theories and their historical development, (indeed this might be a reason why few researchers today dare try to contribute to such problems). In my opinion it is important that sciences do not disappear in fragmented knowledge, but try to understand the major lines in the development, structure and organization of knowledge. Isn't this what research in knowledge organization and classification should be about? Research on classification of knowledge should be relevant for practical purposes like information retrieval. I have tried to outline how research done by using the suggested approach can improve practical library and information services. I do not find that work in the classification field can be done once and for all, but that it should be a continuous activity, which ensures that the value-added information provided by information specialists is of such a quality that the users find it relevant. LIS does have a basic structure of institutions, researchers, journals, and so on. I think we should play a stronger role in classification. Not in monopolizing it but in coordinating research and development. Making our journals relevant for people who seek information about the organization of knowledge. This article is an attempt to do just that. Appendix 1 Empirical Approaches to the Classification of Psychology Empiricism is a philosophy that claims that all knowledge originates from the observations made by individual human beings. All kinds of knowledge established by traditions or inborn in humans are regarded with great skepticism. All knowledge based on experience concerns something particular (isolated); empirical knowledge is therefore fragmented. Empiricism seeks to establish general knowledge through induction made from observed data. For the empiricist there is no necessity in the world, everything that is observed could be different in new observations. The Twentieth century has been much dominated by empiricism (in the form of logical empiricism/logical positivism) and not least in psychology and in information retrieval theory. The prevailing approach to psychology has been behaviorism, which represents an extreme empiristic view of human nature. In spite of its influence, most philosophers of science agree that empiricism/positivism is in very great trouble and has been so at least since about 1950. Even though empiricism and behaviorism are extremely skeptical towards all forms of inborn knowledge or cognitive functioning, intelligence testing has flourished. Among the methods developed to work with statistical data in psychology is a specific classificatory method known as factor analysis. This method has later been exported to many other sciences including information science. Psychologists have tried to define basic categories of intelligence and to classify kinds of human talents and performances. One method has been by using factor analysis of a huge amount of empirical scores in intelligence tests. One well known example is Guilford's "structure-of-intellect" model describing facets of intelligence, which were later expanded to 180 (Guilford 1967 1982). This program has been carried out in a very large scale, but has not been able to maintain the initial optimism with which it was associated, A newer program also building on strictly empirical methods is connectionism where neural networks are taught to categorize input and to react accordingly. This program has also lost much of the initial optimism with which it has been associated. The problems with these kinds of classificatory methods are intimately connected to basic problems in empiricism. Empiricism neglects the fact that every experience does not only depend on the things experienced but also on the organism making the observations. Not even the simplest observations can be madewithout an organism capable of categorizing and interpreting the observations. On the basis of a given set of empirical data many different generalizations can be made. That depends on which attributes are chosen as the most relevant characteristics. Empirical classifications rest on "similarity" or "resemblance". What one should regard as "similar" data is, however, not a question which empiricism itself can answer. In fact it turns into a question concerning the purpose and aim of the classification. Observations are theory-driven, and so are classifications. CONTENT PRODUCTS Units Classes Relations Systems Transformations Implications OPERATIONS Evaluation Convergent production Divergent production Memory retention Memory recording Cognition Guilford's "structure-of-intellect"-model (Guilford 1967) The empirical methods of the classification of literature have especially been used in the bibliometric tradition by using the citation indexes as the database. Concrete examples can be found in Garfield (1976; 1979; 1992; see examples in figures below). The strengths and limitations of this approach in information science is discussed in more detail in Hjørland 1997. Also algorithms in information retrieval are actually used to classify the literature in a database into sets of relevant/non-relevant records. Such techniques must therefore share the same fundamental problems as all methods using purely empiristic methods do. Appendix 2 Rationalistic Approaches to the Classification of Psychology Rationalism is a philosophy which emphasises reasoning and a priori theorizing. Rationalism is - like empiricism - an objectivistic, reductionistic, foundationalistic, ahistorical and apolitical approach. Rationalism analyzes concepts from a logical point of view and tries to organize concepts in one all-embracing structure. In knowledge classification it tries to classify all sciences in one all-embracing structure. This structure is objective, and does not depend on different points of view, purposes or interests. Classification that is based purely on logical division into mutual exclusive and exhaustive classes reflects a rationalistic approach. The facet-analytic tradition in classification research is the most typical representative of rationalism. Theories of language, concepts and thoughts which try to identify an "absolute syntax" or universal laws and principles that do not depend on the context or cultural background of the users are rationalistic by nature. Rationalism finds that the predisposition to realize basic concepts that do not originate from experience must be inborn. It is our inborn way of forming concepts which determines the essential connections between the things we can learn. The rationalistic point of view also presumes that some kind of abstract analysis or fixed procedure could be used to penetrate the surface of documents, thereby revealing their true subjects. Its method is characterized by the tendency to formulate and follow rules and principles. According to rationalism it is possible to organize knowledge in axioms, definitions and theorems in every domain of knowledge. Below is given some illustrations of how to apply the rationalistic method to classify psychological knowledge. A comprehensive critique of the limitations of this approach (and of the concrete examples) will not be given in this appendix. It should, however, be evident that rationalism on the one hand has problems in relating its theoretical approach to empirical reality (it lacks an empirical foundation). On the other hand rationalism has problems in its claim on the nonhistorical character and the disinterestedness of the relationships between concepts. Rationalism presupposes that the principles of division are of an external character, that subject matter is not formed by organic relations of an internal character. The inherent limitations will be evident in the comparison with the other approaches. A Facet Analysis of Psychology In Hjørland (1997) I have argued that the theory of facet analysis in LIS classification developed by Ranganathan, the British "Classification Research Group" and others can be interpreted as a typical rationalistic method for the development of a classification scheme. Mills and Broughton (1978) represents the best attempt to apply this method to developing a classification of psychology. In Hjørland (1988) I designed - inspired by Mills and Broughton (1978) - the following facet model for psychology. A more detailed discussion can be found in Hjørland (1988). Facet 1: Research methods Facet 2: Theoretical orientation Facet 3: Time, place and form Facet 4: Psychological processes Facet 5: Psychobiology Facet 6: Individuals and personality Facet 7: Social and cultural conditions Facet 8: Sphere of application Facet model for the classification of psychology This model can of course be further elaborated. For example, Facet 7 could be subdivided according to a kind of "systems approach" to the social systems, of which the individual is a part, and which influences the individual. Possible levels could be: The individual person The dyad Families Small groups Organizations Communities Countries Cultures These levels can be treated top down or bottom up. The chosen approach is related to epistemological issues regarding "methodological individualism" versus "methodological collectivism" in psychology and the social sciences. Mills and Broughton (1978). IFR Imagination, symbolism, imagery, Bliss Bibliographic Classification. 2. Ed. intuition Class I: Psychology and Psychiatry. Outline IFV Learning and memory IFW Memory I PSYCHOLOGY IG Learning IAA Philosophical Concepts IGM Conditioned learning (as Philosophy AA/ AI) IH Thinking, reasoning, judgment, IAJ Schools of psychology problem solving IAK Psycho-analysis IHT Parapsychology: hypnotism, ESP ... IAR Behaviourism IJ The Subconscious and Unconscious, IE Research and experiment, experimental depth psychology psychology IJK Subconscious, extraconscious IBN Tests, measurement, assessment, lJL Defense mechanisms scales IJP Unconscious, sleep, dreams IC Animal psychology IJV The individual, individual psychology ICC Human psychological processes and IJW Self concept, personal identity attributes, behaviour IK Personality, character, temperament ICD Influences, determinants, IKA W Personal construct theory enVironment IKK Traits, characteristics ICE Stress (general) IKM B Psychoanalytic personality factors: ICE X Physical, physiographical factors Id, Ego ... ICI Biological factors, physiological IKO Differential psychology, individual psychology, psychosomatics differences ICJ Psychological factors (as IC/IX) IKO T Typologies, types of persons ICK Social factors IKQ The sexes, sex psychology ICL B (Attributes of psychological processes) IKS Sex behaviour Norms, variations .... IL Types of persons other than by (Types of processes) sex or age ICM Developmental, psychogenesis (general) (by deprivation, occupation, ICM R Differentiation (general) religion, family membership ... ) ICP Sensation and perception, ILK (By age) Developmental sensory processes psychology ICQ Stimulus and response (general) ILY Age, age groups ICT (Types of response) 1M Children, child psychology ICV Conditional reflexes, conditioning IMM The family, family relations lD Perceptual and motor processes, IMN (Particular ages) Infants, sensorimotor activity adolescents ... lDD Ability, aptitudes, skills, IMU Handicapped persons intelligence IMV Exceptional persons: geniuses ... lDJ Performance, achievement IN Social psychology, social behaviour lDQ Senses: proprioceptive, INL Attitudes ... conformity ... power and someathetic, visual... influence ... IE Motor, psychomotor processes INO Social interaction IEH Motivation, drives, desires INP Interpersonal interaction IEJ Involuntary behaviour, instincts, 10 Communication habits 10P F Signs, symbols lEN Voluntary actions lOR Verbal, language lEV Will, volition, choice, decision 10V Media: audience, information IF Affective psychology: emotions, IP Socialisation feelings IPR Differentiation and stratification, IFG (Types of emotion) roles IFK Cognitive, higher mental processes IPY Psychology of everyday life IFM Associative processes, concept (clothes, appearance ... ) formation IQ Groups IFQ X Ideation IQR Group dynamics IQS J Types of groups 180 IQV IQY IRC IRD IRE IRF IRF RY IRF T IRG IRJ ISB ISF ISG ISH ISP ISR K ISW ISY IUB IUB R IUM IUN IUO N IUP IUP G IUP O IUY IVB IVC IVC Y IVD IVH IVN IVQ IVR IVS Etnopsychology: national, racial, cultural (By types of persons) (As IK/IM) Applied psychology (Applied psychology) Clinical psychology (general) Mental health, hygiene Medical psychology, psychiatry, abnormal psychology, psychopathology Therapeutic environments Hospitals, clinics ... Psychiatric practice, clinical action Diagnosis, systems Types of treatment, therapy Community mental health Physical Drug therapy, pharmacotherapy Psychotherapy Psycho-analysis, analytical psychotherapy Group therapy Brief psycho-therapy Mental disorders, types of disorders Nature of mental illness, aetiology (By cause) Organic, physical causes >'Alternative (preferred in Class H Medicine) (By cause and manifestation) Mental retardation, idiocy, imbecility ... Psychosomatics :?Alternati7)e (preferred in Class H Medicine) (Disorders of psychological processes) (as IA/IQ) Learning disorders Communication disorders, aphasia Autism Psychoses Organic psychoses: toxic ... Functional Affective: mania-depressive ... Paranoia Schizophrenia Neuroses Anxiety neuroses Hysterical neuroses IVT IWB IWD IWX B IWX P IWX W IWY IX IY Phobic neuroses, phobias Personality disorders: psychopaths ... Behavioural disorders, psychopathology Non-psychiatric conditions Types of persons, psychiatric patients (By sex) (as IKW IIKX) (By various characteristics) e.g. Socially deprived ... (as IL) (By age) Children (as 1M) Other applied psychologies >'Alternative (divided like whole classification - e.g. Psychology of law IYS) ::- This is an inverted schedule and filing order of facets & arrays is the reverse of their citation order. ::- Compound classes are built by retroactive synthesis - terms lover in schedule cite first - e.g.: - Personality traits in children = Children - personality traits IMK K; - Group therapy in paranoia = Paranoia - Group therapy IVH SW; - Performance measurement with the mental retarded = Mental retarded - Performance - Measurement IUN DJB N A Dichotomy Classification of Psychology Below is shown a dichotomy classification of psychology. These rationalistic principles of division can be found in many concrete classifications. They have, however, great difficulties in dealing with the internal relationships in the subdisciplines of psychology. They display formal relationships, not organic relations. Human psychology Theoretical psychology "General psychology" "Individual psychology" Adult psychology Higher psychological processes Cognitive psychology Experimental psychology Positivistic psychology Non-positivistic psychology Non-experimental psychology Emotional & motivational psychology Lower psychological processes Child & adolescent psychology Social psychology Cultural psychology Applied psychology Animal psychology Appendix 3 Major Theoretical Approaches in the History of Modern Psychology Psychophysics/Structuralism (1879-1920) Founded by Ernest Weber (1795-1878), Gustav Theodor Fechner (1801-1887) and Wilhelm Wundt (1832-1920). Wundt founded the first psychological laboratory in Leipzig 1879. In America Edward Bradford Titchener (1867-1927) saw himself as a true successor of Wundt (but in fact he altered Wundt's views dramatically). The term "structuralism" was coined by Titchener (not to be confused with the later structuralism inspired by the linguist Ferdinand de Saussure, a movement that also Jean Piaget felt himself associated with). The influence of structuralism declined with the breakthrough of behaviorism about 1913 and it almost disappeared with the death of Titchener 1927. Functionalism/Pragmatism (1896-1930) Pragmatism as philosophy was founded by Charles Sanders Pierce (1839- 1914). It was supported and continued by William James (1842-1910) and John Dewey (1859-1952). James and Dewey developed the pragmatic approach in psychology. Pragmatism was in the USA an alternative approach to Titchner's "Structuralism" from about 1896 until it was gradually replaced by behaviorism from about 1913 to 1930. It is closely related to the functionalistic school in Chicago at the end of the nineteenth century and influenced by John Dewey and James R. Angell. However, pragmatic and functionalistic approaches have influenced psychology outside the explicit schools from Charles Darwin and until this day. Today we see a revival or "neo-pragmatic" tendency. Schultz & Schultz (1996) mention James, Hall, Angell, Cattell, Woodworth & Carr under this heading. They find that functionalism disappeared about 1950. Psychoanalysis/Depth Psychology (1895-) Founded by Sigmund Freud (1856-1939) about 1895. Has gradually influenced academic psychology. It has divided itself in a large number of competing schools. According to Schultz and Schultz (1996, inside cover) it disappeared about 1975. In my opinion, however, it is still an important approach in psychology (see also appendix 4). (The term "psychoanalysis" does not include the psychology of C. G. Jung. The term "Depth psychology" is suggested as a generic term including psychoanalysis, the analytic psychology of Jung, and others). Behaviorism (1913-1965) John B. Watson's "manifest" for a behaviorist psychology from 1913 can be pointed out as a formal establishment of the behaviorist movement, even such persons as Pavlov and Thorndike had contributed essentially at an earlier time. About 1965 behaviorism seemed to be succeeded by cognitivism. However, essential characteristics of behaviorism have, since about 1930 and until now, influenced psychology according to Danziger (1997). This is a broad, eclectic, implicit "variable psychology". According to Danziger all modern psychology has adapted the basic views of behaviorism, which has become "the language of psychology" even among psychologists who do not regard themselves as behaviorists. Byrne (1995, p. 132) writes: "Introductory texts in the philosophy of mind often begin with a discussion of behaviorism, presented as one of the few theories of mind that have been conclusively refuted. But matters are not that simple: behaviorism, in one form or another, is still alive and kicking". Phenomenological psychology (1912-1940) Philosophically related to phenomenology as founded year 1900 by Edmund Husserl (1859-1938), but in experimental psychology an approach related to Gestalt psychology. The Danish psychologist Edgar Rubin (1876-1951) wrote in 1915 a famous book about visual perception in which he described the figure and ground phenomena. In Europe phenomenology remained an important philosophy until the start of World War II (1940), where it was forced out by analytical philosophy. Since then phenomenological psychology has internationally remained a small enclave. In Copenhagen the influence of phenomenological psychology is still perceptible In the USA "Journal of Phenomenological Psychology" was founded in 1970. Today the influence of phenomenology tends to increase. Gestalt psychology (1912-1940) Gestalt psychology is rooted in phenomenological philosophy (Franz Brentano, 1838-1917 and Edmund Husserl). The term "gestalt quality" was first coined by C. von Ehrenfels in 1890 (a student of Brentano). It became a formal school of psychology in 1912 when M. Wertheimer, Wolfgang Kohler and Kurt Koffka studied apparent movements using stroboscopic experiments. It subsequently headquartered in Berlin. Humanistic Psychology (1962-) Humanistic psychology was established as an approach in modern psychology with an independent organization and journals from 1962. ("Association for Humanistic Psychology" and "Journal of Humanistic Psychology"). Abraham Maslow (1908-1970) belongs to the pioneers of this movement. Humanistic psychology was founded as "the third force", in explicit dissatisfaction with both behaviorism and psychoanalysis (especially the deterministic view of human nature). It based itself on principles from the philosophy of humanism (which goes back to the latter half of the Fifteenth century). The history of psychology might be reinterpreted from a humanist point of view. It is rooted in the Renaissance and in a continental European tradition with names such as Gottfried Wilhelm Leibniz (1646-1716), Immanuel Kant (1724-1804), Søren Aabye Kierkegaard (1813-1855), Wilhelm Dilthey (1833-1911), Edmund Husserl (1859-1938), William James (1842- 1910), William Stern (1871-1938), Carl Rogers (1902-1986), Rollo May (1909-), Abraham Maslow (1908-1970), and Gordon W. Allport (1897-1967). According to Schultz & Schultz (1996, inside cover) humanistic psychology disappeared as an approach in psychology about 1985. In my opinion, however, humanistic psychology (including existentialism) is still influential in psychology. Genetic Epistemology (1960-1990) Founded by Jean Piaget (1896-1980). The first books by Piaget were published in the 1920'ties, but his international influence (especially in the USA) became dominant about 1960. It culminated with the death of Piaget in 1980. Since then Piaget's influence has decreased. An important "neo-piagetian" enclave is still influential. Cognitivism (1965-) Cognitivism is an approach influenced by information theory, cybernetics, and systems theory (developed around 1948). As a starting point the year 1956 may be mentioned. This year J crome Bruner published "A Study of Thinking" and Chomsky's "Logical Structure of Linguistic Theory" circulated in a preliminary edition. This approach developed very forcefully, and was from 1965 the dominant view in American and International psychology. One of the pioneers, Herbert A. Simon received the Nobel price (1978 in Economics). From the late 1980s this approach has been met with increasing criticism, and its influence has markedly decreased. Among the critics are also pioneers as Jerome Bruner and Ulrich Neisser. Cultural historical psychology / Activity theory (1990-) Founded as a school of psychology in Russia by Lev. S. Vygotsky (1896- 1934), A. N. Leontjev (1903-1979), and A. R. Luria (1902-1977). It was suppressed in the Stalinist period but regained influence in the Soviet Union during the thaw. Later it first influenced European and then American psychology. In the 1990s it has become a mainstream in international psychology. It is related to American pragmatism. Poststructural psychology / Discursive psychology / Postmodern psychology (1990-) [Other table appendixes] Notes 1. This paper is not about the psychology of classification, which is a major subject in cognitive psychology. 2. As a contribution to LIS this article could be seen as an example of what I have earlier named "domain analysis" (Hjørland and Albrechtsen 1995). This paper concentrates on just one aspect of such an analysis: the classification of a subject area. There have been other methodological contributors to classification of knowledge fields, mainly in the facet analytical tradition (e.g., Mills 1957; Vickery 1960). 3. Jørgensen (1963) is another exception from this general tendency. In this work Jørgensen suggested a very scholarly classification of psychological phenomena based on a combination of behavioral, phenomenological, and physiological criteria. Jørgensen was both a leading person in the movement of logical positivism and also a Marxist. In my opinion these two influences can both be traced in Jørgensen's original definition and cbssification of psychology, and they are in conflict (in a way postulated by me, but not clear to himself). As a logical positivist/empiricist Jørgensen could only accept things that can be observed. Therefore he defined the human psyche by observational criteria. As a Marxist he was influenced by scientific realism and looked after mechanisms behind what is observed. It is, however, outside the scope of this article to discuss his proposal for the classification of psychology. Jørgensen's interest in classification in a period of time where the general interest in classification was low, can partly be explained by the fact that he was a philosopher, and not a psychologist. There was (and still is) a certain division of labor between psychology as an empirical science and (mental) philosophy as a theoretical science. 4. Whitley (1984) is an example of an important sociological contribution. 5. It is outside the scope of this article to discuss the concept of classification itself. However, two recent contributions should be mentioned: Adams and Adams (1991) differentiate between classification and taxonomy. They see a taxonomy as a special kind of classification with specific hierarchical characteristics, a classification in which smaller and more specific classes or taxa (singular: taxon) are grouped to bigger and more general ditto. A typology they conceive as a form of classification which is specially designed with the aim of sorting the elements in mutually exclusive categories called types. Jacob (1991; 1994) finds that categorization is the fundamental cognitive tool that facilitates the organization, storage and retrieval of information. Traditional classification is an arbitrary and artificial tool that is used to structure a specific knowledge domain while ensuring consistency and stability of meaning. Because a classificatory structure serves to identify relationships between entities and to set the boundaries for a specific area of inquiry, it establishes a worldview that limits the recognition of similarities. To encourage dialogue across disciplinary boundaries, the specificity of meaning inherent in the classificatory structure must be replaced by more general meanings that function across disciplines. 6. I consider classifications made without any explicit methodology, just based on the view or horizon of the persons who are doing the classification to be "subjective" because they do not apply the public knowledge. Therefore they tend to reflect specific persons view (or perhaps a widespread ideology) rather than the public or scientific view revealed by scholarly/scientific studies. I also regard such studies as "pragmatic" in one sense of this word. It is, however, very important not to confuse this meaning of "pragmatism" with the view developed by pragmatic philosophers, such as Charles Sanders Peirce or John Dewey around the beginning of the 20th century. 7. The most influential classification of mental disorders is --> Diagnostic and Statistical Manual of Mental Disorders (DSM) published by American Psychiatric Association. Its 4th edition (DSM-IV) is from 1994. This system is widely recognized and debated in the literature (214 book reviews in Social Sciences Citation Index and 10.692 entries in the PsyclNFO database). This system is mainly based on an empirical epistemology and has been discussed by people informed by other epistemologies. 8. However, it is done in different ways, and epistemological trends in the sciences do affect the use of classificatory methods, and even the attitude towards the relevance of classifying at all. According to Kuiken, Wild and Schopflocher [11] positivism has had negative influence on classification in the sciences. 9. Classification theory may find itself in a situation very similar to that of lexicography. Eco (1984, p. 68) shows how the idea of a dictionary runs into theoretical difficulties: "The tree of genera and species, the tree of substances, blows up in a dust of differentiae, in a turmoil of infinite accidents, in a nonhierarchial network of quality. The dictionary is dissolved into a potentially unordered and unrestricted galaxy of pieces of world knowledge. The dictionary thus becomes an encyclopedia, because it was in fact a disguised encyclopedia" (emphasis in original). On p. 84 he writes: "Such a notion of encyclopedia does not deny the existence of structured knowledge; it only suggests that such a knowledge cannot be recognized and organized as a global system; it provides only "local" cultural organizations; every attempt to recognize these local organizations as unique and "global" - ignoring their partiality - produces an ideological bias. The Porphyrian tree tried to tame the labyrinth. It did not succeed because it could not, but many contemporary theories of language are still trying to revive this impossible dream" (emphasis in original). And, one could add that almost all contemporary theories on information retrieval and knowledge organization - in building on empirism or rationalism - are also trying to revive the same impossible dream. 10. About the etymology of the term "psychology" see Lapointe (1973) and Vandekemp (1980; 1983). 11. Many different terms can be used as a label to describe the same subject area. Examples are "psychology", "the study of the soul", "the science of consciousness", "psychoanalysis", "the science of behavior", "cognitive science", "the science of the mental", "the study of subjectivity", "persanology", or "microsociology". If they are used about the same subject area, they are semantically synonymous. However, the definition of psychology is really difficult because it is not given that they refer to the same knowledge field. They represent more or less different or identical views on man, which imply more or less different of identical subject dom;ains. The question about the synonymity of those terms is therefore not objective or a priory, but a theoretical question depending on the development of the theories about psychology. Also the term "psychology" has not one meaning but many meanings. The use of the term is theory laden. For example, some users of the term psychology are more related to the study of physiological processes, while others are more related to the study of personality processes from a more sociological point of view. The philosopher of science Dudley Shapere (1984) writes about "domains": "Although in more primitive stages of science (or, perhaps better, of what will become a science), obvious sensory similarities or general presuppositions usually determine whether certain items of experience will be considered as forming a body or domain, this less and less true as science progresses (or, as one might say, as it becomes more unambiguously scientific). As part of the growing sophistication of science, such associations of items are subjected to criticism, and are often revised on the basis of considerations, which are far from obvious and na"ive. Differences, which seemed to distinguish items from one another are concluded to be superficial; similarities which were previously unrecognized, or, if recognized, considered superficial, become fundamental. Conversely, similarities, which formerly served as bases for association of items come to be considered superficial, and the items formerly associated are no longer, and form independent groupings or come to be associated with other groups. The items themselves often, in the process, come to be redescribed, often, for scientific purposes, in very unfamiliar ways" (Shapere 1984, 323). This is important because it implies that the classification of knowledge domains cannot be done independently of the claims or views of the knowledge in the domain: "as science proceeds, the connection between knowledge-claims, domain groupings, and descriptions {and often naming} tend to become tighter and tighter" (Shapere 1984, 324). According to Shapere, domains are bodies of subjectmatter that have become delineated by the way in which the history of scientific methodology, theory and discovery has developed over many centuries, and even at a mature stage they are in a process of constant refinement and occasional wholesale reordering and unification. There seems to be a long term tendency to develop more comprehensive theories which unite domains of subject matter. The definition of psychology therefore cannot be made from etymological studies alone, but requires a theory about the domain which the term is meant to cover. Such a theory should be able to explain what is to be included and what is to be excluded from psychology. A theory about the classification of psychology must develop a metatheory of what is by different views included and excluded from psychology. In order to argue what should be excluded from psychology the theory must be able to explain where these phenomena belong in the universe of knowledge. Its theory about the domain of psychology must not only imply a classification of what is regarded as psychological phenomena according to the theory itself, but it should also be able to classify phenomena which other theories regard as psychological phenomena. In my view, the ontological and epistemological theories are the basic theories, which can be used as psychological meta theories to produce the basic outline of a classification of psychological knowledge. This view is, however, itself a theory which is opposed to the anti-metaphysicalism of empiricism and positivism. 12. Aristotle was of the opinion that the book "De anima" treated the most important of all subjects. 13. The psychological systematization presupposes some philosophical clarifications of the nature of psychological phenomena and the psycho- physical problem. Do psychological phenomena only exist in man (called "anthropopsychism") as, for example, Descartes thought, or - to take the opposite extreme theory - is the whole world, including non-living objects, endowed with spirit (called "panpsychism") or do psychological phenomena only exist in living beings (called "biopsychism")? Or do they only exist in certain higher animals? 111e Psycho-Physical Problem ('111e Mind-Body Problem) MONISTIC DUALISTIC PLURALISTIC THEORIES THEORIES THEORIES Materialistic theories The Theory of In· (Karl Marx, John B. teraction \Y/atson?-) (Descartes) Spiritualistic or ide- The Theory of Paralistic theories (Berkeley) The theory of Identity (Spinoza) allelism (Leibnitz) The Epiphenomenological Theory (P,vlov) Karl Popper & J.Ecdes "Three Worlds" Karl Pribram (According to Tolman (1992) John B. Watson and the behaviorism he founded were not materialistic, but phenomenologistic and hence philosophically idealistic. This view is connected to an important distinction between empiricism/positivism on the one side and scientific realism on the other. Empiricism/positivism tries to base all knowledge on sense impressions and to avoid any "metaphysical" assumptions. Sense impressions are, however, purely subjective, and without language to ensure the objectivity of observations empiricism cannot avoid the idealistic trap.) Immanuel Kant put forward a classical categorization of psychological phenomena in his book "Kritik der Urtheilskraft" (1790), in which he divided psychical phenomena into "Erkenntnisvermogen" (Cognition), "Gefühl der Lust und Unlust" (Emotion) and "Begehrungsvermogen" (Motivation). Different psychological theories have attempted to reduce these phenomena to one. Different psychological approaches tend to overestimate the importance of some processes at the expense of others. For example, psychoanalysis tends to overestimate motivation, while cognitivism tends to overestimate the role of cognitive processes. It is also important to notice that in the "classical psychology" (e.g by Wundt) cognition, emotion and motivation were regarded as a microcosmos, a world closed on itself, apan from bodily phenomena (based on the dualism of Decanes). It was thought that self-observation (introspection) could be used to uncover the basic elements of consciousness, the special psychical "matter". Leontiev writes: "The classical rationalistic psychology was as an attempt to uncover our psychical world, the world of our ideas, our emotions and our thought, and here find the laws which express its nature. They believed that they only had to observe and come to terms with the fluctuating and unclear subjective psychical experiences in a sufficiently rational and careful way in order to find the laws and causes which govern the "small world" of our consciousness. This could be done in the same way as the observation of the stars had let humankind to the discoveries of the laws that govern the physical universe, the "big world". This psychological idea was never realized, and in fact can never be realized. The world of our consciousness has no resemblance with the world of the planets. You cannot consider consciousness a closed existence in itself and search for independent connections in it. They do not exist. If you talk about "spiritual movements" or "spiritual forces" then you are just using simple metaphors. The expressions of consciousness are always connected to one thing or another and reflect something. Therefore there does neither exist a physics of conscious expressions nor a mathematics of ideas or a geometrical or pure form of the spirit. The expressions of our consciousness are always determinated by the external reality of objects, which is reflected in it" (Leontjev 1977, 86, my translation). It is not until the development of the functionalistic and behaviouristic approach that psychology relates consciousness to behavior and acts to the body and to the adaptation of the individual to his/her environment (cf. Jarosjevski 1980). This is one reason why categories like behavior, acting and activity are missing in the classical tripartirion (cognition/ emotion/motivation). 14. To say that psychology became split is slightly wrong, because a split between, for example, biological and sociological conceptions of psychological phenomena is already visible in the works of Aristotle. 15. The reason that is seldom explicated very well that different approaches to psychology imply different views of subject matter and classification is that each approach tends to consider itself the only real scientific approach, and that psychology as a new empirical science disconnected from philosophy tended to make all such questions merely empirical questions. 16. The formal establishment of psychology as a science was done in universities. First in Leipzig, and very soon in all major universities in the rest of the world. When a discipline is established at a university it becomes both an area of research and of teaching. The kind of psychology, which is taught at universities is called "academic psychology". Psychoanalysis (founded by Sigmund Freud around 1900) was not part of academic psychology for a very long time, and is still mainly a non- academic field of knowledge with strong connections to applied psychotherapy. One indication of the historical impact of a field of knowledge such as psychoanalysis on academia is the number of dissertations written with that approach. In Denmark a dissertation by Lise O/stergaard from 1959 is the first one that documents the acceptance of the psychoanalytic viewpoint in academic psychology. Another indication of the relationship between academic and non-academic knowledge is the system of journals. Psychoanalytic contributions can, for example, be found in academic psychological journals, as well as in academic medical journals (as well as other disciplines such as philosophy and literature). The core set of psychoanalytic journals are, however, not affiliated with any academic discipline, but are connected to psychoanalytic societies in different countries. It is a characteristic of the relation between academic mainstream and psychoanalysis that the American Psychological Association did not establish a psychoanalytically oriented journal until 1984 (Psychoanalytic Psychology, sponsor: American Psychological Association, Division of Psychoanalysis). By the way it is only sponsored, not like many other journals published, by the American Psychological Association (which represents members educated in psychology at the universities). 17. Wundt himself - physician and physiologist by education - also regarded experimental psychology as a very limited area of psychology. He wrote a major work "Volkerpsychologie" (vol. 1-10) based on more humanistic methods, and he did not want that psychology should become "an independent discipline", but he wanted it as part of philosophy with close connections to other humanistic disciplines. He had, however, started a process which he himself could not control. Psychology became an independent field of study partly as a result of the competition for jobs, and partly because of the development of antipsychologistic tendencies in philosophy itself (see Danziger 1990, 41). 18. [Danziger underlines the discipline's tendency to annex new areas such as educational psychology, which existed before the formal establishing of psychology as a science. One may ask: Where did these areas come from? How did they arise as parts of human discourses? How did man become an object of research? In asking such questions, the research done by Michel Foucault is extremely relevant (e.g., Foucault 1967; 1970; 1972; 1973; 1976; 1977; 1978).] 19. Danziger has here a note 1: "More recently, the category of "cognition" has played a similar role. The idea of abstract laws uniting many domains of psychological functioning, irrespective of content, reappeared in the form of the category of "cognition" just when "learning" could no longer play this role effectively. But this development falls outside the time period of this book" (Danziger 1997, 108). 20. In an earlier book Danziger reveals a slightly different view on the possibility of establishing realistic psychological categories: "One thing that has emerged from rhe historical analysis of changes in investigative practices is the importance of the social alliances formed by the discipline as a whole and by subgroups within the discipline. These alliances tended to favor and to maintain certain practices over others. An escape from methodological solipsism is likely to depend on the variety of alliances that members of the discipline manage to forge. Working relationships and alliances are formed not only with other professional groups. We have seen that the social context in which psychological knowledge products are ultimately applied have an effect on the kind of knowledge product for which there is a demand and hence on the practices that must be used to produce it. In that connection it is difficult to ignore the dominant role administrative useful knowledge has played in the past. As long as that state of affairs persisted, the kind of reality to which much of psychological investigation provided access was an administratively created reality. Even when it was not directly tied to actual administrative reality, this kind of research created its own replica of such a reality, as in early American personality research. The administrative context of application cast its shadow over significant parts of the comext of investigation, which did not help to broaden the latter's access to the real world outside such contexts. The prospects of that happening would seem to depend on the extension of disciplinary alliances to groups of people who are more interested in psychological knowledge as a possible factor in their own emancipation than as a factor in their management and control of others. The worldly success of modern psychology was built on a narrow social basis. That entailed a very considerable narrowing of epistemic access to the variety of psychological realities. Critical analysis can give us some knowledge of that which has been excluded - in other words, knowledge that has emerged in different social contexts. The receptivity of discipline to such knowledge, however, would seem to be tied to changes in its social and cultural commitments" (Danziger 1990, 197). 21. If this should not be totally clear from the citations from Danziger (1997) I strongly recommend that everybody should read the whole book. 22. The implication of what is said here is that I disagree with the organizational principle put forward by the German Enzyklopädie der Psychologie 1-88 (see Appendix 2). 23. It may not seem an easy approach. But who has said that doing scientific and scholarly works and making real contributions to knowledge should be an easy task? Researchers are expected to educate themselves so that they are qualified to attack the problems with the kind of methods required by the problems (and not vice versa)! 24. One research object in psychology is, for example, emotions such as anxiety. The basic approaches in psychology towards anxiety are very different. Behaviorism approaches anxiety as a kind of conditioning or as a learned behavioral pattern, ethology approaches anxiety as a kind of instinctive behavior developed phylogenetically, psychopharmachology and biologically oriented psychiatry look at anxiety as a defect in the brain that might be cured by medical treatment. Psychoanalysis sees anxiety as determined by personality structure and forgotten childhood experiences, cognitivism sees anxiety as a kind of reasoning with negative consequences for the self. Existentialism may look at anxiety from a religious perspective or from the perspective of the meaning of human life, social constructivism looks at anxiety as a repertoire of different culturally shaped ways at behaving, Marxism may look at anxiety as the individual's response to a real threat in a capiralist society, and so on. In my view all these approaches have something important to say about anxiety. Their generalizations are, however, problematic. The important thing to realize is that anxiety is a biological, a cultural and an individual phenomenon. It is important not to reduce anxiety to only one aspect. 25. This is further affirmed by an analysis of psychoanalysis put forward by Schultz (1988), who sees the solid core of psychoanalysis as the realistic and cultural study of human symbolic cognition. His main criticism of Sigmund Freud is that Freud "built his ship on the wrong yard". Schultz bases his psychology much on Freud and Leontiev (e.g., Leontiev 1977) and finds that Leontiev and his tradition has too little to say about a very central issue in psychology: the symbolic activities of human beings (in which psychoanalysis and also phenomenology are strong). However, Freud's psychology should be based on a realistic epistemology, Psychoanalysis is then mainly the realistic study of human symbolic cognition. Schultz also proposes a classification of psychology as shown below. A Realist View on the Classification of Psychology (Modified after Schultz 1988, 270) Science of Life (Study of life activity) Biology Psychology ([he science of sim- (Science of the cognitive activity) pIe life activity) Psychology of Personality Ethology (The Science of person activity) (Science of subject-activity) Dynamic Psychology of psychology thinking or (The science reflection of symbolic (The Science cognition) of categorical cognition) 26. Kuhn's view did not, however, influence the basic thinking about users and cognition in cognitive psychology (or in information science) for a long time. The influence of historicism in this discipline was in particular introduced by the criticism which Winograd and Flores (1986) posed to traditional "rationalism" (in the form of "artificial intelligence" and cognitivism). This implies a new view of users as social and cultural beings and of a more sociologicalepistemological view on information seeking. 27. To identify a pragmatic understanding of psychology with the understanding of psychology as an applied science would be a rather vulgar interpretation of pragmatism, but it is, however, more true for some pragmatist researchers than for others. 28. An important issue in all classifications is what are the "elements" or units to be classified. Very few psychological theories use this word or have an explicit definition of this concept. Fechner did, and Mammen (1993; 1998) has made a new proposal which I find extremely important. The meaning of a thing, say "an axe", is not just a result of some sensory properties, but of a standardized human practice and a tradition for producing this tool. The sense of a thing, for example "an axe" is the result of my personal relation to a concrete axe or to axes in general. Humans are different from animals in that they not only sense things because of their universal attributes. Man has a special ability to perceive the numerical identity of objects and thus to follow the history of objects (which may be lost in some cases of schizophrenia). The unit of the specific human psychology is thus a subject (1) in relation to an object (e.g. a tree), mediated by a tool (e.g., an axe) this relatedness is influenced by other human subjects (at least one, subject 2). According to Mammen, these are the necessary, basic elements in human psychology. This view changes fundamentally the traditional classification of psychology: "To conclude: One advantage of using the same basic "units of analysis" in all these different cases is that some of the traditional distinctions between cognitive and "motivational" psychology, or psychology of personality, vanish" (Mammen 1993, 41). 29. Among important classification systems not discussed in this paper is Medical Subject Headings. Tree structures published currently by National Library of Medicine in Washington. 30. This overall structure might be an exception from the rule that such systems are not built on principles developed in or conununicated by LIS. 31. How is this scheme of classification actually applied? Are each of the records in the PsycINFO database given one and only one classification code/subject heading? A simple analysis in the database shows that the number of records with a subject heading/classification code (SH=) is 1.528.542 or equal to the total number of records in the database. Also the number of records printed in 1996 equals the number of records with a subject heading from this year, showing that each record has at least one subject heading. S1 56437 PY = 1996 S2 1528542 SH = ? S3 56437 S1 AND S2 However, some records must have more than one classification code because the truncated set of classification codes beginning with the numbers 2, 3 and 4 do have an overlap (92.393 records both have a classification code beginning with the number 2 and the number 3). At least 102.173 out of 1.528.542 records or 6.7% have more than one classification code (this small investigation cannot reveal duplicates beginning with the same number). The fact that each record can have more than one classification code ascribed to it means that the function of the classification system is not sharply defined in relation to the descriptor system (the thesaurus) where the norm is to apply a number of different descriptors to each record. Set Items Description S1 0 5H=0? S2 0 SH=1? S3 668637 SH=2? S4 932331 SH=3? S5 29747 SH=4) S6 0 SH=5? S7 0 SH=6? S8 0 SH=7? S9 0 SH=8? S10 0 SH=9? S11 1528542 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 S12 92393 S3 AND S4 S13 3430 S3 AND S5 S14 6350 S4 AND S5 S15 10217.1 S12 OR S13 OR S14 S16 0 S12 AND S13 AND S14 32. This retroconversion to provide each record with a new field displaying age groups was done mechanically, in such a way that all records which did not have descriptors for children were defined as concerning adults. This is not an elegant solution, because many writings (such as this paper) are neither about children nor about adults. 33. The coverage of PsycINFO is not only psychology, but also related disciplines, for example, psychiatry or educational research. This represents a clear dilemma. We have many databases covering these fields and from the user's perspective it is better that each database has a deeper coverage than having great overlap between databases at the same time as having great gaps in the coverage of the core disciplinary literature. (This may not be the case from a commercial perspective.) Interdisciplinarity in coverage also makes it harder to make an adequate subject representation (e.g., classification) and to provide the users with a clear picture of the size, extent, structure and content of the current psychological knowledge. 34. I disregard the fact that 2100 is only used for metadisciplines. This is a thing that has happened in the history of the system, and which, in my opinion, is a fault. 35. Poulsen (1984, 169-) writes: "If scientific psychology is to be a general psychology in the meaning that it is only concerned with the aspects of human psychology which are common to all people and characteristic for the human species, this will have unacceptable consequences. Psychology will be a science which denies caring about psychological functions in the specific forms in which they exist in specific societies and in specific individuals. [...] scientific psychology becomes irrelevant and relevant psychology non-scientific" (my translation). 36. What does the term "general" mean. This is not a trivial question, but a complicated philosophical one. I once made a major analysis of the meaning of this term when it was used as adjective in relation to the name of a science (e.g. general psychology, general physiology, or general sociology). Part of this is published in Hjørland (1982), but some parts remain unpublished. My major result was that the meaning of the terms varies with the philosophical view in the sciences. Within the dominant empiristic view a general discipline is seen reductionistic, as the part of the discipline which was considered foundational. Thus general psychology meant physiological psychology, general physiology meant chemical physiology, and general sociology meant psychosociology. Within a non-reductionistic view, however, a general discipline means the part of the discipline in which its general principles are formulated. At the University of Copenhagen (and elsewhere?) the term "General psychology" has been considered a subdiscipline of psychology. As such it is not identical with experimental psychology or physiologically oriented psychology bur is supposed to formulate the basic principles of psychology spanning both biological and sociological views. Leontievs' (1977) general psychology is an influential example of this last meaning. References and Selected Bibliography The bibliography contains many references not mentioned in the text. All references have in this ieko edition been updated to current standard in knowledge organization. Reference added or changed in the ieko edition are marked with *. (In the 1998 edition a CD-ROM version of encyclopedia britannica had been used; many of these references have not been possible to verify today, but the printed and online editions of this encyclopedia contain valuable articles by other authors). Adams, Charles Joseph. 1999. “Principles of Classification (Normative, Geographic, Ethnographic, Linguistic, Philosophical, Phenomenological and Other”. In: Religions, the Study and Classification of. Britannica®CD 99 Mllitimedia Edition. Chicago, II.: Encyclopaedia Britannica, Inc., © 1994-1999. Adams, William Y. and Ernest W. Adams. 1991. Archaeological Typology and Practical Reality. A Dialectical Approach to Artifact Classification and Sorting. Cambridge, UK: Cambridge University Press. Allport, Gordon. W. 1937. Personality: A Psychological Interpretation. New York: Holt, Rinehart & Winston. Andersen, Niels O., Jørgen Arnberg, Jens Thoring Hansen and Lisbeth Marcussen. 1983. Erkendelsesteori og Vidensklassifikation. Copenhagen: Royal School of Library and Information Science. (Danmarks Biblioteksskoles A4-serie nr. 23) Anderson, John Robert. 1983. Architecture of Cognition. Cambridge, MA: Harvard University Press. Andkjær Olsen, Ole and Simo Køppe. 1996. Psykoanalysen efter Freud. Vol. 1-2. Copenhagen : Gyldendal. *Arnfield, A. John. 2003. “Climate Classifcation”. In The New Encyclopedia Britannica. 15 th Edition. Chicago: Encyclopedia Britannica, vol. 16: 497-504; refs. p. 522. Asendorpf, Jens. 1991. Die differentielle Sichtweise in der Psychologie. Göttingen: Verlag für Psychologie. Baldwin, J. Mark. 1901. ”A Scheme of Classification for Psychology”. Princton Contributions to Psycbology 3, no. 1: 16-9. Balmer, Heinrich et al. (Ed.). 1976-1979. Die Psychologie des 20. Jahrhunderts. Band 1-15. Zurich: Kindler. Battan, L. J. (1999). “Approaches to climatic classification. Genetic classifications. Empiric classifications”. In: Climate and weather. Britannica ® CD 99 Mllitimedza Edition. Chicago, II.: Encyclopaedia Britannica, Inc., © 1994-1999 (see *Arnfield 2003). *Bayne, Helen and Ilse Bry. 1954. ”Problems and Projects in the Bibliography of Psychiatry and Psychology”. Libri: International Journal of Libraries and Information Studies 3: 363-87. Braun, Claude M. J. and Jacinthe M.C. Baribeau. 1984. “The Classification of Psychology among the Sciences from Francis Bacon to Boniface Kedrov”. Journal of Mind and Behavior 5, no. 3: 245-59. Brocke, Burkhard. 1980. “Wissenschaftstheoretische Grundlagenprobleme der Angewandten Psychologie: Das Abgrenzungs-, Konstituenten- und Fundierbarkeitsproblem”. Zeitschrift für Sozialpsychologie 11, no. 4: 207-224. Bry, Ilse and Lois Afflerbach. 1968. “In Search of an Organizing Principle for the Behavioral Science Literature”. Community Mental Health Journal 4, no. 1: 75-84. Burne, Alex. 1995. “Behaviourism”. In A Companion to the Philosophy of Mind, edited by Samuel D. Guttenplan Oxford, England: Blackwell Publishers, 132-40. (Blackwell Companions to Philosophy). Chafetz, Michael D. 1986. “Taxonomy in Psychology: Looking for Subatomic Units”. The Journal of Psychology 120, no. 2: 121-35. Danziger, Kurt. 1990. Constructing the Subject: Historical Origins of Psychological Research. Cambridge: Cambridge University Press. Danziger, Kurt. 1997. Naming the Mind. How Psychology Found its Language. London: SAGE. Dewey, John. 1948. Reconstruction in Philosophy. Enlarged edition. New York: Beacon. (Original work published 1920). Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4th ed. (1994). Washington, D. C.: American Psychiatric Association. Dreger, Ralph Mason. 1968. “Aristotle, Linnaeus, and Lewin, or the Place of Classification in the Evaluative Therapeutic Process”. Journal of General Psychology 78, no. 1: 41-59. Dubois, Danièle. 1994. “Identity and Autonomy of Psychology in Cognitive Sciences: Some Remarks from Language Processing and Knowledge Representation.” World Futures 42, nos. 1-2: 71-8. Eco, Umberto. 1984. Semiotics and the Philosophy of Language. London: The MacMillan Press. Fiedler, Frank. 1990. “Klassifikation der Wissenschaften”. In Europaische Enzyklopadie zu Philosophie und Wissenschaften Band 1·4, edited by Hans Jörn Sandkühler. Hamburg: Felix Meiner Verlag, Band 2: 812-15. Fiske, Donald W. 1991. “Macropsychology and Micropsychology: Natural Categories and Natural Kinds”. In: Improving Inquiry in Social Science: A Volume in Honor of Lee J. Cronbach. Edited by Richard. E. Snow and David E. Wiley. Hillsdale, NJ, US: Lawrence Erlbaum Associates, 61-74. Foucault, Michel. 1967. Madness and Civilization: A History of Insanity in the Age of Reason. London: Tavistock Publications. (Translated from: Folie et Déraison: Histoire de la folie à l'âge classique. Paris: Plon, 1961). Foucault, Michel. 1970. The Order of Things: An Archaeology of the Human Sciences. London: Tavistock Publications. (Translated from Les mots et Ies choses une archeologie des sciences humaines. Paris, Gallimard, 1966). Foucault, Michel. 1972. Archaeology of Knowledge. New York, Pantheon Books. (Translated from the French by A. M. Sheridan Smith from: L’Archéologie du savoir. Paris: Gallimard, 1969). Foucault, Michel. 1973. The Birth of the Clinic: An Archaeology of Medical Perception. New York: Pantheon. (Translated from the French by A. M. Sheridan Smith from Naissance de la Clinique Une Archéologie du Regard Médical. Paris, Presses universitaires de France, 1963). Foucault, Michel. 1976. Mental Illness and Psychology. New York: Harper & Row. (Translated by Alan Sheridan from Maladie mentale et psychologie. Paris: Presses universitaires de France, 3.ed. 1966). Foucault, Michel. 1977. Discipline and Punish: The Birth of the Prison. New York: Random House. (Translation of: Surveiller et punir: naissance de la prison. [Paris] : Gallimard, 1975). Foucault, Michel. 1978. The History of Sexuality vol. 1-3. New York: Pantheon Books. (Vol. 1. An Introduction. Vol. 2. The Use of Pleasure. Vol. 3. The Care of the Self.) (Translation from: Histoire de la sexualite. Paris: Gallimard, 1976. 1. La volonte de savoir. 2. L'usage des plaisirs. 3. Le souci de soi). Garfield, Eugene. 1976. «Social-Sciences Citation Index Clusters”. Current Contents, no 27: 5-11. Garfield, Eugene. 1979. Citation Indexing: Its Theory and Application in Science, Technology, and Humanities. New York: John Wiley. Garfield, Eugene. 1992. “Psychology Research, 1986-1990: A Citationist Perspective on the Highest Impact Papers, Institutions, and Authors”. Current Contents, no. 41: 5-13. Goody, Jack. 1987. The Interface between the Written and the Oral. Cambridge: Cambridge University Press. Graumann, Carl-F. et al. (Eds.). 1981- . Enzyklopädie der Psychologie, Band 1-92. Göttingen: Verlag für Psychologie. (Overview see: http://www.webcitation.org/75y4mm9wG ) Gregg, Alan, Chester I. Barnard, Leonard Carmichael, Thomas M. French, Walter S. Hunter, Louis L. Thurstone, Detlev W. Bronk, John Dollard, Ernest R. Hilgard, Edward L. Thorndike, John C. Whitehorn and Robert M. Yerkes. 1947. The Place of Psychology in an Ideal University: The Report of the University Commission to Advise on the Future of Psychology at Harvard. Cambridge, MA: Harvard University Press. (Reprinted 1970 in American Psychologist 25, no. 5: 391-420, in a thematic issue containing a contemporary discussion of the place of psychology in the university). Guilford, Joy Paul. 1967. The Nature of Human Intelligence. New York: McGraw-Hill. Guilford, Joy Paul. 1982. “Cognitive Psychology's Ambiguities: Some Suggested Remedies”. Psychological Review 89, no. 1: 48-59. Harré, Rom. 1991. “Psychology”. In Handbook of Metaphysics and Ontology, edited by Hans Burkhardt and Barry Smith. Vol 1-2, A-Z. Munich, Vienna: Philosophia Verlag, vol. 2: 736-42. Hjørland, Birger. 1980. Classification Scheme developed for "Psykologisk litteratur i danske forskningsbiblioteker" (psychological literature in Danish Research Libraries). Copenhagen: The Royal Library. (Translated from Danish). Hjørland, Birger. 1982. ”Hvad er AImen Psykologi?” (What is General Psychology?). Psyke & Logos 3, 377- 385. Hjørland, Birger. 1988. «Information Retrieval in Psychology». Behavioral and Social Sciences Librarian 6, nos. 3/4: 39-64. Hjørland, Birger. 1992. “The Concept of ‘Subject’ in Information Science”. Journal of Documentation 48, no. 2: 172-200. Hjørland, Birger. 1997. Information Seeking and Subject Representation. An Activity-Theoretical Approach to Information Science. Westport, Conn.: Greenwood Press. Hjørland, Birger. 1998a. “Information Retrieval, Text Composition, and Semantics”. Knowledge Organization 25, nos. 1-2: 16-31. Hjørland, Birger. 1998b. “Theory and Metatheory of Information Science. A New Interpretation”. Journal of Documentation 54, no. 5: 606-21. *Hjørland, Birger. 1999. “Book Review of Miksa (1998): The DDC, the Universe of Knowledge, and the Post- Modern Library”. Journal of the American Society for Information Science 50, no. 5: 475–8. Hjørland, Birger and Hanne Albrechtsen. 1995. „Toward a New Horizon in Information Science: Domain Analysis”. Journal of the American Society for Information Science 46, no. 6: 400-425. Hoffman, Robert. R. and Kenneth A. Deffenbacher. 1994. “Alternative Schemes for the Analysis of Basic Versus Applied Science: Beauty Is Still in the Eye of the Beholder”. Ecological Psychology 6, no. 2: 125-130. Howarth, C. I. and W. E. Gillham (Eds.). 1981. The Structure of Psychology: An Introductory Text. London: Allen & Unwin. Ingwersen, Peter. 1994. “Polyrepresentation of Information Needs and Semantic Entities: Elements of a Cognitive Theory for Information Retrieval Interaction”. In SIGIR '94 Proceedings of the 17th annual international ACM SIGIR conference on Research and development in information retrieval Dublin, Ireland —July 03 - 06, 1994. New York: Springer, 101-110. Ingwersen, Peter. 1996. “Cognitive Perspectives of Information Retrieval Interaction: Elements of a Cognitive IR Theory”. Journal of Documentation 52, no. 1: 3-50. Jacob, Elin K. 1991. “Classification and Categorization: Drawing the Line”. Advances in Classification Research. Vol. 2. Proceedings of the 2.nd. ASIS SIGICR Classification Research Workshop. Washington, D.C.: American Society for Information Science, 67-83. Jacob, Elin K. 1994. “Classification and Crossdisciplinary Communication: Breaching the Boundaries Imposed by Classificatory Structure”. Advances in Knowledge Organization 4: 101-8. Jaroševskij, Michail G. 1980. Psykologien det 20. århundrede: Psykologiens Udvikling i Europa, USA og Sovjetunionen. Translated and abbreviated from Psichologija v XX stoletii by Povl Terkelsen and Eric Danielsen. København: Forum. Jaroševskij, Michail G. 1990. A History of Psychology. Translated from the Russian by Ruth English. Moscow: Progress Publishers. Johnson, David Martel and Christina E. Erneling (Eds.). 1997. The Future of the Cognitive Revolution. Oxford: Oxford University Press. Jørgensen, Jørgen. 1963: “Hovedtræk af Bevidsthedsytringernes Klassifikation” [Main Outline of the Classification of Expressions of Consciousness]. In: Psykologi paa Biologisk Grundlag, by Jørgen Jørgensen. Copenhagen: Munksgaard, chapter VIII+IX: 189-242. Kant, Immanuel. 1790. Kritik der Urtheilskraft. (Translated by J. Creed Meredith: The Critique of Judgement. Oxford: Clarenton Press, 1969). Karamuftuoglu, Murat. 1998. Knowledge Based Information Retrieval: A Semiotic Approach. Thesis submitted for the degree of Doctor of Philosophy. Department of Information Science, City University, London. Kedrow, Bonifaty Mikhai?lovich. 1976. Klassifizierung der Wissenschaften. Band 1·2, Koln: Pahl-Rugenstein. (Translated from Russian). [A third volume is mentioned in vol. 2] Kirk, Barbara A. 1958. “Classifying the Literature in Counseling Psychology”. Journal of Counseling Psychology 5, 89-97. Krauth, Joachim. 1981. “Techniques of Classification in Psychology I: Factor Analysis, Facet Analysis, Multidimensional Scaling, Latent Structure Analysis”. International Classification 8, no. 3: 126-32. Krauth, Joachim. 1982. “Techniques of Classification in Psychology II: Cluster Analysis, Typal Analysis, Configural Frequency Analysis, Discriminant Analysis, Regression Analysis”. International Classification 9, no. 1: 1-10. Kuhn, Thomas S. 1970. The Structure of Scientific Revolutions, Second edition. Chicago: University of Chicago Press. Kuiken, Don T., Cameron Wild and Don Schopflocher. 1992. “Positivist Conceptions of Induction and the Rejection of Classificatory Methods in Psychological Research”. In Positivism in Psychology Historical and Contemporary Problems, edited by Charles W. Tolman. New York: Springer, 47-56. *Lapointe, François H. 1970. “Origin and Evolution of the Term Psychology”. American Psychologist 25, no. 7: 640-6. Lapointe, François H. 1973. “Origin and Evolution of the Term Psychology”. Psychologia 16, no. 1: 1-16. Leont'ev, Aleksei N. 1977. Problemer i det Psykiskes Udvikling I-III. Copenhagen: Rhodos. Leontyev, Aleksei Nikolaevich. 1981. Problems of the Development of the Mind. Translated from Russian from Problemy razvitiia psikhiki. Moscow: Progress. Louttit, Chauncey McKinley. 1933. “The Dewey Decimal System and Psychology”. Journal of General Psychology 9, 234-238. Louttit, Chauncey McKinley. 1941. “Library Classification for Psychological Literature”. Psychological Record 4, 350-364. Luriya, A. R. (1977). “On the Place of Psychology in the Social and Biological Sciences”. (In Russian, translated title). Voprosy Filosofii 9, 68-76. Madsen, K. B. 1975. Systematologi. Sammenlignende Videnskabsteori for Psykologer og Beslægtede Faggrupper. Copenhagen: Munksgaard. Mammen, Jens. 1993. “The Elements of Psychology”. In The Societal Subject, edited by Niels Engelsted, Mariane Hedegaard, Benny Karpatschof and Axel Mortensen. Århus: Aarhus University Press, 29-44. Mammen, Jens. 1998. “The Psyche as ‘Res Extensa’ “. In Fourth Congress of the International Society for Cultural Research and Activity 77JeOlY (Jllne 7·11 1998 Aarhlls, Denmark). Abstracts (pp. 449-450). Aarhus: Institute of Psychology. 4th ISCRAT Congress, Århus, 10. juni Medical Subject Headings. Tree Structures. 1998. Washington, D.C: National Library of Medicine. Miksa, Francis L. 1998. The DDC, the Universe of Knowledge, and the Post-Modern Library. Albany, NY: Forrest Press. Mills, Jack. 1957. “The Classification of a Subject Field”. In Proceedings of the International Study Conference on Classification for Information Retrieval held at Beatrice Webb House, Dorking, England 13th- 17th May 1957. London: Aslib, 29-42. Mills, Jack and Vanda Broughton. 1977. Bliss Bibliographical Classification. 2. ed. Introduction and Auxilliary Schedules. London: Butterworths. Mills, Jack and Vanda Broughton. 1978. Bliss Bibliographical Classification. 2. ed. Class I: Psychology and Psychiatry. London: Butterworth. Munday, Elizabeth. 1979. “When is a Child a ‘Child’? Alternative Systems and Classification”. Journal of Anthropology Society of Oxford, 10, no. 3: 161-72. *Needham, Joseph. 1937. Integrative Levels: A Revaluation of the Idea of Progress. Oxford: Clarendon Press. Ovesen, Jan. 1989. ”Classificare Necesse Est. Lidt om Klassifikation i Antropologien - og også Ellers”. In Orden i papirerne - en hilsen til J. B. Friis·Hansen. Edited by Ole Harbo and Leif Kajberg. Copenhagen: Danmarks Biblioteksskole, 117-122. Pedersen, Johannes. 1982. “Concerning the Main Structure of a New Schedule for Psychology within the Universal Decimal Classification». International Forum on Information and Documentation 7, no. 2: 16-21. Platonov, K. K. and M. S. Mirimanova. 1983. “Objectivated Thesaurus in Psychology” (in Russian, translated title). Nauchno-Tekhnicheskaya Informatsiya, Seriya 2: Informatsionnye Protsessy i Sistemy 17, no. 3: 14-18. Poffenberger, Albert T. 1917. “A Library Classification for Books on Psychology”. Psychological Bulletin 14, no. 9: 328-31. Poulsen, Henrik. 1982. ”Leontjev, Genspejlingsbegrebet og den Almene Psykologi. Kommentarer til Ole Dreier og Benny Karpatschof”. Psyke & Logos 3, 161-75. Rescher, Nicholas. 1979. Cognitive Systematization: A Systems-theoretic Approach to a Coherentist Theory of Knowledge. Totowa, N.J.: Rowman and Littlefield. Rice, Robert W. 1978. “Formal Classification of Research Information. An Empirical Test of the McGrath- Altman Approach and an Illustrative Case”. American Psychologist 33, no. 3: 249-64. Rosenzweig, Mark R. 1992. “Unity and diversity of psychology”. International journal of Psychology 27, no. 5: 283-90. Ruckmick, C. A. 1926. “A Schematic Classification of General Psychology.” Psychological Review 33, no. 5: 397-406. Sams, Thomas, Peter Laursen and Leena Eskelinen. 1994. “Response Classification in Psychological Testing Using a Neural Network”. International Journal of Neural Systems 5, no. 3: 253-6. Šamurin, Evgenij I. 1964–1967. Geschichte der bibliothekarisch-bibliographischen Klassifikation. Zwei Bande. Leipzing: VEB Bibliographisches Institut. (Translated from Russian). Scheerer, Eckart. 1989. „Psychologie“. In: Historisches Wörterbuch der Philosophie. Edited by Joachim Ritter and Karlfried Grunder. Darmstadt: Wissenschaftliche Buchgesellschaft, vol. 7: 1599-653. Schonpflug, Wolfgang. 1993. „Applied Psychology: Newcomer with a Long Tradition.” Applied Psychology: An International Review 42, no. 1: 5-30. [This article is debated by de Wolff, Landy, Louw and Semmer in the same issue of the journal]. Schultz, Duane P. and Sydney Ellen Schultz. 1996. A History of Modem Psychology. Sixth edition. Philadelphia: Harcourt Brace College Publishers. Schultz, Erik. 1988. Personlighedspsykologi på Erkendelsesteoretisk Grundlag - eller Mysteriet om Personen der Forsvandt. Copenhagen: Dansk Psykologisk Forlag. Shapere, Dudley. 1977. “Scientific Theories and Their Domains”. In The Structure of Scientific Theories, Second edition. Edited by Frederick Suppe. Urbana: University of Illinois Press, 518-99. Shapere, Dudley. 1984. Reason and the Search for Knowledge: Investigations in the Philosophy of Science. (Boston Studies in the Philosophy of Science vol. 78). Dordrecht: D. Reidel. Shapere, Dudley. 1989. “Evolution and Continuity in Scientific Change”. Philosophy of Science 56, no3: 419- 37. Soudek, Miluse. 1980. “On the Classification of Psychology in General Library Classification Schemes”. Library Resources and Technical Services 24, no. 2: 114-128. Soudek, Miluse. 1983. “The Inadequate Treatment of Psychology in General Library Classification Schemes.” Canadian Psychology 24, no. 3: 181-9. Thorne, Frederick C. 1964. “Diagnostic Classification and Nomenclature for Psychological States”. Journal of Clinical Psychology 20, no. 1: 3-60. Tolman, Charles W. 1992. “Watson’s Positivism: Materialist or Phenomenalist?” In Positivism in Psychology. Historical and Contemporary Problems, edited by Charles W. Tolman. New York: Springer, 83-102. Tuttle, R. H. 1999. “Taxonomy. The Objectives of Biological Classification”. In: The Biological Sciences. Britannica ® CD 99 Multimedia Edition. Chicago, II.: Encyclopaedia Britannica, Inc., © 1994-1999. Vandekemp, Hendrika Vande. 1983. “A Note on the Term Psychology in English titles - Predecessors of Rauch”. Journal of the History of the Behavioral Sciences 19, no. 2: 185. Vandekemp, Hendrika Vande. 1980. “Origin and Evolution of the Term Psychology – Addenda”. American Psychologist 35, no. 8: 774. Vickery, Brian C. 1960. Faceted Classification. A Guide to the Construction and Use of Special Schemes. London: ASLIB. Walker, Alvin (Ed.). 1991. Thesaurus of Psychological Index Terms. 6th. ed. Washington, D.C.: American Psychological Association. Warnke, Camilla. 1985. “Klassifikation der Wissenschaften“. In: Philosophisches Wörterbuch 1-2, edited by Georg Klaus and Manfred Buhr. 13th. Edition. Berlin: Das Europäische Buch, vol. 1: 629-33. White, Howard D. and Katherine W. McCain. 1998. “Visualizing a Discipline: An Author Co-Citation Analysis of Information Science, 1972-1995.” Journal of the American Society for Information Science 49, no. 4: 327- 55. Whitley, Richard. 1984. The Intellectual and Social Organization of the Sciences. Oxford: Clarendon. Willoughby, Raymond Royce. 1932. “A Classification for Psychology”. Journal of General Psychology 7, no.1: 179-85. Winograd, Terry and Fernando Flores. 1986. Understanding Computers and Cognition: A New Foundation for Design. New York: Addison-Wesley.