In the preceding chapters, I have formulated the object of my investigation in terms of expansion from the level of prevalent individual actions to the level of novel collective activity. Such transitions have commonly taken place as if above the heads of the affected individuals and groups, in the form of historical tragedies and puppet shows of varying scales.
I have argued that a new type of 'learning by expanding' is emerging in the current phase of human history. This implies that the transitions mentioned above are becoming potential objects of conscious or intuitive mastery.
Conscious goal-directed processes are situated on the level of actions, or secondary instruments. This level is the homestead of thinking. Thinking is most typically described as a series of relatively discrete actions of 'gap filling' or problem solving. The emergence of thoughtfully mastered learning activity or 'learning by expanding' implies the extension of thinking into an activity, and the merger of learning and thinking into one unified process on this level.
The problem is to identify the specific instruments of this new type of expansive learning and thinking. For this purpose, I shall first critically examine certain dominant modes of theorizing about thinking.
In his book The Foundations of Primitive Thought, C. R. Hallpike (1979) defines the characteristics of 'primitive' and 'advanced' thinking as follows (Table 4.1; compiled by Atlas 1985, 336).
Table 4.1
Characteristics of 'primitive' and 'advanced' thought after Hallpike (1979)
___________________________________________________________________
Type of thought
'Primitive' 'Advanced'
Domain of thought
Symbolism image-based, affective linguistic
Classification associational taxonomic
Number and
measurement concrete, absolute abstract, relative
Space perceptual conceptual
Time qualitative, quantitative, capable of
incommensurable comparison
Conceptions and fusion of the psychical mind/body duality;
representations of and physical; private distinction between private
the person states not verbally and public awareness
elaborated
Causality essentialist impersonal, probabilistic
___________________________________________________________________
Hallpike uses the Piagetian cognitive stages as his analytical framework. According to him, inhabitants of 'primitive' societies are for the most part characterized by preoperational thought, not reaching the level of concrete operational thinking typical to children of seven years and older living in 'advanced' societies.
For Hallpike, life in 'primitive' societies is cognitively less demanding than life in 'advanced' societies. One source of higher cognitive demands in 'advanced' societies is the presence of mechanical devices and complex technical implements. Substitutability of labor, impersonal productive relations and the rationalization of activity are the features of civilization celebrated by Hallpike. As Atlas (1986, 335) notes in his review, Hallpike's book echoes old mainstream ideas on 'primitive' mentality. The novelty is his wedding of Piaget to this tradition.
The form of theorizing demonstrated by Hallpike is deeply rooted in our psychological reasoning. It is salient in many current discussions of the psychology of human thinking, including attempts with aims opposite to those of Hallpike's. This general form of theorizing is the pervasive use of dichotomies as explanatory constructs.
In his pioneering study of the cultural foundations of cognition, A. R. Luria (1976) distinguished between two broad types of thinking: one concrete, situational and 'graphic-functional', the other abstract, categorical and logical. The protocol of a subject called Rakmat, produced as a response to a classification task, is a famous example of the former type.
"Subject: Rakmat., age thirty-nine, illiterate peasant from an outlying district; has seldom been in Fergana, never in any other city. He was shown drawings of the following: hammer - saw - log - hatchet.
'They are all alike. I think all of them have to be here. See, if you're going to saw, you need a saw, and if you have to split something you need a hatchet. So they're all needed here.'
Employs the principle of 'necessity' to group objects in a practical situation.
(...)
Which of these things could you call by one word?
'How's that? If you call all three of them a 'hammer,' that won't be right either.'
Rejects use of general term.
But one fellow picked three things - the hammer, saw, and hatchet - and said they were alike.
'A saw, a hammer, and a hatchet all have to work together. But the log has to be there, too.'
Reverts to situational thinking.
Why do you think he picked up these three things and not the log?
'Probably he's got a lot of firewood, but if we'll be left without firewood, we won't be able to do anything.'
Explains selection in strictly practical terms.
True, but a hammer, a saw, and a hatchet are all tools.
'Yes, but even if we have tools, we still need wood - otherwise, we can't build anything.'
Persists in situational thinking despite disclosure of categorical term." (Luria 1976, 55-56.)
Luria's schooled subjects behaved differently. To them, the task of isolating a particular attribute as a basis of categorization seemed "a natural, self-evident procedure" (Luria 1976, 78). These schooled subjects actually represented a historical phase entirely different from that represented by Rakmat. Rakmat was a man of a pre-industrial and pre-literate age. The schooled subjects were men and women of socialism and industrialization in the takeoff.
Luria's conclusions imply that concrete situational thinking is something lower or less developed than abstract categorical thinking. This has prompted Cole and Griffin (1980, 352) to note that the qualitative changes in cognition that Luria sought to demonstrate led him into comparisons "that were distressingly quantitative in their implications".
The problem is: can development be conceived of as a linear process where certain valuable ingredients (such as the 'abstractness' of thinking) gradually or abruptly increase while other, restrictive ingredients (such as the 'concreteness' of thinking) decrease? The answer given by Cole and Griffin is negative. While being forced to admit that technologies have evolved from the simple to the complex and more powerful, they point out that in spheres like politics or family life such linear evolutionary schemes are inappropriate (Cole & Griffin 1980, 362).
The justified opposition to linear schemes easily leads to a denial of all logic or lawfulness in history. The result may be a pluralistic ahistorical constructivism along the lines of Nelson Goodman's (1978) 'worldmaking'. The idea that anything may be constructed from what is given and that no constructed world is instrinsically more true than any other is refreshing and spiritually liberating. But it is not very powerful in the face of the overwhelming movement of societal reality. And it helps us very little in our attempts to understand how our societies have evolved.
So Luria's weakness is not the same as that of Hallpike's who presents his dichotomy in essentially ahistorical terms. Luria's dichotomy is an attempt to understand historically the transformation of thinking. It is precisely this that makes Luria's study a pathbreaking classic. Luria's trouble is on a different level. It is a question of what is the logic of history - if it is not linear.
THE SECOND DICHOTOMY: EXPERIENCE VERSUS ANALYSIS
Hallpike's dichotomy sees the concrete thought of the 'primitive' societies as something essentially lower than the abstract thought of the 'advanced' societies. Some recent treatises take a different standpoint, actually praising the neglected virtues of various forms of concrete, tacit and non-analytical thought (though not necessarily connecting these forms with so called 'primitive' societies).
In their book Mind over Machine (1986) Hubert and Stuart Dreyfus discuss the nature and acquisition of expertise in the era of the computer. Their argument is that we cannot explain human expertise as behavior based on explicable principles and rules. A true expert makes decisions on intuitive basis. The psychological mechanism behind intuition is experience-based wholistic recognition of similarity, producing deep situational understanding and fluid, rapid behavior. Through experience, we store in our memories large amounts of typical situations which bear no names and defy complete verbal description: "experience seems immeasurably more important than any form of verbal description" (Dreyfus & Dreyfus 1986, ....).
"While most expert performance is ongoing and nonreflective, when time permits and outcomes are crucial an expert will deliberate before acting. But (...) this deliberation does not require calculative problem-solving, but rather involves critically reflecting on one's intuitions." (Dreyfus & Dreyfus 1986, ....)
The authors describe the process of becoming an expert as consisting of five stages of skill acquisition (Table 4.2).
Table 4.2
The five stages of skill acquisition after Dreyfus & Dreyfus (1986)
__________________________________________________________________
Skill level Components Perspective Decision Commitment
1. Novice context-free none analytical detached
2. Advanced beginner context-free and " " "
situational
3. Competent " chosen " detached understanding
and deciding; involved
in outcome
4. Proficient " experienced " involved understanding;
detached deciding
5. Expert " " intuitive involved
__________________________________________________________________
The acquisition process is depicted as a linear sequence from the analytical to the intuitive, from the rule-guided 'kowing that' to the experience-based know-how. It is essentially a process of internalization.
"An expert's skill has become so much a part of him that he need be no more aware of it than he is of his own body." (Dreyfus & Dreyfus 1986, ....)
The process is not only linear. For the authors it also seems to be automatic and self-evident in every case of expertise acquisition. Experience is the golden key to the consequtive steps of this path.
This assumption fails to explain why so many people never become fluid intuitive experts in spite of years and years of experience. Somehow the authors seem to forget all about the rigidity associated with extensive routinization.
The Dreyfus brothers' singular praise of experience may be contrasted with the findings produced over the years by research on learning from experience in probabilistic situations (see Brehmer 1980; Kahneman, Slovic & Tversky 1982).
Brehmer (1980, 224-227) points out the weakness of the psychological research supposedly demonstrating how people learn from experience. The tasks used in that kind of research, such as paired associates and classification tasks, typically employ materials where the truth is manifest. In the word lists of paired associates the subject immediately knows what he is supposed to learn. Similarly, in the classification tasks the common components of the stimuli, such as color and form, are already well formed concepts and the experimenter is certain that the subjects already have the hypotheses relevant to the task. Thus, the guarantee of the validity of the solution in these tasks does not come from experience. It comes from the experimenter through his choice of materials.
"The paradigm may thus very well model the situation in teaching, where the teacher decides for the pupil what the truth should be in a given case, but it certainly does not model the situation in which a person is learning from experience." (Brehmer 1980, 225.)
The situation is different when subjects face complex probabilistic tasks, such as diagnostic decision making. The truth is not manifest. Nobody tells the practitioner what there is to learn, or even whether there is anything for him to learn. The fact that the chosen treatment leads to recovery does not mean that the decision was correct, for (a) the recovery may have had other causes, (b) other kinds of treatment might have been equally or more effective, and (c) the chosen treatment may eventually have other unwanted effects which are, however, difficult if not impossible to trace back to their cause with full certainty. Even if the chosen treatment works, the explanation for why it works may be very different from what the practitioner thinks it is. But the practitioner learning from experience learns mainly from the outcomes of his actions. As Dreyfus & Dreyfus (1986, .....) put it, "the proficient performer has experienced similar situations in the past and memories of them trigger plans similar to those that worked in the past".
"When we learn from outcomes, it may, in fact, be almost impossible to discover that one really does not know anything. This is especially true when the concepts are very complex in the sense that each instance contains many dimensions. In this case, there are too many ways of explaining why a certain outcome occurred, and to explain away failures of predicting the correct outcome. Because of this, the need to change may not be apparent to us, and we may fail to learn that our rule is invalid, not only for particular cases but for the general case also." (Brehmer 1980, 228-229.)
Mere experience, even of probabilistic tasks, seems only to strengthen the subjects' non-pobabilistic thinking. Subjects prefer to assume that there is a deterministic causal rule behind every task. When their assumed deterministic rules fail, they tend to assume that there is no rule at all, rather than seriously consider the possibility that the rule may be probabilistic in character.
"These results, then, support the earlier results on clinical inference in that they show that people do not learn optimal strategies from experience even if they are given massive amounts of practice. The reason why the subjects fail to improve in these tasks seems to be that they lack the necessary basic schemata to help them understand and use the information provided by their experience. Rather than using the appropriate statistical schemata, subjects use an inappropriate causal or deterministic shema. (...) The characteristic of probabilism is, of course, not manifest, but it has to be inferred. (...) for a person with a firm belief in the deterministic character of the world, there is nothing in his experience that would force him to discover that the task is probabilistic and to give up the notion of determinism. (...) In short, probabilism must be invented before it can be detected." (Brehmer 1980, 233-235.)
The problem with learning from experience is actually the classical problem of induction. According to the classical theory of induction, we make generalizations on the basis of experiencing many things of a similar kind. Dreyfus and Dreyfus (1986, ....) seem to subscribe to the classical position: "through practical experience in concrete situations with meaningful elements, which neither the instructor nor the learner can define in terms of objectively recognizable context-free features, the advanced beginner starts to recognize those elements when they are present". And this happens "thanks to a perceived similarity with prior examples".
This kind of empirical generalization seems to work reasonably well when we are dealing with simple stimuli and well established conventions. But when the cases we observe are complex and novel, how do we know that things are really similar and instances of the same general class? For that, we need to know what the relevant characteristics are in the first place. We have to define what we are to learn before we can learn it. Pure induction turns out to be a fallacy, as Nelson Goodman demonstrated long ago.
"To say that valid predictions are those based on past regularities, without being able to say which regularities, is thus quite pointless. Regularities are where you find them, and you can find them anywhere." (Goodman 1983, 82.)
What we commonly think is pure experience is actually sense data selected and interpreted by our culturally molded but not necessarily modern schemata and mental models. Probability calculus was invented in the seventeenth century, perhaps because "it was only at this point in time that the notion of causality had reached such a level that it could provide a suitable contrast against which to evaluate disorder" (Brehmer 1980, 235). In individual practitioners, the old cultural model of linear causality has tremendous persistence. This illuminates the conservatist bias of experience. Recalling the Dreyfus brothers' unreserved belief in the power of experience, Brehmer's (1980, 224) conclusion that we have come to have "a perverse conception of the nature of experience" is not unfounded.
The Dreyfus brothers' dichotomy is experience-based intuitive expertise vs. rule-based analytical expertise. Employing Brehmer's critique of experience, we obtain a further dichotomy: experience as casual growth of wholistic intuition vs. experience as strengthening of rigid and biased routines. So we are still stuck with a dichotomy.
One final version of the dichotomy deserves our attention. It is the split between scientific and artistic, or paradigmatic and narrative thought, recently revitalized by Jerome Bruner (1986).
According to Bruner (1986, 12), the paradigmatic or logico-scientific mode of thought attempts to fulfill the ideal of a formal, mathematical system of description and explanation. It employs categorization or conceptualization and the operations by which categories are established, instantiated, idealized, and related one to the other to form a system. Propositions are extracted from statements in their particular contexts. The logico-scientific mode deals in general causes, and in their establishment. It makes use of procedures to assure verifiable reference and to test for empirical truth. Its language is regulated by requirements of consistency and noncontradiction.
The narrative type of thought has opposite characteristics.
"The imaginative application of the narrative mode leads instead to good stories, gripping drama, believable (though not necessarily 'true') historical accounts. It deals in human or human-like intention and action (...). It strives to put its timeless miracles into the particulars of experience, and to locate the experience in time and place. (...) The paradigmatic mode, by contrast, seeks to transcend the particular by higher and higher reach for abstraction, and in the end disclaims in principle any explanatory value at all where the particular is concerned." (Bruner 1986, 13.)
Bruner's book is a continuation of a distinguished series on essentially analogous dichotomies: science vs. humanities; nomothetic vs. idiographic; concepts vs. images; positivism vs. phenomenology. In psychology, the same basic division was championed already by Wundt.
Many recent efforts to deal with these dichotomies aim at balancing or combining the two sides. In his important book on Imagery in Scientific Thought, Arthur I. Miller (1984, 312) concludes that "when scientists hold a theory, they hold a particular mode of imagery as well". Herbert Simon (1983, 28) joins the credo by stating that there is no contradiction between the intuitive model and the behavioral model of thinking, since "all serious thinking calls on both modes, both search-like processes and the sudden recognition of familiar patterns". In his theory of fantasy, Roset (1984) presents the two sides as alternating phases of intuitive production and analytical control, or 'anaxiomatization' and 'hyperaxiomatization'.
These combinations leave us with constructions of the type 'both-and' instead of mere 'either-or'. But the abstract dichotomous structure remains at the heart of the argument.
The problem with the dichotomies is that they depict movement as mechanical opposition, summation or oscillation between two fixed poles, thus effectively excluding the dimension of concrete historical development.
'Either-or' and 'both-and' are closed and timeless structures. Within them, there is no room for something qualitatively new emerging first as a subordinated mediator between the two poles and being transformed into a determining factor that will eventually change the character of the whole structural configuration. There is no room for thirdness.
In classical treatises on the psychology of thinking one finds, however, intriguing attempts to overcome the dichotomous structure. My first example is John Dewey's (1910) How We Think. In this book, Dewey takes up the problem of experience. He first criticizes our belief in experience much in the manner Brehmer did it 70 years later.
"But even the most reliable beliefs of this type fail when they confront the novel. Since they rest upon past uniformities, they are useless when further experience departs in any considerable measure from ancient incident and wonted precedent. (...)
Mental inertia, laziness, unjustifiable conservatism, are its probable accompaniments. Its general effect upon mental attitude is more serious than even the specific wrong conclusions in which it has landed. Wherever the chief dependence in forming inferences is upon the conjunctions observed in past experience, failures to agree with the usual order are slurred over, cases of successful confirmation are exaggerated. Since the mind naturally demands some principle of continuity, some connecting link between separate facts and causes, forces are arbitrarily invented for that purpose." (Dewey 1910, 148.)
But Dewey is not satisfied with this. He realizes that people may also become truly flexible and inventive experts. This kind of development is based on experimentation and hypothesis testing. But that in turn cannot be explained as something given from above, mechanically separated from experience. Thus, experience acquires a deeper double meaning.
"In short, the term experience may be interpreted either with reference to the empirical or the experimental attitude of mind. Experience is not a rigid and closed thing; it is vital, and hence growing. When dominated by the past, by custom and routine, it is often opposed to the reasonable, the thoughtful. But experience also includes the reflection that sets us free from the limiting influence of sense, appetite, and tradition. Experience may welcome and assimilate all that the most exact and penetrating thought discovers. Indeed, the business of education might be defined as just such an emancipation and enlargement of experience." (Dewey 1910, 156.)
The external opposition of experience versus analysis has thus been transformed into an internal contradiction within experience itself. But the mediating thirdness is still lacking.
To find the way out, we must take a critical look at the logic implicitly attributed to both components of experience, or to both intuitive and analytical thinking. Max Wertheimer's Productive Thinking (1945) is a classical work which provides us with this critique.
"There are several objects. (The way in which they are segregated, and why just so, how an object constitutes itself in separation from other objects, is a question neglected in traditional logic, is taken for granted without real investigation.) I compare them. In their qualities of their parts I find similarities and differences. Abstracting from the differences, and concentrating on common qualities or parts in the objects, I get a general concept. The content is given by these common parts. This is the 'intension.' The 'extension' is the manifold of objects embraced by the class concept.
If we call the common element m, and the other elements x, an exact expression for the class (or for any object as conceived under the class concept) is
m + x.
Between the m and the x is an 'and.' The m is what is common in the contents of the objects; the x is what there is besides the m and may vary in the contents of the various objects. The conceived datum, m, is independent of its setting to the left and right, and apparently must be so for the sake of exact use of the concept in inference, syllogisms, etc. There is no reference to whatever else there may be in the object besides, no references to the role which m plays in this object, no reference to its meaning as a part among the other parts of the same entity, no reference to the structure of this entity. This abstraction is substractive; it simply isolates the m. For the m it does not matter what the x is. (...)
In the historic development difficulties have arisen as to the adequacy of the procedure (...). The problem was whether such a procedure, although exact, does not easily combine objects which are basically different in nature and, on the other hand, sharply separate objects which belong to each other in fact. The logician seeks help in the term 'essential.' There always was emphasis on this point; but although for common sense the meaning of 'essential' is often clear enough, unfortunately it was and has remained extremely controversial in logic. It has served to name the problem rather than to solve it. It has consequently been rejected again, excluded in newer developments of logic." (Wertheimer 1945, 207-208.)
Another grave feature in traditional logic is its insistence that the items of discourse - concepts, propositions, etc. - must remain rigidly identical if repeated. In real thinking processes, items do not remain identical. To the contrary, precisely their change is required. Their functional and structural meaning changes, and blindness to such change impedes productive processes. Formal logic is incapable of grasping development because it disregards "the intense directedness of live thought processes as they improve a given situation" (Wertheimer 1945, 215).
This fundamental insight has long been neglected in cognitive and anthropological studies of classification. A recent study of the conceptual organization of practicing blacksmiths indicates that the emphasis may be changing. "What leads to highly effective means of blacksmithing is flexibility in classification. There is no one basic structure to which we can turn as the key to the practice of blacksmithing. Blacksmithing, like other behavior, is characterized by productivity." (Dougherty & Keller 1985, 170-171; see also Gatewood 1985.)
Wertheimer (1945, 10) concludes that in comparison with actual thought processes, the rules and examples of traditional logic look "dull, insipid, lifeless". If one tries to describe processes of productive thinking in terms of formal logic, one may have a series of correct operations but the sense of the process, what is vital and creative in it, is lost.
One factor behind the persistence of formal-logical conceptions of thinking is their correspondence to certain deep-seated modes of real thought processes. The outstanding instance is our habit to proceed only successively, step by step, in an 'and-summative' fashion. According to Wertheimer (1945, 88), this may be due to the fact that "we cannot write down two propositions simultaneously, that in reports we have to proceed one thing after the other". In other words, the concrete-historical instrument of written language enters as a structural determinant of thinking. Unfortunately Wertheimer does not continue this line of analysis. It remains an intriguing hint, a sidetrack without consequence.
For Wertheimer, there is something essential behind the endless multitude of external properties of objects. This essential includes the following aspects:
- the wholeness or whole-quality of the object or situation, as opposed to a mere additive listing of its parts;
- the clear, complete and consistent structure of the object, as opposed to an incomplete or shallow structure;
- the inner relatedness of the parts of the whole, as opposed to their separation or discreteness;
- the center, core or radix of the whole, as opposed to a stuctrue without center.
The essential is thus the 'good gestalt', and productive thinking is transition from a bad gestalt to a good one. Wertheimer summarizes his idea in the following description.
"Thinking consists in
envisaging, realizing structural features and structural requirements; proceeding in accordance with, and determined by, these requirements; thereby changing the situation in the direction of structural improvements, which involves:
that gaps, trouble-regions, disturbances, superficialities, etc., be viewed and dealt with structurally;
that inner structural relations - fitting or not fitting - be sought among such disturbances and the given situation as a whole and among its various parts;
that there be operations of structural grouping and segregation, of centering, etc.;
that operations be viewed and treated in their structural place, role, dynamic meaning,
including realization of the changes which this involves;
realizing structural transposability, structural hierarchy, and separating structurally peripheral from fundamental features - a special case of grouping; looking for structural rather than piecemeal truth." (Wertheimer 1945, 190-191.)
For a modern cognitive scientist, characterizations like the one cited above are aggravating, if not totally meaningless. It is hard to find tangible operational, not to speak of measurable, counterparts or indices for Wertheimer's concepts. For Wertheimer, this kind of reaction would rather prove his point, being another example of the dominant piecemeal, and-summative way of thinking.
It is not justified to nullify Wertheimer's work on account of its lacking concreteness. Wertheimer does present a very convincing series of concrete examples, ranging from the famous parallelograms to the unique account of Einstein's way to the discovery of relativity. In these examples, he demonstrates how productive thinking proceeds. But he does not demonstrate what primary and secondary instruments could be used to enhance this type of thinking. Obviously this is why gestalt theory was overrun by the many variants of behaviorism. Skinner offered the world concrete tools with which one could do something practical. Wertheimer did not.
This criticism could be interpreted as crude utilitarianism. But there is more at stake here. The question of instruments is above all a theoretical weakness in Wertheimer's work. As I noted earlier, he only accidentally touched the role of cultural instruments - namely written language - as determinants of the development of thinking. He did not seek the expansive perspective by way of a historical analysis of the emerging new instrumentalities of thought. His conception was presented as an unhistorical, eternal solution. Productive thinking, aimed at the 'good gestalt', was for him a moral imperative, something stemming from inside, being already planted deeply in the human nature: "in humans there is at bottom the desire, the craving to face the true issue, the structural core, the radix of the situation" (Wertheimer 1945, 191). Thus, at least implicitly, the emergence of productive thinking was something to be realized by individual willpower.
And yet, in spite of this critical weakness, there is something prophetic in Wertheimer's vision of thinking as expansion.
"In such processes of thinking the solution of an actual task, 'Problem solved, task finished,' is not the end. The way of solution, its fundamental features, the problem with its solution function as parts of a large expanding realm. Here the function of thinking is not just solving an actual problem, but discovering, envisaging, going into deeper questions. Often in great discoveries the most important thing is that a certain question is found. (...)
Often such a process takes a long time; it is drama with setbacks and struggles. There are fine cases in which the process proceeds irresistibly, through months, through years, never losing sight of the deeper issue, never getting lost in petty details, in detours, bypaths." (Wertheimer 1945, 122-123.)
So Wertheimer gives us prophecy, but not instruments. To get some idea of the latter, I'll consult a third classic, namely Sir Frederic Bartlett's book Thinking (1958).
According to Bartlett (1958, 182), much of what is called inductive generalizing is "no more than the acceptance, with biased selection, of already formed social conventions". These generalizations "have little to do with transfer of practice or training save that they make it more difficult" (Bartlett 1958, 184).
There is, however, also an exploratory or experimental type of generalization which may lead to genuinely new discoveries and concepts. But even this is not accomplished by 'purifying' the sensory data from cultural conventions. To the contrary.
"However 'pure' his aims may be he has to be able to practise a technique and to handle a technology. Far the most important aspect of the experimenter's need to master method and to handle apparatus is that in the majority of cases (...) the method and the instrumentation are brought into his field of work from the outside." (Bartlett 1958, 133.)
Why are instruments so important in experimental thinking? Successful experimenting requires that the experimenter knows where to look for new crucial findings. This step becomes possible when apparatus, methods, hints, or established findings are taken over from some field different from that in which they are applied. They subsequently function as 'lenses' that allow for a novel perspective. (Bartlett 1958, 134.)
Bartlett's insight has recently been restated by Tweney, Doherty and Mynatt (1981, 411-412).
"Scientists self-consciously bring a store of knowledge to bear on the task at hand, as well as a highly developed set of intellectual tools. They may use extensive note-taking, carefully organized records of data, files, and libraries, as 'external memories.' They use blackboards, mathematics, even formal logic. Latour and Woolgar (1979) noted that nearly all of the behavioral activity in a major laboratory consisted of manipulation of symbols, and only a tiny fraction involved direct contact with the phenomena under investigation. Cognitive psychologists have typically studied 'prescientific man.' The typical subject in a psychological laboratory has access only to presented stimuli and almost never to memory aids or other heuristics. The intent has been to study basic cognitive processes, unencumbered by cultural artifacts or aids. A cognitive psychology of sciencewill have to focus instead on aided cognition, on the psychology of scientists when all the tools of the trade are at their disposal."
However, the above statement falls short of grasping the gist of Bartlett's idea. For Bartlett, the specific instrumentality of exploratory thinking implies also its specific sociality. The sociality of experimental thinking is not of an immediate, face-to-face nature (though it may certainly include that, too). The necessity of taking over instruments from other, overlapping fields means that experimental thinking is "fundamentally co-operative, social, and cannot proceed far without the stimulus of outside contacts" (Bartlett 1958, 123).
This is a specific extended kind of sociality. It indicates the expansive, cyclic nature of experimental thinking. According to Bartlett (1958, 136), when any experimental science is ripe for marked advance, a mass of routine thinking has come near to wearing itself out by exploiting a limited range of technique to establish more and more minute and specialized detail.
"However, at the same time, perhaps in some other branch of science, and perhaps in some hitherto disconnected part of what is treated as the same branch, there are other techniques generating their own problems, opening up their own gaps. An original mind, never wholly contained in any conventionally enclosed field of interest, now seizes upon the possibility that there may be some unsuspected overlap, takes the risk whether there is or not, and gives the old subject-matter a new look. (...)"
This passage takes us back to Figure 3.2. The phase of repetitive production of more and more specialized detail precedes the phase of the need state, or the primary contradiction. The phase where a new instrument is seized upon and taken over from an overlapping field corresponds to the emergence of the secondary contradiction where a foreign element is introduced into the prevalent activity structure.
In Chapter 3, I used the metaphor of a voyage to characterize the zone of proximal development. Bartlett (1958, 137) describes the course of experimental thinking much in the same manner.
"The experimental thinker is in the position of somebody who must use whatever tools may be available for adding to some structure that is not yet finished, and that he himself is certainly not going to complete. Because the materials that he must use have properties of their own, many of which he cannot know before he uses them, and some of which in all likelihood are actually generated in the course of their use, he is in the position of an explorer rather than that of a spectator."
Notice the expression "he himself is certainly not going to complete". Here Bartlett hints at the social dimension of the expansion. The qualitatively new scientific concept - or the qualitatively new form of scientific activity - is going to be a collective formation that goes beyond all the individual actions that gave rise to it.
At the end of his book, Bartlett discusses artistic thinking. He notices that when an artist has got his work well under way, "it very often appears to him that something outside himself has taken charge and is now settling everything that happens" (Bartlett 1958, 192). This experience is not foreign to scientists either. The phenomenon is due to the anticipation of the essentially collective and societal, tertiary character embedded within a work of art (or science) under creation (recall Zinchenko's 'liberated actions' and Bateson's loss of the 'I'). The double nature of this expansion is evident in a work of art in that "it is at once convincing and satisfying, and yet question-making and disturbing" (Bartlett 1958, 196). In other words, it requires simultaenously acceptance of a convention - the given new - and passing beyond it "towards whatever standard it serves" (Bartlett 1958, 193) - the created new.
When is the artifact an instrument? In the realm of primary artifacts, we speak of objects of consumption, rawmaterials of production, and instruments of production. There are rapidly and slowly renewed objects of consumption: a loaf of bread belongs to the former, a television set belongs to the latter. A piece of wood and a bag of flour are rawmaterials of production. A hammer is supposedly an example of instruments of production.
One hesitates to make sharp distinctions like those suggested above. The differences between these types of artifacts are relative, and the same artifact may have different meanings depending on the context. For a television critic, the TV set is an instrument of production. For a collector of old tools, the hammer may be an object consumption.
In the realm of secondary artifacts, similar types may be tentatively distinguished.
Firstly, the continuously changing flow of information, consisting of specific opinions, news, descriptions, advertisements, etc. may be identified as the rapidly renewed objects of consumption on the secondary level.
Secondly, the relatively stable and general representations with which we filter and modulate our daily information flow may be indentified as the slowly renewed objects of consumption on the secondary level.
Thirdly, both above-mentioned types may be turned into objects or rawmaterials of production, to be molded and transformed into something new.
Fourthly, sign systems such as gestures, spoken and written language, or mathematical and musical notation may be identified as typical, continuously available instruments of production on the secondary level.
The relations between the types of artifacts are not 'and-summative' but genuinely complementary (see Otte 1980; 1984). They both presuppose and struggle with each other. In the course of development, the different types are truly transformed into each other.
Expansive thinking requires that relatively stable objects of consumption and production are transformed into instruments of production. Cycles of expansion, or zones of proximal development, activate the "complementarity of representational and instrumental aspects" (Otte 1980, 64) of such conceptual objects. The representational concept, as a static and uncritically accepted frame, must be transformed into an instrumental concept, critically reflected, molded and applied, and back to a new representational frame.
According to the standard view, a concept is a verbal label that encompasses an array of diverse instances deemed to be related. The array must have coherence or family resemblance. Concepts are formed by comparing particular objects with one another and finding their common features. Concepts are thus memoranda of identical features in objects perceived. They are means for bunching together objects scattered in experience. The process necessary and sufficient to generate concepts is classifying. (Sigel 1983, 242-245.)
This standard view has been remarkably persistent in psychology and education. Within the mainstream cognitive psychology, it has been seriously challenged only quite recently.
The first challenge comes from the Piagetian impulse. Katherine Nelson is a well known representative of this challenge. According to her, "the child's initial mental representations are in the form of scripts for familiar events involving social interaction and communication" (Nelson 1983, 135). A script is a structured whole, a generalized representation of a sequence of activity that has occurred more than once. Therefore, the basic and initial form of conceptual representation is that of event representation. Concepts of particular objects are later achievements. In other words, paradigmatic categories are extracted from syntagmatic representations. Finally contextfree categories are formed.
"Note that what is not involved in any of the operations outlined thus far is an analysis in terms of the similarity of object types independent of their functional relations. The analysis assumes instead that the child operates for a very long time with a conceptual representation that defines object categories in terms of their relationships and not in terms of their internal qualities (...). The establishment of similarity relations is assumed to be a more advanced cognitive operation that takes place only after the basic categories have been formed." (Nelson 1983, 141; italics in the original.)
Nelson's critique of the standard view is that the abstraction and generalization of similarity features is assumed to be initial. For Nelson, this type of concept formation becomes dominant only later.
Nelson's view leaves the old belief in induction intact. According to her, children acquire their scripts through the same kind of induction as the standard view attributes to the acquisition of similarity features - only the unit is more holistic, namely a social event script. In other words, Nelson's critique accepts the basic logic of the standard view.
This point has recently been made very clearly by Ivana Marková (1982, 59; italics added).
"We can thus conclude that although 'scripts' and 'plans' and perhaps some other terms introduce 'context' and 'experience' into the understanding of language and events, the conceptual framework has not changed. We may say that the theory of 'scripts' and 'plans' is an example of the attempts to save the collapsing Cartesian paradigm. (...) scripts and plans exist only because a person has been in that particular situation before and is simply matching the pre-stored representations to his new experience. People can cope with new situations because they can understand them in terms of their previous experience, because they can re-organize the pieces of information they already have in their internal representations. No actual development is taking place: the apparent development of plans and scripts is really only a regrouping of static and predetermined elements of information."
This very logic has been partially questioned in two new contributions to the problem of conceptual thinking. These are Susan Carey's (1985) monograph Conceptual Change in Childhood and the paper The Role of Theories in Conceptual Coherence by G. I. Murphy and D. L. Medin (1985).
Not surprisingly, the authors of both contributions take their philosophical stance from Nelson Goodman's critique of induction. They point out that any two entities can be found arbitrarily similar or dissimilar by changing the criterion of what counts as a relevant attribute. There is always an infinity of features in terms of which two objects may be compared. There is no ontologically given, theory-neutral arbiter of projectability. Thus, there is no pure induction. Abstraction and concept formation is always theory-driven.
Accordingly, concepts must be identified by the roles they play in theories (Carey 1985, 198). Representations of concepts are best thought of as theoretical knowledge or, at least, as embedded in knowledge that embodies a theory about the world (Murphy & Medin 1985, 298).
So what is a theory? And how do theories emerge in the first place?
Carey (1985, 201) points out that explanation is at the core of theories. Explanatory mechanisms distinguish theories from other types of conceptual structures, such as scripts. The cognitive psychologists' famous restaurant script tells us what happens and in which order when we go to a restaurant. But it does not explain why we pay for our food, for example.
Murphy and Medin also see explanatory relations and causal connections - 'underlying principles' in their choice of words - as the essence of theories. They note that "one might have a theory that could connect (to some degree) objects that seem to share very few features" (Murphy & Medin 1985, 298). But they disagree with Carey in that they accept also scripts as theories. After all, "scripts may contain an implicit theory of the entailment relations of mundane events" (Murphy & Medin 1985, 290). Indeed, even the restaurant script contains one kind of an explanation to Carey's 'why' question: we pay because we are expected or asked to do that.
In other words, the presence of explanation does not seem to be a sufficient criterion of a theory. What is more important, we'll probably never find a clear and sufficient criterion by following the approach taken by Carey, Murphy and Medin. These authors try to define theory by looking at knowledge and mental representations as self-sufficient bodies or things stored within the head of the individual. They fall prey to the cognitivist or Cartesian fallacy, exhibited by Nelson Goodman, too. Theory is conceived as an entity the individual 'has'. When a theory emerges or is acquired, it may be stored and begins to function as a filter or lens, constraining the individual's inductive projections. Such a constructivism is mere mental constructivism, worldmaking in the mind only.
This cognitivist conception is unable to say anything interesting about how theories actually emerge in the first place. Carey (1985, 200) takes recourse to a moderate innatism: "my guess is that the 'initial state' of human children can be described by saying that they are innately endowed with two theoretical systems: a naive physics and a naive psychology". Murphy and Medin (1985, 311) are even more vague: "it is certainly possible that children's prototheories of the functions, relations, and importance of objects have effects quite early" - but "exactly when they do is an empirical question".
In this respect, Nelson's contention that event scripts are the initial form of conceptual representation is much more advanced than the conceptions of Carey, Murphy and Medin. It avoids the dead end of innatism by acknowledging a simple but powerful idea: in the beginning there was an act.
At one point in their paper, however, Murphy and Medin step beyond the cognitivist confines. They take up Bulmer's (1967) anthropological study of the Karam of New Guinea who do not consider a cassowary a bird. Bulmer argued that that this is not merely because the cassowary does not fly, but because of its special role as a forest creature and its resulting participation in an antithesis in Karam thought between forest and cultivation. This antithesis is further related to basic concerns with kinship roles and rights. Myrphy and Medin (1985, 305) correctly note that "apparently, the Karam's theories about forest life and cultivation produce different classifications than do our culture's biological theories".
This conclusion implies that theory is no more seen as a self-sufficient entity within the idividual mind but rather as a social activity system in itself. In this view, theories and concepts can only be understood as the representational, secondary aspect of sensuous, material activity systems. This has nothing to do with the mechanical idea of theories as somehow direct copies of material objects. But theories live and develop only integrally embedded in activities. Theories may be separated from activities - forgotten and hidden in obscure books, for example - but contrary to Popper's view, this means that they are in effect dead or frozen, barren from life and development at least temporarily.
In Vygotsky's late work Thinking and Speech, the problem of concepts was central. Vygotsky rejected the traditional inductivist notion of concepts. He pointed out that for the traditional view concept formation is similar to Galton's composite 'family portraits'. These are made by taking pictures of different members of a family on the same plate, so that the traits common to several people stand out vividly while differing individual traits are blurred by the superimposition. For the traditional view, the totality of the common traits is the concept.
"One cannot depict the real process of concept formation in a more mistaken way than this logified picture.
It was found already long ago, and our experiments have shown in clearly, that concept formation in adolescents never conforms to the logical process which traditional psychology describes." (Wygotski 1977, 160-161; italics in the original.)
Vygotsky cites Vogel's and Bühler's findings according to which children do not start with mere particulars but use general concepts from the beginning. The child acquires the word 'flower' earlier than the names of various particular flowers. Or if it acquires first the name of a particular flower, say 'rose', it uses this word not only for roses but for all flowers. (Wygotski 1977, 162.)
Vygotsky concludes that concept formation is a two-way movement within a pyramid of concepts: from the particular to the general and from the general to the particular at the same time. This fundamental idea is further elaborated in an analysis of the relationship of everyday and scientific concepts.
"From the stanpoint of dialectical logic, our everyday concepts are not concepts in the proper sense of the word. They are rather general notions of objects." (Wygotski 1977, 150.)
This important statement implies that we have to work out and apply a logic qualitatively different from the traditional formal logic if we are to grasp the nature of genuine, scientific concepts. This demand was, however, never met in Vygotsky's own analysis. As a matter of fact, later in his book he states that "one can say that the logical side of this question has been fully treated and investigated" while the genetic and psychological aspect remains open (Wygotski 1977, 263). Thus, Vygotsky did not work out an alternative logic of genuine concepts.
According to Vygotsky, scientific concepts work their way downward from the general to the particulars. Everyday concepts develop the opposite way. As the two meet, they penetrate and transform each other.
There are three characteristics which make scientific concepts distinctive. Firstly, scientific concepts are always included in a conceptual system. Secondly, scientific concepts require that the learner is conscious of them; their formation begins with the word, with the definition, and the learner is required to work on the concepts themselves. Thirdly, scientific concepts are thus not acquired spontaneously but through instruction..
V. V. Davydov points out the weakness of this definition. First of all, even empirical concepts possess a system which may take the form of elaborate classificatory dependencies of the 'genus-species' type. Furthermore, such descriptive concepts, or 'general notions', are systematically transmitted in school instruction. As a matter of fact, they dominate the subject matter of primary school instruction. The two-way movement in a conceptual pyramid is fully possible within a purely empirical or descriptive structure of concepts.
"The acquisition which begins with the 'general' verbal definition by no means characterizes the scientific nature of a concept; also arbitrary everyday notions, empirical general notions can be transmitted this way in instruction." (Dawydow 1977, 162-163.)
In other words, Vygotsky could not solve the problem of the specific contents of scientific concepts. His definitions remained formal - a little like those put forward by Carey and Murphy & Medin more than 50 years later. Surely Vygotsky was right when he wrote that scientific concepts are systemic - but what is the specific quality of their systems? In a like manner, Carey, Murphy and Medin are right in stating that theories contain explanatory mechanisms or principles - but what distinguishes a theoretical explanatory mechanism from an empirical one?
Within the cultural-historical school, V. V. Davydov was the first psychologist who broke out of the confines of traditional formal logic in the problem of concept formation. The importance of this step has not been widely understood, and Davydov's fundamental work has still not been translated into English though it appeared in 1972. The far-reaching instructional implications of Davydov's work have often met with aggressive resistance and misinterpretation, both in his own culture and in the west.
But Davydov's achievement was made possible by certain advances in the philosophy and epistemology of dialectical materialism. The two works that had the strongest effect of Davydov seem to have been E. V. Il'enkov's (1982 [in Russian 1960]) book The Dialectics of the Abstract and the Concrete in Marx's Capital and the volume Analysis of the Developing Concept written jointly by A. S. Arsen'ev, V. S. Bibler and B. M. Kedrov (1967). As philosophical works, both books are exceptional in that they are based on detailed analysis of important developments in the history of science. The former investigates the formation of the concept of value in Marx's research in political economy. The latter analyzes the development of central concepts in mechanics and chemistry.
The point of departure in Il'enkov's work is a redifinition of the meaning of 'concrete' and 'abstract'. Contrary to the common notions, dialectics does not see 'concrete' as something sensually palpable and 'abstract' as something conceptual or mentally constructed. 'Concrete' is rather the holistic quality of systemic interconnectedness.
"(...) if consciousness has perceived an individual thing as such, without grasping the whole concrete chain of interconnections within which the thing actually exists, that means it has perceived the thing in an extremely abstract way despite the fact that it has perceived it in direct concrete sensual observation, in all the fullness of its sensually tangible image.
On the contrary, when consciousness has perceived a thing in its interconnections with all the other, just as individual things, facts, phenomena, if it has grasped the individual through its universal interconnections, then it has for the first time perceived it concretely, even if a notion of it was formed not through direct contemplation, touching or smelling but rather through speech from other individuals and is consequently devoid of immediately sensual features." (Ilyenkov 1982, 87-88.)
General notions are formal abstractions since they separate arbitrary features of objects form their interconnections. Genuine concepts are concrete abstractions since they reflect and reconstruct the systemic and interconnected nature of the objects. This systemic nature is not of the static classificatory 'genus-species' type but of a genetic and dynamic type. Il'enkov uses Marx's concept of the proletariat to illustrate this.
"When Marx and Engels worked out the concept of the proletariat as the most revolutionary class of bourgeois society, as the gravedigger of capitalism, it was in principle impossible to obtain this concept by considering an abstractly general trait inherent in each separate proletarian and each particular stratum of the proletariat. A formal abstraction which could be made in the mid-19th century by comparing all individual representatives of the proletariat, by the kind of abstracting recommended by non-dialectical logic, would have characterised the proletariat as the most oppressed passively suffering poverty-ridden class capable, at best, only of a desperate hungry rebellion.
This concept [better: general notion; Y.E.] of the proletariat was current in the innumerable studies of that time, in the philanthropic writings of the contemporaries of Marx and Engels, and in the works of utopian socialists. This abstraction was a precise reflection of the empirically general. But it was only Marx and Engels who obtained a theoretical expression of these empirical facts (...).
The concept of the proletariat, as distinct from the empirical general notion of it, was not a formal abstraction here but a theoretical expression of the objective conditions of its development containing a comprehension of its objective role and of the latter's tendency of development. (...)
The truth of this concept was shown, as is well known, by the real transformation of the proletariat from a 'class in itself' into a 'class for itself'. The proletariat developed, in the full sense of the term, towards a correspondence with 'its own concept' (...)." (Ilyenkov 1982, 130-131.)
In other words, the systemic nature of the genuine concept is essentially temporal, historical and developmental. The concept expresses the origin and the developmental tendency of the totality it reconstructs.
"To comprehend a phenomenon means to discover the mode of its origin, the rule according to which the phenomenon emerges with necessity rooted in the concrete totality of conditions, it means to analyse the very conditions of the origin of phenomena. That is the general formula for the formation of a concept (...)." (Ilyenkov 1982, 177.)
Moreover, the concept "expresses a reality which, while being quite a particular phenomenon among other particular phenomena, is at the same time a genuinely universal element, a 'cell' in all the other particular phenomena" (Ilyenkov 1982, 79). The task of genuine concept formation is thus to find out the developmental 'germ cell', the initial genetic abstraction, of the totality under investigation and to develop it into its full concrete diversity. Herein lies the kernel of the 'other logic' Vygotsky pleaded for but could never formulate. "The logical development of categories (...) must coincide with the historical development of the object (Ilyenkov 1982, 215-216; italics added)." In other words, we are not talking of an eternal and content-indifferent logic but of a developmental logic of the object itself. This logic is stored nowhere in the form of ready-made formulas to be imposed upon the object. To the contrary, "the concrete history of a concrete object should be considered in each particular case rather than history in general" (Ilyenkov 1982, 215).
In dialectical logic, the concrete is an interconnected systemic whole. But the interconnections are not of any arbitrary kind. At the core of the interconnections there are internal contradictions.
"Concreteness is in general identity of opposites, whereas the abstract general is obtained according to the principle of bare identity, identity without contradiction." (Ilyenkov 1982, 272.)
Contradictions become central if we are to handle movement, development and change conceptually.
"Any utterance expressing the very moment, the very act of transition (and not the result of this transition only) inevitably contains an explicit or implicit contradiction, and a contradiction 'at one and the same time' (that is, during transition, at the moment of transition) and 'in one and the same relation' (precisely with regard to the transition of the opposites into each other)." (Ilyenkov 1982, 251.)
The struggle and mutual dependency of opposite forces or elements is the developmental driving force within objective systems. To create a genuine concept is to grasp and fixate this inner contradiction of the object system and to derive the system's subsequent developmental manifestations from that initial contradiction.
"The dialectical materialist method of resolution of contradictions in theoretical definitions thus consists in tracing the process by which the movement of reality itself resolves them in a new form of expression. Expressed objectively, the goal lies in tracing, through analysis of new empirical materials, the emergence of reality in which an earlier established contradiction finds its relative resolution in a new objective form of its realisation." (Ilyenkov 1982, 262-263.)
For Arsen'ev, Bibler and Kedrov, a genuine scientific-theoretical concept is always the simple, initial germ of a whole complex theory. The characteristic of a genuine concept is its expansive "potency of concretization, tendency of developing into a theory" (Arsen'ev, Bibler & Kedrov 1967, 15). It tends to generate a multitude of successive developmental elaborations and conceptual offshoots out of itself. This view is actually opposite to that of Carey, Murphy and Medin who see concepts as products generated by initial theories.
But if concept is the initial form of a theory, how does the concept emerge in the first place? Here Arsen'ev, Bibler and Kedrov disagree with Nelson's inductivist view, according to which the initial scripts emerge as mental recollections of repeated familiar events. Arsen'ev, Bibler and Kedrov argue that the initial concepts emerge out of the interplay of two psychic processes constitutive in any practical productive activity: (1) the continuous construction of the anticipated future object (outcome) of the activity through active material and mental experimentation, and (2) the equally continuous sensuous or contemplative experiencing and observation of the object 'as it is'. In other words, the initial concepts are not just reproductions of events as experienced. Already from the very beginning they possess also the tendency of creating something not yet observed and experienced.
Arsen'ev, Bibler and Kedrov do not ascribe this potency solely to the concepts developed and used within the historically formed activity called science. "From our standpoint, any thinking and any concept is in its potentiality, i.e., in its essence, scientific-theoretical" (Arsen'ev, Bibler & Kedrov 1967, 14). Thus, so called everyday concepts have in principle the same expansive quality as the consciously elaborated concepts of science. A similar point is made by Il'enkov.
"It stands to reason that the universal laws of thought are the same both in the scientific and so-called everyday thinking. But they are easier to discern in scientific thought for the same reason for which the universal laws of the development of the capitalist formation could be easier established, in mid-19th century, by the analysis of English capitalism rather than Russian or Italian." (Ilyenkov 1982, 100.)
DAVYDOV AND THE PROBLEM OF CONCEPTS
Davydov characterizes the theoretical concept as follows.
"This type of concept functions as a completely specified and concrete means of connecting the general and the specific, a means of deducing particular and specific phenomena from their general basis. Due to this, the development of an object functions as the content of the theoretical concept.
The concept is a procedure of realizing a substantial generalization, a means of transition from the essence to the phenomena. It fixates the conditions and means of such transformation, such deduction of the individual from the universal." (Dawydow 1977, 305.)
Genuine concept formation and conceptual thinking ascends first from the perceptually concrete phenomena to the substantial abstraction, the 'germ cell' which expresses the genetically original inner contradiction of the system under scrutiny. It then proceeds to concrete generalization by deducing the various particular manifestations from this developmental basis. Following Hegel and Marx, this procedure is called ascending from the abstract to the concrete. Davydov points out that outside this process the concept becomes "a mere word" (Dawydow 1977, 308).
"To have a concept of an object means that one is able to use the general method of its construction, the knowledge of its origination. This method is a specific thinking activity of human beings which itself is formed as a derivative of object-oriented action reproducing its object of cognition.
(...)Thus, behind every concept there is a specific hidden object-oriented action (or a system of such actions), the discovery of which is a special research task." (Dawydow 1977, 309.)
Davydov summarizes the qualities of empirical and theoretical knowledge and thought in six points.
1. Empirical knowledge is produced by comparing objects and their representations which makes is possible to discern in them common general traits. Theoretical knowledge arises on the basis of an analysis of the role and function of a certain relation of things inside a structured system.
2. Comparison discerns the formally common trait which makes it possible to classify separate objects under a certain formal class irrespective of their being interconnected. By means of the analysis, the real, specific relation of things is found which is the genetic foundation of all other manifestations of the system. This relation functions as the general form or essence of the mentally reproduced totality.
3. Empirical knowledge, based on observation, reflects only external traits of objects and relies on perceptual notions. Theoretical knowledge, based on the transformation of objects, reflects their internal relations and interconnections. In the reproduction of an object, theoretical thinking exceeds the limits of perceptual presentations.
4. The formally common trait is separated from the particular features of the objects. In theoretical knowledge, the connection between the real general relation and its various manifestations, i.e., the connection of the general and the specific, is fixated.
5. The concretization of empirical knowledge consists in the gathering of illustrations or examples which belong to a formally derived category. The concretization of theoretical knowledge presupposes its conversion into a developed theory by deducing and explaining the specific manifestations from their general foundation.
6. The necessary means of fixating empirical knowledge is the word, the term. Theoretical knowledge is primarily expressed in the methods of intellectual activity and subsequently in various systems of signs and symbols, especially in artificial and natural languages. The theoretical concept may exist as a method of deducing the specific from the general before it has acquired a terminological formulation. (Dawydow 1977, 310-312.)
Davydov's argumentation suffers here from a dichotomous structure. Empirical thinking and theoretical thinking are presented as mutually exclusive alternatives. Their mutual dependency and mutual penetration are temporarily set aside. However, Davydov discusses this problem of complementarity earlier in his book.
"Man's sensuousness as objective-practical activity is inherently contradictory. Sensation and perception in themselves reflect things as immediately given. But through the practical action which brings things purposefully into contact with each other (object and tool), another content 'penetrates' into sensuousness - the mediated and interconnected character of things, their inner substance. The practical action as sensuous-objective action unifies in itself contradictory contents - the external and the internal, the immediately given and the mediated, the specific and the general." (Dawydow 1977, 261-262.)
Later, both phylogenetically and ontogenetically, these original moments of practical action are differentiated into separately identifiable fundamental modes of thought, the empirical and the theoretical, or the classificatory and the experimental. Still, neither of these can be conceived of as fully independent of the other. Both modes contain latent forms or seeds of the other. This does not mean that they are developmentally on the same level. To the contrary, it is theoretical thought that contains the instrumentality necessary for expansive development, for the production of the new.
Davydov's central point is theoretically compelling. And it leads to practical consequences.
"From a logico-psychological point of view, a person's true understanding of a subject can be shown by the ability to reproduce and demonstrate to another person the entire process of its origin. In the case of the concept of number, this means that a student should be able to demonstrate independently to a teacher, using appropriate actions upon objects, why it is both possible and necessary to form this concept. Further, the student should also be able to utilize the numerical properties of any quantifiable set for any specified purpose. For example, whether or not a child understands the concept of number can be shown by the proper execution of tasks like the following:
1. Require the child to pour into a second container the same amount of water provided in a first container that differs is form from the second. (The first container is a narrow, graduated cylinder, the second a wide-mouthed glass.) A child who can really isolate the conditions for obtaining a number, that is, who really understands its meaning, should use some intermediate measure, such as a small glass, to determine the amount of water the narrow cylinder contains (for example, five small glasses) and then pour the same number of glasses into the wide-mouthed glass.
2. Require a child to determine how many large glasses of water are contained i a series of three large and four small glasses if a small glass is equal to one half of a large one. Here, the child must count two small glasses as one large one and obtain the result of five.
3. Using a single set of blocks, require the child to determine various conditions under which several different numerical attributes would be defined. In this task, the child must construct equal groups of blocks and then use those groups as a unit of measure to determine different numbers. For instance, if 24 blocks are grouped by twos, then the number 12 will be expressed; if grouped by fours, then the number will be six; and so on.
4. Require a child to show how, using a single volume of water in a glass, different numerical descriptions of that same volume of water can be expressed. This task is similar to Task 3 but uses a continuous quantity instead of discrete objects. Different measures (for example, different sized small glasses) must be used to determine several different numbers.
For each of these tasks, the child must recognize the multiple relationship that can exist between a continuous or discrete object (as expressed by its numerical measure) and some part of that object that has been used as the unit of measure. In so doing, it is of particular importance that the child realize the arbitrary nature of the size of the part (the unit of measure) that is used to determine the measure of the entire object. When measuring, the child should be able to exchange one unit size for another and thereby determine different measures for the same object. In this exercise, the child needs a clear understanding of the origin of numerical measure to generate various concrete numerical representations of the object. Only when a child can carry out these fundamental steps can one speak of the child's understanding of number as a general mathematical method of expressing quantitative relationships within and between objects.
(...) Initially, we found that a majority of children enrolled in traditional programs could not carry out these tasks. For instance, in the first and fourth tasks they had no idea of how to proceed. In the second task they counted each glass, large or small, as a separate unit and thus obtained an answer of seven rather than the correct response of five. In the third task they counted the blocks singly to obtain 24 and were not able to group out any other unit of counting." (Davydov 1982, 225-227.)
Although these children were able to use a limited notion of number to deal with day-to-day and school problems, they really did not exhibit a true mathematical understanding of the number concept. This was due to the teachers' use of 'familiar' numbers as the starting point for instruction within the traditional program. On this basis first-grade children quickly proceeded to addition and substraction of numbers known to them only on an experiential basis. Davydov cites the famous mathematician Kolmogorov (1960, 10): "Divorcing mathematical concepts from their origins in teaching results in a course with a complete absence of principles and with defective logic."
In recent cognitive psychological research, interest in so called mental models has increased notably. Alone in 1983, two major volumes appeared under the title Mental Models (Gentner & Stevens 1983; Johnson-Laird 1983). In their review, Rouse and Morris (1985, 7) propose the following definition of mental models: they are mechanisms whereby humans are able to generate descriptions of system purpose and form, explanations of system functioning and observed system states, and predictions of future system states. Norman (1983, 7) adds an important point.
"Mental models are naturally evolving models. That is, through interaction with a target system, people formulate mental models of that system."
Norman distinguishes mental models from conceptual models. The latter are consciously invented by teachers, designers, scientists and engineers. But it remains unclear whether conceptual models are also mental - or perhaps non-mental. The difference seems to be merely one of the degree of consciousness and presentational rigor.
In many ways, the recent discussion on mental models is a new version of the 'model muddle' prevalent in philosophy during the 1960's. Wartofsky summarizes the muddle as follows.
"In much of model-talk, models inhabit a limbo between worlds. On the one hand, they are not citizens of the blood-and-guts world of real objects and processes; or at best have only a derived citizenship by way of their reference to such a world. On the other hand, they are denied full equality in the cognitive world of purported truths, assigned only the function of instruments of such cognition: crutches, aids to the imagination, inference-machines, heuristic devices, data-ordering frameworks and whatnot." (Wartofsky 1979, 3.)
The problem with the cognitive psychological notion of mental models is that it is static, dead in a twofold sense.
Firstly, mental models are conceived of as something evolving spontaneously within individual heads, on the basis of individual experience. This evolution consists of two basic processes (De Kleer & Brown 1983, 156): constructing or envisioning the mental model; and simulating the result of of this construction or running the model. However, both these processes are cut off from the construction and use of external, material, socio-cultural models. How these external, objectified models are generated and how they interact with individual mental models remains a mysterious sphere outside the interest of mainstream cognitive psychology. This isolationist mode of inquiry renders the mental models of cognitive psychology mere filters, slowly renewed objects of consumption. Models are deprived of their productive and instrumental aspect.
Secondly, in consequence of the first delimitation, there seem to be no satisfactory ways of assessing the qualitative level or type of a mental model. A host of different classifications and typologies have been offered, but each one of them seems to be equally arbitrary. The reason is that the classifications and typologies have no historical basis. As long as the historical steps of the societal production of models remain obscure, psychologists are bound to keep inventing their private favorite typologies ad nauseam. They will also remain incapable of foreseeing and enhancing the necessary future qualities of models.
Earlier in this chapter I noted that expansive thinking demands that the consumptive objects of thought are transformed into productive instruments of thought. Representational concepts must be transformed into instrumental concepts. This transformation requires a specific type of objectivity-instrumentality. Models are specifically simplified and 'purified' reconstructions of the perceptual-concrete object, created for the purpose of gaining unexpected information or working out unforeseen potentialities of the object. Models are an integral moment of experimentation. Being transparent and compact at the same time, models function both as projections and as means of constructing and realizing the projections (Dawydow 1977, 260-261).
Wartofsky sees models much in the same way. For him, a model is not simply the entity we take as a model but, potentially, rather the mode of action that such an entity itself represents. In this sense, "models are embodiments of purpose and, at the same time, instruments for carrying out such purposes" (Wartofsky 1979, 142).
Models are the specifically theoretical or expansive mode of ideality. The ideal exists only in man. But man is to be understood not as one individual with a brain, but as a real aggregate of people collectively realizing their human life activity, as the aggregate of social relations arising between people around the process of the social production of their life. Only in this sense is the ideal inside man.
"(...) 'inside' man thus understood are all the things that 'mediate' the individuals that are socially producing their life: words, books, statues, churches, community centres, television towers and (above all!) the instruments of labour (...).
The ideal form is a form of a thing, but a form that is outside the thing, and is to be found in man as a form of his dynamic life activity, as goals and needs. Or conversely, it is a form of man's life activity, but outside man, in the form of the thing he creates. 'Ideality' as such exists only in the constant succession and replacement of these two forms of its 'external embodiment' and does not coincide with either of them taken separately. It exists only through the unceasing process of the transformation of the form of activity - into the form of a thing and back - the form of a thing into the form of activity (of social man, of course)." (Ilyenkov 1977a, 98.)
I suggest that models as the specificially theoretical type of ideality may be fruitfully analyzed from two angles: the functional and the historical.
THE FUNCTIONING OF MODELS IN THEORETICAL THINKING - PRESENTED AND QUESTIONED
From the functional angle, three general steps of model construction and application may be identified. Theoretical thinking starts with the constitution of its object. The object of inquiry is delineated with the help of available previous knowledge concerning the problem domain. This constitution of the object often takes place in a tacit fashion, without the individual's conscious effort, as an unreflected projection of the social conventions and relations in which the individual is embedded. However, the object is never just there, without constitutive actions of the subject - without being identified and named. This first step of object constitution or problem identification may be depicted diagrammatically as follows (Figure 4.1).
Now this, often tacit or implicit, step does not discriminate between theory construction and any everyday problem solving. Theoretical thinking differs from other types of thinking in that it constructs a model of the object, attempting to uncover and make visible the hidden relations or regularities behind the observable behavior of the object. This model construction is achieved with the help of analogy: "thus, at the heart of a theory are various modelling relations which are types of analogy" (Harré 1970, 35).
Analogy as an instrument is closely related to play and imagination. In both, the subject is making visible the 'rules of the game' or the hidden relations of the object transparent and visible through various forms of practical and mental experimentation.
This second step of theory construction is a step to the realm of 'secondary processes' in Bateson's (1972, 185) terminology, i.e., a step to consciously externalized and objectified abstractions. This step is diagrammatically depicted in Figure 4.2.
A model is not yet a full-blown theory. The theoretical model may be considered as an instrument for developing and applying the theory at the same time. The model invites and provokes thought experiments and concretizations. As Wartofsky (1979, 142) says, it is "a creation of something working toward the future". In this working toward the future, the subject not only elaborates the object with the help of the model. He also elaborates the model, modifies it into new, more complex developmental forms and variations. In other words, he builds a theory on the basis and with the help of the model. This is the third step of theory construction proper (Figure 4.3). In this view, a theory is an active, evolving relationship of the model to the things the model is supposed to
represent. In its embodiments, it takes the form of statements, categories, rules and procedures.
The stepwise process described above is neat and compact. It corresponds to the manner the process of ascending from the abstract to the concrete is often depicted in Marxist literature: as an essentially individual and mental process of expansion. However, it is too clean and regular to account for the cognitive-instrumental aspect of the ruptures involved in the creation of societally new activity structures.
A glance at Figures 4.1 to 4.3 reveals that thinking is here restricted to a solitary process.The subject remains individual. No community is involved. Similarly, the object remains the same from the beginning to the end. There is no structural expansion in these corners. Furthermore, theory as the end product of the process consists of new representations of reality; change in the reality itself is not implied.
To summarize, the steps described above depict theoretical thinking as a series on individually situated mental actions. The process corresponds to that of Learning IIb, as described above in Chapter 3.
What remains to be explained is the qualitative transition from series of individual, mental actions to a new collective, material activity system. In diagrammatic terms (Figure 4.4):
The general cyclic model of this expansive transition, or zone of proximal development, has been presented in Figure 3.3 in the preceding chapter. However, we are now dealing with the specific cognitive instruments needed for the conscious or intuitive mastery of the transition. So far, I have indicated that a new conception of concepts is required. I have also indicated that models play a special role within this new conception of concepts. But in order to characterize the cognitive instruments and their functioning more concretely, the successive dominant historical forms of the transition have to be analyzed. In other words, my functional analysis will necessarily acquire a historical dimension. The strict boundary between functional and historical analysis must be tendentially overcome.
To enter this kind of functional-historical analysis, I shall reinterpret B. M. Kedrov's (1966-67; 1972) famous historical account of the discovery of the periodic law of the elements by D. I. Mendeleev in 1868.
THE DISCOVERY OF THE PERIODIC LAW AS AN INSTANCE OF EXPANSIVE TRANSITION
D. I. Mendeleev discovered the periodic law in 1869. As an extraordinarily accurate person, he kept and stored without exception all documents and rough notes related to his work, even relatively minor and insignificant ones. From the late 1940's, Kedrov has conducted intensive investigations in the D. I. Mendeleev Museum at the University of Leningrad to reconstruct the course of the great discovery on the basis of these archive materials.
Mendeleev's discovery can be divided into four periods. Firstly, there was a preparatory period of about 15 years (1854-1869). Secondly, the discovery itself took shape within one day, February 17, 1869. Thirdly, the discovery was elaborated and refined during a period of approximately three years (February 1869 - December 1871). Finally, Mendeleev used the remaining 35 years of his life for the less intensive and condensed tasks of proving and completing the theory and pushing it through in the scientific community.
The activity system under our scrutiny here is that of the 'invisible college' or community of chemical researchers in the second half of the 19th century. Typically to science as universal labor, this activity system consisted of extremely distributed parallel working units. But these were still relatively autonomous and independent of each other.
"The first level, on which the overwhelming majority of chemists of this time stood, amounted to sorting the elements into natural groups ('specific') without relating them in a single unity. The second level involved laying bare the general law relating all of the elements ('general'), hence relating the groups in which they were already classified." (Kedrov 1966-67, 33.)
The barrier preventing this expansive breakthrough from the specific to the general level consisted of the relative encapsulation of the standard procedure.
"The grouping of elements according to their specific features became a tradition and stabilized itself in the consciousness of chemists. It finally became the strongest hindernis to the further development of the science (...). In fact, the grouping of elements according to the specific features requires that only chemically similar elements are compared and associated with each other, while chemically dissimilar and especially chemically opposite elements are not compared and definitely not associated with each other.
Contrariwise, the transition to the general level, i.e., the discovery of a general natural law covering all elements (...), necessarily requires the association of not only similar but importantly also of dissimilar elements." (Kedrov 1972, 88.)
The inductive tradition made it impossible to use the atomic weight, at that time the only known feature common to all elements, as the unifying basis for constructing a comprehensive system of the elements. It would have brought chemically dissimilar, even polarly opposite elements next to each other.
As the knowledge of the elements and their specific natural groups increased and became technically easier to obtain (electrolysis, spectroscopy), the disadvantages resulting of the lack of a general principle for arranging the elements gradually became visible.
"By the sixth decade of the nineteenth century, chemistry had reached such a stage that chemists ought to have discovered and brought about, by some means, a shift from the first level to the second. This was the task placed before chemistry by the objective line of development of the science itself." (Kedrov 1966-67, 33.)
As a matter of fact, at least two other scientists, Newlands in England and De Shancourt in France, came very close to the discovery at the same time as Mendeleev.
This general need state within the activity of chemical research was specifically aggravated in the case of Mendeleev. At the time of the discovery, he was writing his major textbook Fundamentals of Chemistry.
"The first part of this work was completed at the end of 1868, its final chapters being devoted to the group of very strong non-metallic haloids (halogens). Directly after the haloids followed the group of very strong metals - alkaline metals - to which the author allotted the first two chapters of the second part of his work.
It can be assumed that by the middle of February 1869 both chapters were finished, and the task confronted the author, with all insistence, of deciding which group of elements should follow the alkaline metals in the book. But to decide this it was necessary to elucidate which metals adjoined the alkaline closest of all. (...) To answer it, it was necessary to find some general principle according to which the elements should be arranged in their groups in a definite order (...)." (Kedrov 1966-67, 19.)
Mendeleev's chemical research activity may be characterized with the help of the triangle model developed in Chapter 2.
Now this personal aggravation of the general need state was not enough. A foreign element entered the system of research activity and intensified the conflict into a secondary contradiction. Parallel with his research, Mendeleev was passionately involved in co-operative programs of practical agricultural development. He had planned to carry out an inspection trip to some dairy artels in central Russia between February 17 and February 28, 1869. However, having just finished the two first chapters mentioned above, Mendeleev's mind was intensely preoccupied with the problem of the continuation of the book. At the same time, the departure for the inspection trip was coming closer.
"Thanks to these purely accidental coincidences, on the 17th of February, unexpectedly for Mendeleev, both lines of his activity during this period came in conflict and crossed: first, writing the Fundamental of Chemistry and, second, his trip to the dairy co-op. Since the trip was agreed upon with the interested organizations, Mendeleev could not avoid his obligation to go on a specific day. This circumstance strictly limited the time he could set aside for solving the problem confronting him (...). In other words, Mendeleev achieved the discovery of the periodic law under conditions of the most severe Zeitnot (time pressure), which gave rise to a very distinctive character and path in its development. The general psychological situation of Mendeleev on the day of the discovery can be compared with the situation of a chess master, caught at the very beginning of a game in Zeitnot, but striving at all costs to achieve a victory in spite of the unfavorable conditions." (Kedrov 1966-67, 20.)
Thus, the foreign element that entered the research activity was a new rule, namely that of a time limit. This, however, is still an insufficient picture of the conditions of the discovery. Mendeleev actually found quickly a partial, half-intuitive solution to the particular problem concerning the next chapter of his book.
"When Mendeleev found the answer to the question that had interested him - what group of metals should be treated after the alkaline metals in the Fundamentals of Chemistry - he did not regard his work as finished. (...) The concern was now with carrying out to the end the discovery of the lawfulness, already found in the first approximation. (...)
However, the method initially selected for constructing the table of elements by entering elements in it successively (one after the other), although it was successful in the first stage of discovery, turned out to be inapplicable for the whole set. The point is: while Mendeleev was operating on the well-known elements, all of them, with few exceptions, took their places in the table; even if their places had to be changed subsequently, such failures were few and did not obscure the whole picture of the organization of the elements which at any moment were included in the table. But when Mendeleev tried by these same means to find a basis for including in the table the poorly studied elements, the number of necessary corrections, transpositions, and deletions became so great that it began to interfere with the progress of the discovery. To recopy from the beginning in every case the incomplete table (...) was practically impossible. This would have taken so much time that one could not think of completing the whole work in a single day (for he was still to go out to the cooperatives on the following day). The Zeitnot (...) required finding a more convenient method for quickly carrying to completion the developing discovery." (Kedrov 1966-67, 23-24.)
This situation, the aggravated contradiction between an emerging idea and the lacking instruments for its formulation and elaboration, was sharpened to a point where symptoms of a double bind appeared.
"Calling on Mendeleev, it would seem, at just this moment, his friend A. A. Inostrantzev found Mendeleev in a gloomy, depressed state. According to Inostrantzev, Mendeleev began to speak of what was subsequently the embodiment of the periodic system of elements. But at this moment the law was still not formulated and the table still not completed. 'It's all formed in my head,' said Mendeleev with bitterness, 'but I can't express it in the table.'
(...) Mendeleev himself (...) wrote in his diary that after a period of enthusiasm he sometimes fell into a sudden slump, or even depression, ending sometimes in tears." (Kedrov 1966-67, 24.)
How did Mendeleev break the double bind? Here, quite an interesting, seemingly accidental analogy functioned as the springboard.
"It should be mentioned that Mendeleev loved to play the game of patience, where the thoroughly shuffled cards must then be rearranged according to definite rules, resulting in a definite pattern of disposing them by suit and denomination. The analogy with the distribution of elements turns out to be nearly complete; for at the moment when he considered this problem, two incomplete tables of elements were already written down on paper, and in them was already clearly charted a distribution of elements in two dimensions: horizontally, according to their general chemical properties or chemical similarity (which corresponds to arranging the playing cards according to suit), and vertically, according to the closeness of their atomic weights (which corresponds to arranging the playing cards by denomination)." (Kedrov 1966-67, 24.)
The springboard thus consisted of a technique and an image taken from a recreational activity quite remote from research work but thoroughly familiar to the subject. This kind of association may look purely accidental and arbitrary. But that is not the whole truth. Basically the same analogy had earlier been used in another problem by the famous scientist Gerhardt. Gerhardt drew the parallel between arranging cards by suit and denomination, on the one hand, and arranging organic substances in homologous and genetic series on the other. Mendeleev counted himself one of Gerhardt's convinced adherents, and of course was acquainted with this earlier application.
With the help of this springboard, Mendeleev constructed his famous 'chemical patience' which quickly "grew into a general picture of the future system of elements in its completeness" (Kedrov 1966-67, 26). In other words, the new general model of the object of chemical research activity was formulated.
The course of the discovery may now be summarized with the help of a table (Table 4.3) similar to those presented in the cases of Huckleberry Finn (Table 3.3) and Seven Brothers (Table 3.4).
Table 4.3 graphically reveals the problem peculiar not only to Kedrov's account but to most descriptions of scientific discoveries. The transition from the singular and specific to the general is followed only half way, to the point of the formulation and modelling of the new law or principle. But this is not yet the true level of generality. How does the new general model transform the structure and content of the practical scientific research activity in question? What is the nature of the tertiary and quaternary contradictions? These questions are left open, as if they were considered unessential for the understanding of scientific creativity. In my analysis, these questions should be recognized as all-important.
There is a reason for the general omission of these phases from the accounts of discoveries. A 'classical' discovery, such as that of Mendeleev's, is typically made by an ingenious craftsman-like individual scientist. Such a discovery, at least its intensive course, actually seems to terminate at the point where the individual craftsman-scientist publishes his revolutionary findings and, metaphorically speaking, hands them over to the scientific community (and indirectly to the society in general) for judgement and eventual application. The discontinuous nature of this historical type of transition makes it hard to realize the tremendous potential embedded in the emergence of the created new through the tertiary contradictions.
Table 4.3
The sequential structure of the discovery of the periodic law
__________________________________________________________
CONTRADICTION PHASE CONTENT ACCORDING TO KEDROV
Primary within Need state Generally: the inductive classificatory
the components tradition vs. the need to master the growing
of the old activity amount and complexity of the elements
Individually: the choice of the group of
elements for the next chapter of the book
Secondary Double bind The intruding new rule (time limit) vs.
between old instruments (inductive classification,
the components serial one-by-one procedure)
of the old activity
Object/motive The idea of patience as the springboard
construction New object: all elements in a comprehensive
system
New general model: the periodic law,
embodied in the periodic table
Tertiary Application, ?
between the old generalization
and the given new
activity
Quaternary Activity 2: ?
between the new reflection,
activity and its consolidation
neighbour
activities
__________________________________________________________________
But does it make any sense to talk about the given new in the case of a great scientific discovery? Isn't it all created new?
My contention is that scientific discoveries just like expansive developmental transitions in more mundane activity systems are to a large extent achievements of synthesizing and crystallizing elements that were already 'there'. In Mendeleev's case, atomic weights were already known. Surely in this case the given new is different from that of the seven brothers, for example. Science as universal labor produces strong generalizations. But the most dynamic and revolutionary aspect of scientific discoveries resides in the unexpected questions and ideas they evoke while being assimilated, argued against, generalized and applied. The psychologist or historian studying scientific creativity is usually interested in the creative individual. He thus looses track of the expansive development as soon as the subject of the process is no more just the individual genius but a collective or several collectives.
As Mendeleev's creative process reached its intensive phases, a new rule - the time limit - entered the lower left corner of the triangle in Figure 4.3. To facilitate the solution of the contradiction between the new rule and the old instruments, a provisional new instrument, namely the patience, appeared in the uppermost corner in the function of a springboard. These new prerequisites led to an expansive transition where there was a qualitatively new outcome of Mendeleev's actions: not just new specific classificatory knowledge about the elements but a totally new general principle for the understanding of their relations - the periodic law. This outcome was transformed into a new kind of general instrument, giving eventually rise to a qualitatively new developmental form of chemical research activity.
This historical type of activity and expansive transition corresponds to the classical ideal of university research. The ingenious individual scientist and his selfless striving after pure truth seem to be the prime movers behind great discoveries. In modern days, Michael Polanyi (1964) has made this type of transition into the eternal model of all research work. Polanyi's conception of science as activity may be summarized with the help of Figure 4.4. The noteworthy feature of this model is the lack of internal contradictions. Pressures toward change are seen as external threats, not as manifestations of the inner dynamics of research activity.
Drawing directly upon Polanyi, Jerome Ravetz (1971, 103; emphasis added) concludes that "in every one of its aspects, scientific inquiry is a craft activity, depending on a body of knowledge which is informal and partly tacit".
