From owner-bhaskar@jefferson.village.Virginia.EDU Sun Jan 25 17:51:33 1998 Date: Sun, 25 Jan 1998 15:51:18 -0700 From: Hans Ehrbar <> Subject: BHA: rts2-34 The Logic of Scientific Discovery 185 4. THE SOCIAL PRODUCTION OF KNOWLEDGE BY MEANS OF KNOWLEDGE The basic conception of scientific activity that I have been concerned to advance here is that is (consists in or involves) *work*. Science, I have argued, must be conceived as an ongoing process of transformation, continually or essentially in motion, in an attempt to capture (i.e. penetrate and describe) the stratification of the world. The logical structure of work is Aristotelian. It depends, in particular, upon the co-presence in any given productive episode of both a material and an efficient cause. Science operates on given materials, including pre-existing theory and antecedently established facts, with given materials, i.e. by means of an ensemble of intellectual and technical tools (including among the former paradigms, models, metaphors and analogies), producing new theories and facts. Science is produced by the imaginative and disciplined work of men on what is given to them. But the instruments of the imagination are themselves provided by knowledge. Thus knowledge is produced by means of knowledge. The objects from, and by, which knowledge is generated are thus always themselves social products (as is the knowledge generated). Thus science as a process is always entirely intrinsic to `thought'. However, by perception and experiment access to objects, viz. things and causal structures, existing independently of thought may be obtained.58 And of such objects knowledge may be achieved. Science is not an epiphenomenon of nature, for knowledge possesses a material cause of its own kind. But neither is nature a product of man, for the intelligibility of the scientific activities of perception and experiment presupposes the intransitive and structured character of the objects of knowledge, viz. that they exist and act independently of the operations of men and the patterns of events alike. 58 As the immediate objects of perception are normally assumed to be short events or momentary states, perceptual access to things presupposes a resolution of the problem of induction (to be discussed in paragraph 6 below). 186 A Realist Theory of Science Thus science, I have argued, presupposes the ontological independence and the possible disjuncture of the domains of the real, the actual and the empirical at every stratum or level of reality. At each stratum scientists attempt to identify the entities responsible for what happens at the less fundamental stratum (their point of departure) and describe *their* normic behaviour. But knowledge of existence, I have argued, cannot be identified with demonstration of it. Causal powers, for example, can only be known, not shown to exist. Hence if, as I have suggested there are grounds for supposing, the ultimate entities in any one branch of science are bare powers, they must necessarily be undemonstrable. However, under certain conditions, some states of things may be perceived, unaided or with the help of sense-extending equipment; and some causes may be demonstrated indirectly, i.e. through the ostension of their effects. But for an existential or a dispositional claim to be confirmed or corroborated the states (or effects) and behaviour must be recognised or identified as being of the asserted type. Hence in general two kinds of criteria, viz. demonstrative and recognitive, must be satisfied for such a claim to be granted. Because the theoretical and technical conditions under which such claims are made (and criteria elaborated) are themselves developing, our knowledge may be extended; and because they may be falsified if the criteria are not met (or revised) our knowledge may be corrected. The paradoxical air of talking of the correction of knowledge vanishes once the demand for extra-theoretical truth and intertheoretical synonymity is rejected. Progress, I shall contend, can be shown to have occurred but only from some particular position, some specific vantage point, as it were, in theoretical time. Science is explanatory, not simply descriptive. Explanation is achieved by reference to enduring mechanisms. Such mechanisms exist as the powers of things and act independently of the conditions that enable us to identify them. Thus there is a direct link between the dynamic realist thesis that the things and causal structures of nature not only exist but act independently of men and the conception of science as a social activity sui generis in which both the facts and the conjunctions that, when attainable, provide the empirical grounds for causal laws are The Logic of Scientific Discovery 187 seen as social products. In classical empiricism, in a subtle interchange, these ideas are crossed: so that facts and their conjunctions appear as naturally given and things and causal structures as experiences of men. (In transcendental idealism, the former is seen as in part imposed by, and the latter as unknowable to, men.) Now the identification of the conditions of (knowledge of) being with the conditions of experience in empirical realism leaves `theory' with a very uncertain status. For it must be either reduced to, or grounded a priori in some necessary condition of, experience; so that it is either reducible or immutable. For transcendental realism theory is both irreducible and mutable. It is always there and liable to change, as part of our socially innate intellectual endowment. It is this endowment that we must draw upon as we attempt to deepen our knowledge of the way things are and act in the world; and in so doing we can continually add to and modify it. The existence of this stock, as a layered structure, is a necessary feature of any human cognitive situation; so knowledge can never be seen as a function of individual sense-experience. The necessity for a scientific training shows that knowledge is a social product and cannot be conceived as purely individual acquisition. For it always stands to the individual as something that must be acquired to be used (for scientific work). That science is ongoing implies that some individuals do so. Knowledge shares a feature common to many social products then: namely that though it exists only in virtue of human activity, it is irreducible to the acts of men. For any cognitive act to be possible there must be a material cause; some knowledge established, given to us, already produced. No sum of individual cognitive acts can yield knowledge, for the first member of the series would already presuppose it. Experience is, on the other hand, susceptible to a purely individualistic analysis; `mass experience' is clearly derivative and analysable as `the experiences of masses of individuals'. It can thus be seen that underpinning empirical realism is an epistemological individualism. That knowledge is not analysable in terms of individual experiences does not imply that it is not analysable in terms of experience. But that the latter is the case can be seen by reflecting upon the consideration that the antecedent cognitive situation of the individual would have at the very least to contain one theoretical 188 A Realist Theory of Science conjecture, viz. that there were experiences of others. Assuming that the category of experience was allowed to apply intersubjectively, if all terms were explicitly (ostensively or operationally) defined all truths would be analytic and all falsehoods contradictory, as a claim about the facts would be implicit in the meaning of every theoretical conjecture, so that the point of appealing to experience would be lost; and theory would be incapable of growing and developing. In short, if antecedently established knowledge is to be capable of functioning as a material cause, the layered stock must contain some terms not completely definable in terms of experience; i.e. it must consist of a web of empirical and theoretical ideas. Recent work establishing that science has a transitive (or sociological) dimension and some facts about its nature has been widely regarded as shocking. That science is a social activity which shares many of the characteristics of, and does not exist in isolation from others; that it depends upon a whole complex of institutions, some of which have little interest in knowledge `for its own sake'; and that in particular circumstances its fortunes can depend upon the accidents of particular men raises serious moral and political questions. Moreover there are some particularly disturbing features about current science. One need only invoke, from the recent literature, the epithets of entrepreneurial science and shoddy science, reckless science and dirty science,59 government science60 and mob science,61 repressive science, Stalinist science and their anodyne anarchist science62 to appreciate this. However these problems do not flow from the social character of science per se, i.e. the mere existence of a transitive dimension, but from the present character of its social character. The realization that science has social problems could only be shocking if one had been tacitly viewing it, in the style of Hume, as a kind of behavioural 59 J. R. Ravetz, op. cit., esp. pp. 47-59. 60 N. Chomsky, `Objectivity and Liberal Scholarship', *American Power and the New Mandarins*, pp. 23-129. 61 K. R. Popper, `Normal Science and Its Dangers', *Criticism and the Growth of Knowledge*, eds. I. Lakatos and A. Musgrave, pp. 51-8. 62 P. K. Feyerabend, `Against Method', *Minnesota Studies in the Philosophy of Science, Vol. IV*, eds. M. Radner and S. Winokur, pp. 17-130. The Logic of Scientific Discovery 189 response to the stimulus of given facts and their conjunctions. This is the positivist concept of behaviourist or automatic science. It is a concept which can itself be used to disguise embarrassing facets and rationalize the practice of a science. Recognition of the transitive dimension implies that scientific beliefs can no longer be distinguished by their content. For experiences and the facts they generate must now be viewed as socially produced and what is socially produced is socially changeable. There are no absolutely privileged statements. The application of the category `empirical' becomes relative and theory-dependent. Hence it cannot be used, without a degree of circularity, to establish the scientificity of one class of statements with respect to another. Knowledge, viewed as a transitive process, has no foundation - only a structure in time. The sciences have histories, which like all histories are characterized by both continuity and change; and in which, as in all histories, certain events stand out retrospectively as especially significant, e.g. the discovery of oxygen, the publication of *The Origin of the Species*, the Michelson-Morley experiment. (Later we shall have to inquire into the grounds for their significance.) Now the fact that scientists do not possess a special attitude or a superior morality does not mean that science does not have a rationality of its own. Nor does the fact that scientific beliefs cannot be distinguished by their content imply that scientific activities cannot be distinguished by their structure or their aim. There are two errors here: the first is to suppose that science is not a social activity in the fullest sense (exactly what this entails we have yet to see). The second is to suppose that it is not equally a social activity, quite unlike any other, sui generis: namely a social activity whose aim is the production of the knowledge, with the cognitive tools at its disposal, of the enduring and transfactually active mechanisms of the production of phenomena in nature. In paragraph 3 I described science as the systematic attempt to capture the stratification of the world. Only the concept of a real stratification allows us to sustain the idea of scientific progress, in a way which is both non-inductivist and consistent with the possibility of scientific change. Knowledge of new strata does not dissolve, though it may occasion a correction of, knowledge of old strata. Nor does it render the less fundamental strata 190 A Realist Theory of Science illusory. We do not need the metaphor to which so many writers have found it necessary to resort in order to reconcile progress with change: viz. that of an asymptotic approach to the truth. Two fallacies must be most assiduously avoided. The first is to suppose that science grows but does not change. The second is to suppose that science changes but does not grow. It is the fact that science grows, i.e. that in the transitive process of science new strata and dimensions (or branches) of reality are discovered, that means that scientific change can be accommodated as a fact of history without sacrificing the idea of scientific progress. But progress can only be shown to have occurred from some substantive theoretical standpoint or position. There is no Archimedean point outside theoretical time. But knowledge changes as it grows. For knowledge at a new level may lead to a revision, correction or modification of knowledge at the previous level. For what is explained is never the `pure' phenomena, but always the phenomena read in a certain way: i.e. facts. The scientist seeks to describe the mechanisms generating the phenomena; but the results of his activity belong to the social world of science, not the intransitive world of things. Does this mean that it is wrong to talk of the scientist explaining events describing mechanisms, etc.? No: provided we remember that what is explained in any concrete scientific episode is always the event known under a particular description. This does not mean that the event is, or that we must think of it as if it were its description. On the contrary, the ontological independence of the event is a condition of the intelligibility of its description. But here, as elsewhere, it is the task of philosophy to analyse concepts, such as that of an event, which can only be used syncategorematically in science. Scientific activity is continuous. This has meant that `refutations' have normally taken the form of `replacements'. Now it has been pointed out that cases of both `inconsistency' and `meaning-change' can be drawn from the history of the sciences. For example, Newtonian physics corrected Kepler's and Galileo's laws;63 and the concepts of `mass' employed in classical dynamics and the theory of relativity are radically 63 See e.g. P. Duhem, op. cit., Chaps. 9 and 10, or K. R. Popper, `The Aim of Science', Objective Knowledge, Chap. 5. The Logic of Scientific Discovery 191 different.64 Now given that these are the most obvious ways in which scientific changes occur, great care must be taken about the way in which they, and their relationship, are formulated. Kuhn, Feyerabend, and others, have claimed that theories may be so radically different in meaning as to be literally `incommensurable'. To this there is the obvious objection that if they were literally incommensurable, i.e. shared no elements of meaning in common, it is difficult to see how scientists could have had grounds for preferring one to another. It is clear that at the moment of `falsification', when one theory is replaced by another, some elements of meaning must be shared in common.65 But that the subsequent divergent development of the theories may result in their eventually becoming `incommensurable'. So that `inconsistency' and `incommensurability' refer to distinct moments of the scientific process.66 (Something similar must be true of the normal process of education.) Theory without experiment is empty. Experiment without theory is blind. But in the historical development of the sciences experiment and theory are often out of step. Michelson and Morley did not see their experiment as a refutation of the other, and Michelson never in fact relinquished his belief in it.67 On the other hand Prout's hypothesis could not be vindicated until the invention of physical techniques of chemical separation, a 64 See P. K. Feyerabend, `Problems of Empiricism', op. Cit., pp. 168-71, and T. S. Kuhn, op. cit., pp. 100-1. 65 I have already suggested in paragraph 3 above the role that the concept of powers might play here. 66 It is his failure to see this that I think leads Feyerabend into error. For he wants to say (a) that there is bound to come a time when the `alternatives' do not share a single statement (including observation statement) in common, yet (b) we could still `choose' between the theories, viz. in terms of the uninterpreted sentences that the scientists testing them would be motivated to produce in observational contexts, op. cit., pp. 214-5. Now what is objectionable about this suggestion is not only that such uninterpreted sentences could never provide *grounds* for a choice (cf. e.g. D. Shapere. `Meaning and Scientific Change', Mind and Cosmos, ed. R. G. Colodny, p. 61) but also the idea that we could ever be in a position to make such a choice. For this involves the hypostatisation of a whole historical process of meaning-change and its encapsulation in a single notional moment of judgement. In this way it involves a relapse back to the pre-relativistic notion that we can make judgements outside some particular theory and some particular position in theoretical time. 67 Cf. I. Lakatos, op. cit., pp. 159-65. 192 A Realist Theory of Science century after its formulation. It is not just the experimental results but what is done with them that counts. On the other hand, the nonfulfilment of experimental expectations can always be explained in terms of the deficiency of experimental techniques. Scientific activity is itself differentiated into periods or better `phases' (so as not to identify them with chronological time): viz. into (a) phases of discovery and/or change and (b) phases of r application. Both are necessary. I use these characterizations in preference to the emotive and somewhat misleading terms `revolutionary' and `normal'. (a) consists in the production of the knowledge of a new stratum or level and/or the radical revision of knowledge at the current one. This is often preceded by the hint or glimpse of a new level or by a crisis induced by the proliferation of anomalous facts of a particularly disturbing kind. The discovery of X-rays illustrates both these facets. Indeed Lord Kelvin initially thought that Roentgen had devised an elaborate hoax.68 (b) consists in the application of the discovery and/or change to account for (and perhaps correct) currently established facts and generate new ones. Needless confusion has been engendered by the failure to distinguish models, theories, paradigms, etc. Very roughly, a theory is a model with existential commitment; that is, a model conceived, and meant to be taken, as true; i.e. a model in which the entities posited and mechanisms described are conceived as real. It is relatively easy for the scientist to invent models, but much more difficult for him to construct theories. There were several models of the aether, but never a satisfactory theory of it. Diagram 3.7 illustrates a schema for the development of science. A general conceptual scheme (abbreviated here to G.C.S.) or metaphysical framework, such as that provided by atomism, `begets' (logically, not temporally) a research programme, such as that associated with the attempt to explain phenomena by reference to the primary qualities of matter. The research programme in turn generates a theory and/or a sequence of theories either intended for different fields (or different strata) and/or in competition with and replacing each other. At the centre of theory-construction is the process of model-building and technical innovation required for the empirical testing of 68 Cf. T. S. Kuhn, op. cit., p. 59. The Logic of Scientific Discovery 193 the various models. These levels are never in complete harmony. Some hypotheses, seemingly necessitated by the facts, are always out of line with the theory generating the facts. Gravity, for example, could never be assimilated to the corpuscularian metaphysical paradigm. Puzzles or problems are the concrete working data of the scientist. His immediate task is their clarification and resolution. For this he must use the tools he has at his disposal: established results, facts, and theories, promising hypotheses and half-tried Diagram 3.7. The Internal Structure of Science C.G.S.________________________________________________________ | ^ | | v | Research Programme_________________________________ | ______ | | v | Sequence of | Theories ................._________________ | _______ ^ ^ | | | | (a) `non-normal'_| | | science | | (b) `normal'_______| | science | Critical Science Nb. so called `non-normal' science corresponds to phase (a) viz. of discovery and change: in it models are invented and subjected to empirical tests with the aim of theory construction. so called normal' science corresponds to phase (b) viz. of application: in it puzzles generated by phase (a) are resolved. the structure of theory-construction: models + facts + theories + techniques [within G.C.S. + Res. Prog.] --> new theories. the structure of puzzle-solving: facts + theories + techniques + methodological (and heuristic) paradigms --> resolution of puzzles. critical science has no formal structure but may call into question any level, including established theories, successful research programmes and even the G.C.S. or metaphysical framework, leading to the replacement or development of theory, research programme or G.C.S. (e.g. Mach, Einstein). 194 A Realist Theory of Science (or forgotten) ideas, the available formal and technical equipment and usually some methodological (or heuristic) paradigm, concretely embodied perhaps in some exemplary piece of work. It is because such paradigms are shared that there are intersubjective criteria for assessing the scientific adequacy of his work. The scientist's work normally takes place within a definite institution, a disciplinary matrix and is governed by what some continental writers have called the problematic of his science (that is, roughly the structured field within which alone meaningful questions can be asked or problems posed, expressing the dominant theoretical concerns of the time). The research programme may have its own methodological paradigm, such as Newton's *Principia* or Durkheim's *Suicide*. Now if the problems generated by work at phase (a) cannot be resolved by the material available to the scientist in his own field he must necessarily draw on another. This provides the rationale for paramorphic model-building, a role for analogies and metaphors (the models of discursive thought), leading to existential novelty (e.g. `what are the punctuation marks in a genetic code?'). Incidentally, this also helps to explain why theoretical innovations are often made by individuals originally working in fields adjacent to the field they innovate, as in the case of Dalton (a meteorologist); or with a strong `prejudice' taken from an adjacent field, as in the case of Pasteur who was convinced, in opposition to orthodox opinion at the time, that fermentation could not be caused by chemical agents alone but had to be explained by dissymmetrical forces associated with the activity of living organisms.69 Conceiving science as work readily lends itself to Aristotelian schematization.70 The material cause is antecedently established knowledge, facts and theories; the efficient cause is the methodological paradigm or generative theory at work in the theoretical and experimental activity of men; the formal cause new knowledge, facts and theories; and the final cause knowledge of the enduring and transfactually active mechanisms of nature. Now from the perspective advanced here an event, such as the discovery of oxygen, is significant not just because it refuted the 69 See e.g. M. J. Mulkay, The Social Process of Innovation, p. 12. 70 This has been noted by J. R. Ravetz, op. cit., pp. 116-18. But his use of Aristotle's schema differs substantially from mine. The Logic of Scientific Discovery 195 phlogiston theory of combustion, but because it constituted a decisive moment in the transformation from one way of doing chemistry, viz. that associated with the theory of elective affinity, to another, viz. that represented by post-Daltonian atomic chemistry; that is, because it constituted a transformation in the ongoing activity of chemistry. Knowledge does not exist in a third world.71 Rather, it exists in our world, embedded in the scientific community. Without men there would be no knowledge, only its traces. In this sense it depends upon men. But though it exists only in virtue of the thoughts and actions and products of men, it is irreducible to them. For though it would not exist without the activity of some men, its pre-existence is a necessary condition of the activity of anyone. It is a public mix that always antedates the individual. Now it is not necessary that science should continue, i.e. be ongoing. It is contingent that it is. But given this men must reproduce (or more or less transform) the knowledge that is given to them. Men do not construct their knowledge: they reproduce or transform it. This is another way of saying that any knowledge that there is must possess a material cause. Now in general for scientific activity to be continuous all levels depicted in Diagram 3.7 must be represented. The trouble with social science, for example, is not that it has no (or too many) paradigms or research programmes; but rather that it lacks an adequate general conceptual scheme. I said in 1.6 that an adequate philosophy of science would depend not only upon the development of an adequate philosophical ontology, but upon the development of an adequate philosophical sociology too. This must consist in an answer to the question: what must society be like if science (as a specific kind of social activity) is to be possible? It must satisfy the desiderata of being a structure irreducible to but present only in its effects. Society can only be known, not shown, to exist. It exists only in virtue of the intentional activity of men but it is not the result (or the cause) of their intentional activity. Sociology and psychology thus constitute distinct branches of science, in the sense of the criterion developed in paragraph 3 above. Sociology is not concerned with masses of individuals or mass behaviour; but with the persistent relationships between 71 See K. R. Popper. Objective Knowledge, esp. Chaps. 3 and 4. 196 A Realist Theory of Science individuals. Such relationships would not exist without their relata but they do not depend for their efficacy upon any particular relata, any particular named individuals. Now the autonomy of sociology and psychology accords well with our intuitions. Thus we do not suppose that the reason why the war is fought is the soldier's reason for fighting it, just as the reason why the bar of chocolate is wrapped need not be the chocolate wrapper's reason for wrapping it (though it depends upon the latter). I do not have to know the laws of supply and demand to buy a mackintosh or to know the deep structure of language in order to use it. The deep structure of language may indeed impose limits (like natural structures) upon the kinds of speech acts I can perform but it does not determine what I say. This conception of social science thus preserves the status of human agency, but does away with the myth of creation (logical or historical), i.e. the possibility of a methodologically individualist reduction. It is not necessary that that society should continue. But if it is to do so then men must reproduce (or more or less transform) the structures (languages, forms of economic and political organization, systems of belief, cultural and ethical norms, etc.) that are given to them. The Newtonian revolution in sociology consists in coming to see that it is not necessary to explain society as such; but only the various structures responsible for different societies and their changes. The problem of how men reproduce any particular society belongs to a linking science of social psychology. As so conceived, society may be regarded as an ensemble of powers which exist, unlike other powers, only as long as they are exercised; and are continually exercised via (i.e. in the last instance through) the intentional action of men.72 Established facts are social products. Understanding their logic may help us to clarify the relationships between men and society and men and the world. Here the metaphor of a reading may be used. Its adequacy depends upon the existence of both a given language and an independent text. (But the metaphor is misleading in one way: in that the text of nature exists independently of any language.) It is this that makes possible talk of a correct, rather than just commutatively successful, 72 For the concept of powers continually exercised we have of course groomed the concept of tendency. The Logic of Scientific Discovery 197 reading. Men never create this language. For it always preexists them. But it exists as an actual, i.e. `living', language only in virtue of, and changes with, their uses of it. Thus if society is represented by the model of a language it may be regarded as a structure which is always there; which men must reproduce or partially transform; but which would not exist without its `functionaries'. It is methodologically incorrect to search for an efficient cause of society, though society depends necessarily upon the efficient activity of men. But a reading depends upon antecedent social activity; the acquisition of a language by the reader. It is in this sense that the facts always depend upon social activity. In experience the skilled scientist reads the world as if it were a text in an attempt to understand the mechanisms of the production of phenomena in nature. But his own reading depends upon the mechanisms of the reproduction and transformation of language, of knowledge and of society. To return now to science: it is contingent that science is continuous but given that it is men must reproduce (or more or less transform) the knowledge that is given to them. The condition that science be continuous is equivalent to the condition that all knowledge possesses a material cause. Hence the criterion of adequacy in the transitive dimension of the philosophy of science is that the account of science should be capable of sustaining the concept of it as an ongoing social activity. It is here that dyadic theories of falsification, that is, theories that conceive falsification as consisting in a confrontation between a single theory and a set of facts fail. If science is to be continuous, refutations must be replacements; which means that always more than one theory must be involved. But, related to this, is an even more basic objection to fallibilism as such. For the refutation of any theory presupposes the acceptance of the refuting observation statement. If everything is conditional nothing can be. If all knowledge is (equally) conjectural, no statement can be refuted. Of course it is always possible that the scientist is mistaken in any particular belief (and a good scientist is continually alive to this possibility). But in order to demonstrate a mistake some proposition must be asserted (some theory accepted and framework worked within). In order to learn from our mistakes we must know that (and when) we are mistaken. Lacking from fallibilism, as from classical empiricism, 198 A Realist Theory of Science is the key concept of knowledge necessarily possessing a material cause: antecedently established knowledges; science's means of production. It is not necessary that a scientist works within an particular framework or accept any particular theory; but it is necessary that he works within (accepts) one. Underpinning empirical realism is a model of man in which men are seen as sensors of given facts and recorders of their constant conjunctions: passive spectators of a given world rather than active agents in a complex one. This model plays a role at least as important as that played by the classical paradigm of action and the celestial closure discussed in Chapter 2. Together they form a complementary triangle (see Diagram 3.8) Classical Paradigm of Action /\ / \ / \ / \ / \ / \ Model /___________ \ Celestial of Man Closure Diagram 3.8. Three Sources of Empirical Realism Implicit in empirical realism is a conflation between a ground of knowledge, viz. experience, and the world. If experience is to be capable of playing the role traditionally assigned to it of grounding our knowledge (in whole or in part) then the items of which it is composed must be perfectly simple and atomistic; i.e. insusceptible to further analysis or justification. But if it is to define the world then the world must be similarly composed: of atomistic and discrete events (or momentary states) independent of each other. If knowledge is to have its foundations in experience and experience is to define the world then both the ultimate items and objects of knowledge must be atomistic and independent of each other. This creates the problem of what grounds we can have for moving from the observed to the unobserved, or from the actual to the possible (and thence to The Logic of Scientific Discovery 199 the counter-factual). Fallibilism, which shares this model, can no less escape this problem. For the refutation of theory T at time t_1 by an observation statement is consistent with its corroboration by that statement after t_1 unless we are justified in moving from the observed to the unobserved, and from the actual to the possible. (Unless induction is justified or nature is uniform we can never know that a `mistake' is a mistake; so we can never put our mistakes behind us.) --- from list --- .