Some Critical Observations

Although by and large I agree with Agazzi’s philosophy of science, I have some doubts concerning his account of the relationship between theory and experiment and of the connection between science and technique.

Concerning the relationship between theory and experiment, Agazzi defends a variant of Kant’s thesis that an experiment is a question put to nature and rightly argues that, while the purpose of theory is to supply meaning and intension, experiments provide theoretical statements with reference.^[40] On the other hand, he seems to share the new experimentalist view that experiments can lead a life of their own. Sometimes he goes so far as to say that an experiment, regarded as a question put to nature, is understandable in itself, apart from any theory: in order to be performed, “an experimental test must be entirely describable in terms of operational concepts and their meanings, which [...] do not depend on the theory.”^[41] The same point has been reiterated by Agazzi more recently:

the theory provides the significance of the question, its point, its purpose and reason, besides providing the question with a global meaning resulting from its position in the general context of the theory. But it is not as though the question would be meaningless, that is, not understandable without the theory. This cannot be the case simply because, in order to be performed, an experimental test must be entirely describable in terms of oper­ational concepts and their meanings which, as we have already remarked several times, do not depend on the theory. This is why we need the theory in order arrive at the experi­ment, in order to design the experiment (hence the experiment depends “genetically” on the theory, as we have already said) but not in order to justify it. Once the experiment has been performed, it assumes an independent existence and is in no need of help, simply because it has the same character as the data which are the indisputable basis a theory is challenged to account for, and which it cannot modify or dispense with.^[42]

This connection between theory and experiment, by itself, comes too close to the thesis of the new experimentalism that experiments can lead a life of their own.

If in this respect Agazzi concedes too much to new experimentalism, on the other hand (especially as to the relation between theoretical and observational, or operational, terms) he inclines to the opposite view, vindicating the independence of theoretical interpretation. Theoretical concepts are necessary because they sub­sume different characteristics under the unity of an object:

every operational procedure reveals one single feature to be attributed to the object so that, after performing all the operations we need, we have a set of such features. But no object of any science is represented by a pure collection of features; it is always a struc­tured collection, in the sense that all these features are mutually connected by certain mathematical and/or logical relations, which are not obtained directly from any instru­ment, but must be arrived at through the intellectual activity of the researcher.^[43]

On closer scrutiny, this argument presupposes that operations can show single deter­minations of reality but not the link that connects them within one scientific object: this link would be in principle irreducible to operations. If this were the case, such a component of the scientific object, rather than something by means of which we seek access to reality, would be itself a term and independent object of the cognitive act, capable as such of existing on its own with no function of connection to the real world.

Moreover, the claim that theoretical terms are needed to express the connection between the properties of an object clashes with another basic thesis of Agazzi’s realism, which I have already mentioned, according to which a scientific object does not have, but is its properties. If this is true, it seems to be unnecessary to assume terms that both designate the way in which predicates are connected in an object, and are irreducible as to their content to operational terms. If we reject an unknown x beyond the phenomena, we must regard an object’s properties not as atomistic, isolated determinations of reality but as internally connected with the other properties that constitute the object (where this connection too must be established operatively by means of experimental testing).^[44]

The second main point on which, in my opinion, Agazzi’s view requires some correction is the connection between science and technique. Agazzi’s account of this connection changed considerably over his philosophical career. In Agazzi (1969) we find a decided predominance of the theoretical and rational element over the technical one: the nature of scientific theory is discussed without men­tioning the importance of its connection with technical applications.^[45] This remained unchanged until Agazzi (1985b), which contains, so far as I know, the first recognition of the crucial importance of technique for science.^[46]

In my opinion, this was an inconsistency in Agazzi’s operationalism. If one recog­nizes that scientific hypotheses have cognitive value only through some connection with our operational interventions on reality, then one must admit that technical reproduc­ibility is not a criterion among many but the distinctive criterion of the truth of scientific propositions. For in experimentation the performance of actions which are in princi­ple intersubjectively repeatable in order to test a hypothesis is a technical intervention in reality. From this point of view, strictly speaking, technical applicability does not depend on theoretical truth ascertained by other means: even though truth is not identi­cal in every sense with technical applicability, reproducible technical applications are, in the experimental natural sciences, the only way of ascertaining and justifying the truth of a theory.

To be more precise, truth is plainly not the same thing as technical usability, just as a theory is not merely an instrument. However, a theory, in so far as it is true and affords us knowledge of the way things actually are, is potentially also useful. On the other hand, the means by which we ascertain the truth of a theory (the ratio cognoscendi, one might say) proceeds in the opposite direction: we can prove that a theory says something true about the world only by showing that it can be translated into operationally, technically reproducible results. The theoretical and technical aspects within experimental natural science can be separated only by methodical abstraction: in the concreteness of doing science, they are inextricably connected. The theoretical aspect is defined as the condition of the possibility of the knowledge of determinate aspects of reality in so far as it allows one to envisage as possible causal connections that must be translatable, in principle without residue, into suc­cessful technical applications. Conversely, the technical aspect has alethic relevance only in so far as it translates into actions a conceptual mediation without which the technical aspect would appear isolated from any causal context—that is, as a mere coincidence, a chance event not reproducible outside the precise and punctual situa­tion in which it occurs (this is probably the case when animals use tools).^[47]

In Il bene, il male e la scienza (Agazzi 1992[2004]), we find a detailed discussion of the connection of science with technique. In one passage Agazzi goes so far as to admit that, if we take into account the fact that the collection of operations that “cut out” a given field of objects from reality “constitutes a network of techniques (that is, a knowledge of how to do or to work) whose goal is to make pure research possible”,^[48] then “technique is “consubstantial” with science itself [la tecnica e consustanziale alla stessa scienza]”.^[49] The last quoted sentence is left out in the English edition, and this omission is no accident.

The distinction between knowing “how” and knowing “why” is acceptable only functionally, as it were, as a distinction concerning both the theoretical and tech­nical aspects. What counts as “knowing why” at a certain cognitive level appears as a given, as “knowing how” at a further level where deeper questions arise; this deeper questioning changes the previous “knowing why” into a given (a “knowing how”) in need of further explanation. For example, one could think that we only have a “knowing that” about the functioning of the more common household appli­ances; but what is prima facie a “knowing that” (say, that the dishwasher is turned on by pushing a certain button, with no deeper knowledge of its functioning) for a child may well be a “knowing why”. To the child’s question why the dishwasher has started making that noise, we may reply, for instance, that this happens just “because” we pushed a certain button which turns it on. Likewise, we can distin­guish between our knowledge “that” the dishwasher does not work and the techni­cian’s knowledge of “why” that is the case (say, “because the condenser is broken”). However, the technician’s “knowing why” is, from the point of view, say, of an elec­trical engineer, a “knowing how” which in turn calls for an explanation as to “why” the condenser is broken—and so on without end, at least in the sense that it is not possible to establish a frontier beyond which science can progress no more.

One could respond by adducing prima facie more convincing examples sup­posedly demonstrating the possibility of a mere “knowing how”. Many technical improvements proceed from chance discoveries and can further improve without probing the reasons behind this improvement. For example, if an angler all of a sudden caught many more fish than usual and noticed that the hook had been acci­dentally bent for some unknown reason, from then on that angler may always use that hook and may also bend it more or in different ways actually producing more efficient hooks.

At first it would appear that the angler has no insight into the reasons of his undeniably technical behaviour. But if we look at the example more closely, it soon becomes evident that this is not the case. The angler would have never embarked on the search for more efficient hooks had he not noticed that the hook worked better because it had been bent; and this is a knowing “why”, it does not matter at how elementary or low a degree. Without this explanatory hypothesis, the angler would not have progressed to using the bent hook systematically, let alone to improving on it technically. In the course of the historical development of scientific knowledge, a split developed between those who operate in the field of basic science and those who operate in the applied sector; but this does not call into question the fact that science can know only by acting and intervening techni­cally in reality, and that this intervention, in so far as it is not blind but has some access to its reasons, is from the very beginning to some degree scientific.

From this point of view, also the distinction between “technique” and “technol­ogy” is legitimate in only one of its meanings, namely as the distinction between technique and discourse on technique: it is clear that the concept of technique can­not be defined by technical means. However, it is in principle impossible to distin­guish between technique and technology by alleging that the former is a pure “knowing how to do” lacking knowledge of the reasons of this doing, while the latter is an “efficient” operating which “is conscious of the reasons for its efficacy and is based upon them, that is, where operation is nourished by its grounding in theoretical knowledge”.^[51] This distinction, intended to separate technology from technique, in actual fact only separates human from animal technique. In the human sphere all “knowing how to do”, even in the weak forms of habit and/or compulsion to repeat, qua knowing, involves at no matter how infinitesimal a level a noetic aspect of critical awareness. Certainly animals too interpret their environ­ment and thereby use something similar to our concepts, but (with all the caution due when talking about animal capacities) these: “concepts” probably lack the human prerogative of criticality, that essential openness that lets them be freely modified according to the changing of situations.

It is certainly legitimate to distinguish between a scientific-theoretical attitude that concerns itself with the way things are (the search for truth) and a practical- technical attitude that aims at transforming things according to certain concepts or values. It is also correct to distinguish between pure and applied science on the basis of the different practical intentionality of the scientists involved.^[52] However, all of this is quite irrelevant to our problem, namely the epistemological relation­ship between science and technique. In particular, none of this excludes the fact that one can know empirical reality only by acting and intervening in nature, and that one can act on nature only by means of meanings or concepts without which acts would be no more than chance events.