Hacking's and Agazzi's Realism

According to Hacking experimentation “has a life of its own”.^{[19] [20] This statement, which has become the motto of the new experimentalism, means that experi­ments, unlike theories, “are organic, develop, change, and yet retain a certain long-term development which makes us talk about repeating and replicating experiments”.11 This entails a sharp distinction between high-level theories and the phenomenological laws and models, endowed with a low level of generality, which provide the basis for the stability of experimental results.}

In this connection, Hacking bases the reality of theoretical entities such as elec­trons on the fact that we can manipulate them in the same way as we manipulate everyday objects. We can put electrons to various uses—for example, we can use them to alter the charge on a niobium ball.^[22] Theoretical entities are real chiefly because they can be used as “instruments” capable of interacting causally so as “to manipulate other parts of nature in a systematic way”. By the time we succeed in doing this, the electron “has ceased to be something hypothetical, something inferred. It has ceased to be theoretical and has become experimental”.^[23]

By embracing these theses, Hacking tries to prevent experimentation from being swallowed up by the theoretical aspect. However, by so doing he ends up endorsing, in an inverted form, a dichotomy of theory and experiment similar to the one he criticises in Popper and in relativistic philosophies of science.

Despite insisting on a technical-operational criterion for the reality of theoreti­cal entities, Hacking stresses the contrast between experiment and theory rather than their connection. This contrast underpins his claim that one can defend real­ism “about entities” but not “about theories”. In this respect Hacking runs into two main difficulties. In the first place, it is unclear what more might a theory say about empirical reality other than what it says about it through its theoretical enti­ties. It is impossible to distinguish the evidence for the existence of certain parti­cles from the evidence for the theory that talks about them, since the content of the theory is just the claim that these particles and their properties exist. As Franklin noted, when experimenters find a particle with a charge, mass and lifetime equal to those ascribed by the theory to K mesons, they rightly assert the existence of K-mesons—just as we would assert the existence of the philosopher Bas van Fraassen if we found an entity with height, weight, gender, hair colour, eye colour, date of birth and home address which were exactly those listed on his driver’s license. The point is that the successful performance of operations and measure­ments on certain particles warrants both the existence of those particles and the truth of the laws about them. The use of the laws that allow us to find out the prop­erties of elementary particles gives these laws the same epistemic status as the par­ticles and their properties.^[24]

Secondly, to maintain that experiments have a life of their own, thus separat­ing theory and experiment, would make action punctual, limited to the entities involved in a particular action here and now. Action would lose its generality, a generality which depends on the encounter between the linguistic-representational sphere and empirical reality, and is expressed by natural laws in the form of theo­ries. Without theory, action could only afford access to a reality which would be indeterminate and formless, an unknown “x” that would be scientifically irrelevant and should be discarded in accordance with the principle of economy since, being compatible with any theory, it is incapable of “taking sides” in disputes between rival theories.

Although Hacking is quite right in rejecting theory-ladenness in its most common interpretation, he neglects its element of truth, which consists in the theoretical-per- spectival character of scientific knowledge.^[25] By changing the theoretical frame­work, two instances of the “same” experiment can become two different experiments.

Therefore, there is an intrinsic connection between theory and experiment. The Neopositivist and Popperian thesis of the independence of theory from experi­ment is as flawed as its mainly new experimentalist converse, the independence of experiment from theory. Both theses presuppose a dichotomy between theory and experiment that makes both concepts incomprehensible.

Returning to our comparison between Hacking’s and Agazzi’s realism, we can say that (despite some inconsistencies which we shall discuss later on) in Agazzi the problems arising from the dichotomy between theory and experiment are, at least in principle, solved. What may be called the “perspectival character” of sci­entific knowledge^[27] is one of the pillars of Agazzi’s philosophy of science, namely the notion of objectivity as reference to “objects”. As we saw, it is in connection with this sense of objectivity that Agazzi insists upon the fact that the different sciences investigate “things” from different points of view, so that the same “thing” can become the object of different sciences when considered from differ­ent “viewpoints”.

Later on we shall touch on some aspects of Agazzi’s thought in which the opera­tional and theoretical aspects are not perfectly harmonized. However, before we do that I wish to say something about the central role played by the perspectival charac­ter of scientific knowledge in Agazzi’s thought. For this purpose it will be interesting to compare Agazzi’s and Searle’s views on the limits of AI. At first sight, this might seem a strange comparison, but a re-examination of Searle’s point of view in the light of the perspectival character of scientific knowledge will prove to be instructive.

3