El desarrollo de la autonomía en la adolescencia

We have seen that there are general philosophical reasons for regarding realism as an attractive position in ontology, epistemology, and philosophy of science. But the applicability of the ideas presented so far should be tested also in relation to theories in various scientific disciplines. Today there is much specialized and highly advanced professional work in the history and philosophy of science about such foundational questions. In this work, I can only give brief and non-technical illustrations of how the issues of realism have arisen in some selected cases: astronomy, quantum mechanics, psychology, and economics.

(a) Astronomy. Pierre Duhem published in 1908 a book on the idea of physical theory from Plato to Galileo (Duhem 1969). Its title, Σ Ω ZEIN TA Φ AINOMENA, was taken from Simplicius' famous description of the task of astronomy in the Platonic tradition:

What circular motions, uniform and perfectly regular, are to be admitted as hypotheses so that it might be possible to save the appearances presented by the planets?

An astronomical theory, like Eudoxus' system of homocentric spheres or Ptolemy's system of eccentrics and epicycles, has attained its goal when it has ‘saved the phenomena’. If there are several such accounts, the simplest hypothesis should be selected. But speculations about the nature of heavenly bodies, or about their true causes and essences, are beyond the tasks of astronomy.

Geminus (1st century bc end p.144

) stated clearly the distinction between a hypothesis introduced merely to save the phenomena and a hypothesis

which is true or false. In later terminology, this is the distinction between instrumentalist and realist interpretations of scientific hypotheses and theories. Geminus also suggested that the ancient Greeks generally interpreted theories in physics in the realist way (following the ideas of Aristotle) and theories in astronomy in the instrumentalist way.

But, as Duhem illustrated, this consensus was later broken: some Aristotelians (e.g.

Averroës) extended realism to astronomy as well, while some early Renaissance philosophers treated all science on the model of saving the appearances.

The contrast between realism and instrumentalism was highlighted by the Copernican Revolution.²⁰

The later partisans of the Copernican theory (Giordano Bruno, Johannes Kepler, Galileo Galilei) adopted the realist interpretation. As Cardinal Bellarmine advised Galileo in 1615, the Catholic Church approved the use of the Copernican system for the calculation of the calendar, but one should not claim ‘absolutely’ that the earth really moves and the sun is the fixed centre of the universe. In 1616 the Holy Office of the Church found the Copernican theory to be incompatible with sound (Aristotelian) physics and with holy scripture. In 1633, a year after Galileo's Dialogue, the doctrine was condemned: Cardinal Barberini (later Pope Urban VIII) had tried in vain to convince Galileo that the

heliocentric system could not be proved. According to his argument, it is not ‘beyond God's power and wisdom’ to arrange alternative ways which would ‘save all the phenomena displayed in the heavens’.

Copernicus himself was convinced that his heliocentric theory not only saved the appearances, but was also true—conformed to the real nature of things and to the principles of physics. Joachim Rheticus urged in 1540 that the Copernican hypotheses were ‘most true’ (verissimae), since he arrived at them by the physicist's method, i.e.

from effects to causes. But when Copernicus' De revolutionibus finally appeared in 1543, the preface by Andreas Osiander claimed that the author's novel proposal was merely a device for ‘calculation’ and ‘correct computation’—it did not tell ‘how things really were’, and was neither true, likely (verisimile), nor certain.

Galileo is usually treated as the hero of modern science. Duhem's view was quite different. He granted Galileo only the insight that all the phenomena of the inanimate universe should be saved together. But in their understanding of the scientific method

‘the Copernicans stubbornly stuck to an illogical realism’: ‘logic sides with Osiander, Bellarmine, and Urban VIII, not with Kepler and Galileo’ (1954: 113).

Duhem's To Save the Phenomena was a part of his own campaign for instrumentalism, advocated in his 1906 book on the aim and structure of physical theories (Duhem 1954).

His historical interpretations have not been unchallenged. Thus, Lehti (1986) argues against the standard view of Ptolemy's Almagest as an instrumentalist work. Musgrave (1991) attacks Duhem by claiming that astronomical instrumentalism—from Plato, Eudoxus, Apollonius, and Hipparcus to Ptolemy—is a myth. Musgrave points out that there is an important difference between conceding the conjectural or hypothetical character of some scientific theory and taking the theory to be an uninterpreted symbolic device. Agreeing with Jardine (1979), he suggests that what Duhem described as debates between realists and instrumentalists were in fact disputes between dogmatic and critical realists.

(b) Quantum theory. The status of scientific theories had become a hot issue in physics already by the late nineteenth century. Robert Kirchhoff and Ernst Mach argued that physical theories should describe, in the most ‘economical’ way, observable phenomena and their functional interconnections. On this basis, Mach criticized the concepts of absolute time and space in Newton's mechanics (Mach 1960)—paving the way for the special theory of relativity of Albert Einstein. Mach rejected the existence of atoms: for him atoms were mere fictions without physical reality. Mach's programme was supported by Heinrich Hertz's reformulation of mechanics which eliminated the theoretical concept of force (Hertz 1956). Duhem's instrumentalism (and Mach's own practice in science) allowed the use of auxiliary terms and theoretical hypotheses to save the phenomena, but

such hypotheses should not be understood as explanations with a truth value. Also Henri Poincaré's conventionalism regarded the axioms of geometry and mechanics to be

‘definitions in disguise’—they are not true or false, but more or less ‘convenient’. The defenders of realism against Machian positivism and instrumentalism included Ludwig Boltzmann, Max Planck, and Albert Einstein (after his work on Brownian motion). What seemed to be the decisive argument for the theory of atoms, and thereby for Planck's realism, came from experimental work in physics, especially Wilson's cloud chamber which almost allowed the physicist ‘to see’ the atoms. Conflicting philosophical

interpretations of this problem situation were given by the neo-Kantians of the Marburg school and by the logical empiricists in Vienna (Moritz Schlick, Rudolf Carnap) and Berlin (Hans Reichenbach).

end p.146

This was the lively historical background of the new quantum theory created in the 1920s by the generation of such great physicists as Albert Einstein, Niels Bohr, Werner

Heisenberg, Erwin Schrödinger, and Wolf-gang Pauli.²¹ It also helps us to locate some well-known and important positions on the nature of quantum theory in the conceptual net developed in earlier chapters (cf. Niiniluoto 1987c). This approach has its limitations:

the philosophical views of real-life physicists are usually ‘mixed’ cases of different influences and doctrines, and they cannot be reduced to the ‘pure’ states defined by the abstract distinctions of the philosophers. But this fact also makes the continuing

re-evaluation of the foundational discussions of physics philosophically rewarding. I believe that, in particular, the elaboration of various kinds of realist and non-realist approaches in recent philosophy of science gives us tools for understanding better the positions taken by the creators of quantum theory.²²

A good starting point for the discussion is Niels Bohr's philosophy of complementarity which continues to be an important object of various interpretations.

23

As Folse's Bohr accepts the existence of quantum-mechanical objects behind the phenomena, he is at least an entity realist. Hence, he cannot be a phenomenalist. While there is a strong Kantian flavour in the

According to Henry J. Folse (1985; 1987), Bohr is a physicist who wishes to avoid ontological discussion—and thus favours a position similar to Fine's NOA (cf. Section 1.4).

However, Folse argues, Bohr is committed to a form of realism, i.e. to ‘upholding the view that it is the task of science to understand the nature of physical reality as it exists independently of the phenomena through which we experience it’. More precisely, Bohr must accept that the same ‘atomic object’ causes, via physical interactions, different complementary ‘phenomenal objects’. The classical description—with the wave and particle pictures—refers to properties belonging only to a phenomenal object. Indeed, no pictorial description of the atomic object is available. While the complementary

phenomena are not representations of the physical object that produces them, still the combination of these complementary descriptions exhausts all that can be known about this independent reality.

end p.147

distinction between atomic and phenomenal objects, where the former are causes of the latter, Bohr is not an epistemological anti-realist: quantum theory in the framework of complementarity gives us knowledge about the atomic objects in terms of their potential for causal interaction.

How could Bohr defend the idea that his atomic objects are not merely unknowable Kantian Dinge an sich, that it is possible to have some knowledge of them, in spite of the fact that he did not interpret the ψ-function as describing the state of quantum-mechanical objects? According to Teller (1981), Bohr treated the state function in an instrumentalist way as ‘a purely symbolic device for calculating the statistics of classically or commonly described experimental outcomes in collections of phenomena grouped by shared

specifications of experimental conditions’. If this is the case, laws involving the state function do not have truth values, which also excludes the possibility that Bohr is a constructive empiricist. Folse (1987) suggests that Bohr not only rejected classical realism (or the substance/property ontology) but the spectator account of knowledge and the correspondence theory of truth as well. Thus, Bohr would turn out to be a sort of pragmatist, epistemic realist, or semantical anti-realist.²⁴

While Bohr rejected classical realism, he also insisted against Einstein that quantum mechanics provides a ‘complete’ framework for describing nature. This means that quantum theory does not assign any non-relational intrinsic properties to atomic objects, but only what might be called relational or interactive properties, i.e. properties which the object possesses or fails to possess only consequent upon the occurrence of a suitable interaction (cf. Healey 1981). Moreover, for Bohr these interactions are observational interactions, i.e. they involve ‘preparation’ or ‘measurement’ by macroscopic devices.

If the claim of completeness were an epistemic thesis, then atomic objects might have intrinsic properties, but they would remain unknown or it would be meaningless to speak about them. This kind of position might be attractive to some logical empiricists with a physicalist outlook or to pragmatists with a verificationist theory of meaning. However, it would be compatible with local hidden variable theories—and thus vulnerable to the objections from Bell's Theorem (cf. Readhead 1987). I do not believe that this is the view of Folse's Bohr. Rather, his claim of completeness is an ontological thesis: there are no

‘pictures’ of atomic objects as they exist in isolation from any observation, since such objects do not

end p.148

have any properties at all. It is only after observational interaction that we can attribute to an atomic object x the relational property ‘x caused this particle-type phenomenon’ or ‘x caused this wave-type phenomenon’.

This conclusion suggests that Folse's Bohr is virtually identical with Feyerabend's Bohr, who maintains that ‘all state descriptions of quantum-mechanical systems are relations between the systems and measuring devices in action’ (Feyerabend 1962b). On this view, a micro-system isolated from measuring devices is a ‘naked individual’, a bare particular without properties. As Hooker (1972) remarks, it is doubtful whether the idea of such

‘propertyless ghosts’ is coherent (cf. Section 2.4).

If this interpretation is correct, Bohr is a radical nominalist with respect to the micro-world. He is also a representative of a special kind of internal realism: the independently

existing naked individuals will be dressed with relational properties only through observational interactions (cf. Ch. 7).

Bohr's complementarity, as interpreted by Folse, is a weak form of realism: only naked atomic objects exist independently of observations, and our knowledge of such objects concerns their phenomenal manifestations.

This view rejects the claims that atomic objects are fictions, mere potentialities (Heisenberg²⁵

However, there are objections to Folse's Bohr which suggest that it is reasonable to search for a version of quantum realism which is stronger than Bohr's but weaker than classical realism. First, it is not plausible that an atomic object does not have any permanent intrinsic properties. It is true that the state of a quantum-mechanical system need not be independent of our measurements: such a system does not have a definite or sharply defined position and momentum, but our measurement may have the effect that the system is ‘localized’. But while position and momentum are ‘non-classical’

interactive properties (cf. Piron 1985), the same does not hold of mass and electric charge. Without such properties, we could not

), exist only during observational interactions (microphenomenalism), or are created in observational interactions. It also gives up attempts to restore classical realism within deterministic (Bohm 1980) or indeterministic (Popper 1985) frameworks.

end p.149

identify atomic objects—it would hardly make sense to say with Bohr that the same object produced two different phenomenal objects.

Secondly, if it is true to say that an atomic object x caused a particular phenomenon y, then it is likewise true that x has a causal power or potency to produce phenomena of type y. The object x can then be said to possess a dispositional property—and such a disposition is real even when x is isolated from measuring devices or from observations (cf. Tuomela 1978). The idea that the isolated x is propertyless, and cannot be described by true statements, seems to be incoherent.

Hence, it is possible to give quantum mechanics such a realist interpretation that atomic objects and some of their properties exist independently of observations. If this is correct, we may also have descriptions of atomic objects—of their properties, relations, and structure—which are true in the correspondence-theoretical sense. We can thus resist the pragmatist move suggested by Folse (1987). With non-classical properties we need a non-standard logic, and perhaps an idea of unsharp reality as well,²⁶

Thirdly, we may accept the existence of interactive properties without restricting our attention on observational interactions, as Bohr without warrant does. For example, an electron may fail to have a sharply defined location in a box; it may become localized by our manipulating activity with measuring devices, but equally well by interactions with other macroscopic or microscopic entities. As Simon Kochen (1985) remarks, there is no reason to give a ‘privileged role to the measurement process among all interactions’:

quantum theory can be interpreted as ‘describing an objective world of individual interacting systems’.

but not a non-realist theory of truth.

Thus, for an interactive realist (see Healey 1981; 1989), the main content of quantum mechanics is expressible by interaction conditionals of the form

• (16)

•

The conditionals need not have anything to do with observations or measurements—but they are true in the quantum world, i.e. in the independent reality which quantum theory is primarily about. (For example, they hold of all unobserved micro-processes inside the table in front of me at this moment.) Such probabilistic laws express invariances that serve to constitute the reality of the micro-world (cf. Section 5.2).

On the other hand, quantum mechanics also entails manipulative measurement conditionals of the form

• (17)

•

(Cf. Heisenberg 1962: 46.) It is typical of descriptivism and instrumentalism to think that such statements as (17) exhaust the content of a scientific theory; for a realist, they are test statements which can be used in the empirical appraisal of the theory (see Fig. 14).

(Cf. Bunge 1973.) In brief, (16) explains (17), and (17) gives observational evidence for (16).

In this view, Bohr's philosophy of complementarity is a valuable and interesting contribution to discussions about the peculiar subject–object relations within the

experimental testing of quantum mechanics. But this philosophy does not imply that we (or our minds) are somehow involved in the unobserved object–object relations described by interactive conditionals of type (16).

Bohr made the well-known remark, in the spirit of Dewey's pragmatism, that we are not only spectators but also actors in the drama of life. It is natural that quantum-theorists are impressed by our ability to manipulate the micro-world—by influencing, transforming, and even ‘creating’ atomic objects and some of their properties. But then they start to speak as if our use of this ability, which could be understood as a proof of realism (see Hacking 1983, and the Argument from Successful Action in Section 2.5), would somehow be responsible for all the novel exciting features of the world that quantum theory has revealed—such as wave–particle dualism, transition from potentiality to actuality, and indeterminism. For a realist, this confuses the directions of explanation and evidence: quantum measurements and observations, as described by (17), are

indeterministic because the ontologically mind-independent quantum world, as described by (16), is indeterministic.²⁷

end p.151

So is critical scientific realism able to survive the challenge of quantum theory? I have here only briefly indicated why I fail to be impressed by the frequently voiced claim that the ‘paradoxes’ in the interpretation of quantum mechanics inevitably lead to anti-realist

conclusions. But I have not discussed all the important aspects of this broad and deep question. Its answer will depend on further work on such topics as the EPR paradox, Bell inequalities, Bohm's version of quantum mechanics, and quantum field theories.²⁸

I conclude with remarks on Bernard d'Espagnat's (1983) concept of ‘veiled reality’—to illustrate how critical realism is more ‘sober’ than some speculations inspired by quantum theory.

D'Espagnat starts from a critical discussion of the problems that ‘forced’ quantum mechanics to give up the idea that atoms and their constituents ‘possess a reality that is completely independent of our means of observing it’. Less critically he then asserts that

‘consciousness exists’ and ‘cannot be reduced to the notions of physics’. In the next step, he postulates the existence of ‘independent’ or ‘nonphysical reality’ which is ‘beyond the frames of space and time and cannot be described by our current concepts’. This

independent reality is ‘veiled’, since it is ‘intrinsically impossible’ to describe it ‘as it really is’. Still, ‘empirical reality’ (the set of all phenomena extended in space and time) and ‘consciousness’ are two complementary ‘reflections’ of this independent reality.

There is again a strong Kantian flavour in d'Espagnat's distinction between an unknowable independent reality and its reflection as a mind-dependent empirical reality.²⁹

end p.152 There is also a clear influence from objective

idealism, since the ‘nonphysical’ reality is referred to by such concepts as Spinoza's substance and ‘God’.

Apart from the other traditional difficulties in Spinozistic and Kantian ontologies, d'Espagnat's theory faces the problem that he seems to multiply domains of reality beyond necessity. His need for saving the possibility of objective knowledge by

postulating an unknowable non-physical reality arises from the fact that he accepts a too mind-dependent account of empirical reality. For example, he claims that ‘our eyes contribute to the creation of atoms’ (ibid. 97). The critical realist will, therefore, reply to d'Espagnat that the concept of empirical reality is unnecessary in the ontological sense: it refers only to our partial knowledge about objective reality.³⁰

With this conclusion, the idea of a veil in front of intrinsically unknowable reality becomes unnecessary. Science is a search for reality—according to realism, the aim of

With this conclusion, the idea of a veil in front of intrinsically unknowable reality becomes unnecessary. Science is a search for reality—according to realism, the aim of