Philosophy and science: Putnam and Friedman
1.
Putnam
and
the
rejection
of
the
“thing-in-itself”
Recognizing the Kantian influence on his work (“my indebtedness to Kant is
very large” – Putnam, Realism with a Human Face, 1990, p. 3) Putnam is antifoundationalist, i.e. he rejects “metaphysical necessity” (see “Is Water necessarily
H2O?” in Putnam 1990) and he puts forward internal realism. Along the same lines as
Goodman, but avoiding the relativism, Putnam rejects the absolutist picture of the
world to be found in realism. Different languages, different theories, and different
representations have the same value in different contexts. “In my picture, objects are
theory-dependent in the sense that theories with incompatible ontologies can both be
right.” (Putnam 1990, p. 30) Both language and science are deeply pluralistic in their
ontology. For Putnam absolute realism or foundationalism has no sense. “I criticize
the idea that the world picture of fundamental physics is metaphysically complete.”
(see “Is the causal structure of the physical itself something physical?” in Putnam
1990, p. 95) Like Quine, Putnam accepts that stars, tables and micro-particles “exist”
only because we can describe all these things in different conceptual schemes. Thus
the concept of existence has to be related to conceptual frames. For Putnam, the
question “How many objects really exist?” is an illusion. (Putnam 1987, p. 20) In fact
the conceptual schemes are very similar to Carnap’s linguistic frameworks.
Putnam internal realism (see “I should have called it pragmatic realism!”
Putnam 1987, p. 17) rejects the notion of the thing-in-itself”. (Putnam 1987, p. 36) In
this sense, realism is not incompatible with conceptual relativism. (Putnam 1987, p.
1
17) Again, from the EDWs perspective, there is the same problematic relationship
between the world and the so-called “relative concepts” or “relative descriptions”.
Conceptual schemes are empty in describing different entities that belong to the
unicorn-world. As we shall see in the next section they are missing a kind of Kantian
constitutive framework.
Putnam asserts that assuming that there is a “thing-in-itself” presupposes a view
from an “Archimedean point”. In this case it is about an ideal or impersonal
knowledge.
The same notion of a “God’s Eye View”, the same epistemic ideal of achieving a view from an
“Archimedean point”—a point from which we can survey observers as if they were not
ourselves, survey them as if we were, so to speak, outside our own skins—is involved in both
cases. The same notion that ideal knowledge is impersonal is involved.
(Putnam 1990, p.
17)
2. Friedman’s relative constitutive a priori principles
Friedman re-introduced and developed Reichenbach’s notion of relativized a priori
principles in explaining Newton’s and Einstein’s theories from mathematical physics.
He replaces Quine’s holism of belief with a dynamical and stratified system of
knowledge constituted on three levels: (1) empirical laws of nature (like Newtonian
laws of universal gravitation or Einstein’s equation for the gravitational field; (2)
constitutive a priori principles, such as principles from geometry and mechanics that
construct paradigms (in Kuhn’s sense) or conceptual frameworks and “define the
fundamental spatio-temporal framework within which alone the rigorous formulation
and empirical testing of the first or base level principles is then possible”; and (3)
2
philosophical meta-paradigms or meta-frameworks that guide the transition from one
paradigm or conceptual framework to another. (Friedman 2001, pp. 45–6)
Extrapolating Kant’s idea, Friedman considers that a priori constitutive principles
define “the fundamental spatio-temporal framework of empirical natural science.”
(Friedman 2001, p. 43) For each scientific theory there are certain a priori
constitutive principles that define its proper space of empirical possibilities.
(Friedman 2001, p. 84) Extrapolating Kant’s idea, Friedman considers that a priori
constitutive principles define “the fundamental spatiotemporal framework of
empirical natural science”. (2001, p. 43) For each scientific theory there are certain a
priori constitutive principles that define its proper space of empirical possibilities.
(Friedman 2001, p. 84) However, the question is how we can define “the space of
empirical possibilities” for each individual theory.
In addressing this question, let us see how Friedman explains Newton’s and
Einstein’s theories within his framework. Newtonian mechanics and Einstein’s
theory of relativity contain two parts: (1) the empirical part containing such laws as
universal gravitation or Einstein’s equations for the gravitational field; (2) the
constitutively a priori part containing both mathematical principles utilized for
constructing the theory (Euclidian geometry vs. the geometry of Minkovski’s spacetime, the Riemannian theory of manifolds) and fundamental physical principles or
the “mechanical part” (the Newtonian laws of motion vs. the light principle, the
equivalence principle). (Friedman 2001, p. 71)1 Newton’s and Einstein’s theories
offer us different spatio-temporal structures.
There are different reasons for which we cannot even compare these two
theories. Einstein has transformed the light principle that was an empirical principle
We have to recall that, in Parvu’s interpretation, for Kant, the “condition of possibility” appears at
two levels: mathematical (axioms of intuition and anticipations of perception) and physical (analogies
of experience, i.e. the transcendental laws of nature).
1
3
for Newton into a constitutively a priori one. “Einstein has ‘elevated’ an empirical
law to the status of a convention or to the status of a coordinating or constitutive
principle”. (Friedman 2001, p. 88) Within the special theory of relativity the spatiotemporal framework is Minkovski’s four-dimensional space-time.2 If in Newton’s
theory the gravitational force is independent of inertial mass, in the general theory
of relativity they are equivalent. “The inertial structure is defined in terms of ‘freefalling’ trajectories in a gravitational field” and so the “gravitational force is directly
incorporated into the geometry of space-time and thus into the constitutive
framework of our theory.” (Friedman 2001, pp. 89–91) In Newton’s period, people
could not even conceive certain notions from Einstein’s theory. The general theory
of relativity requires a different spatio-temporal framework than the one postulated
by Newton’s theory. In this sense, Friedman presents “three revolutionary
advances”: a new field of mathematics, tensor calculus or the general theory of
manifolds (originally elaborated by Riemann), Einstein’s principle of equivalence,
and his equations for the gravitational field. (Friedman 2001, pp. 37-8) The
principle of equivalence and Einstein’s field equations require a curved space-time
structure. Thus geometry “functions” as part of the constitutive framework for our
experience. (Friedman 2001, p. 62) The constitutive a priori part establishes
knowledge about experience and in this sense “they are a priori to or independent of
experience.” (Friedman 2001, p. 73) The a priori physical principles (mechanical
part) mediate between abstract mathematical tools and empirical phenomena. The
function of this mechanical part is to create, in one mathematical-physical theory,
“… for Einstein uses his light principle empirically to define a fundamentally new notion of simultaneity
and, as a consequence, fundamentally new metrical structures for both space and time (more precisely, for
space-time).” (Friedman 2001, p. 88) Friedman shows that the principle of equivalence has the same
status within the general theory. The empirical content of the general theory is given by the application of
the principle of equivalence to the inertial structure of Minkovski’s space-time. (Friedman 2001, p. 115)
Friedman mentions that this idea is from Norton 1985/1989.
2
4
the necessary link between two parts with different structures: abstract and
empirical. But a legitimate question here would be what exactly this concept of the
“empirical part” designates? For Einstein, the coordinating principles constitute a
new
framework
for
space,
time,
and
motion
(Friedman
2001,
p. 107) and therefore all the empirical laws have constitutive meaning only in the
framework created by a priori constitutive principles. Even the individuation of
entities requires such conceptual frameworks. 3 That is necessary not only because
the entities that are in motion belong to a certain spatio-temporal framework, but
also because “the knowledge of physical rigidity presupposes the knowledge of
forces acting on the material constitutions of bodies.” (Friedman 2001, p. 110) For
describing these forces it is a necessary geometry. Essential for Friedman’s
perspective is his footnote on page 55 about Einstein, who adopted a perspective on
the relationship between this necessary geometry and the entities as “practicall y
rigid bodies” that ignores microphysical forces. (Friedman 2001, p. 114) The frame
of reference for both theories is given by space, time, and motion but the theories
refer to the same entities even if we can individuate empirical objects and their
relationship only through such constitutional frameworks. In fact, the spatiotemporal structures of these theories are different mainly regarding their metric.
Empirical tests – the perihelion of Mercury – supported Einstein’s theory.
“The transcendental laws of nature” are, for Kant, the analogies of experience that determine the
formal Nature. (See 2.3)
3
5