Duhem, Pierre. (1905).
La Théorie Physique: Son Objet, Sa Structure.
Paris: Marcel Rivière; 2nd edition, 1914.
Translated by P.P. Wiener as: The Aim and Structure of Physical Theory. Princeton: Princeton University Press, 1954; reprinted New York: Atheneum, 1962, 1977
Pierre Duhem (1861-1916) was a French physicist and devote Catholic who was the principal exponent of a philosophy of knowledge known as conventionalism. Duhem argued that:
"For our inquiry to make sense, we first of all claim that underlying our sense perceptions there is a reality distinct from these appearances." (p.9)
"The question of a reality distinct from our sense perceptions, and another question -- namely What is the nature of this reality? -- cannot be answered by experimental method since this method deals only with appearances and can discover nothing beyond these appearances." (p. 10)
"If we believe that physical theories can explain things, this can only be true if the physical science is subordinate to metaphysics." (p. 10)
"If theoretical physics is subordinate to metaphysics then the divisions which separate different metaphysical views will extend into the domain of physics. Hence physics would be sectarian." (p. 10)
"Each metaphysical school scolds its rivals for appealing in its explanations to notions which are themselves unexplained."
"In each case, given the metaphysical principles, it is not possible to deduce various physical laws. Invariable observation of these physical laws (like the inverse-square law) occur first."
"We cannot therefore derive from a metaphysical system all the elements necessary for the construction of a physical theory. The latter always appeals to propositions which the metaphysical system has not furnished. At the root of the explanations it claims to give there always lies the unexplained." (p. 18)
"Implicit in a metaphysical system is that physical laws are derived from that system."
"Let us attempt to formulate a physical theory which is indpendent of a metaphysical basis: A physical theory is not an explanation. It is a system of mathematical propositions, deduced from a smaller number of principles, which aim to represent as simply, as completely, and as exactly as possible a set of experimental laws."
"Any physical theory which does purport to explain betrays a metaphysical basis."
"In our mathematics, the mathematical symbols have no connection of an intrinsic nature with the properties (i.e. measurement procedures) they represent; they bear to the latter only the relation of sign to signified."
"Likewise, the relations expressed in such mathematics need not reflect the real relations among those realities. Therefore it matters little whether the operations performed do or do not correspond to real or conceivable physical transformations."
"A theory is not judged true or false according to how well it gives an explanation of physical appearances in conformity with reality."
"Agreement with experiment is the only criterion of truth for a physical theory." (p. 21)
What is the use of such an ametaphysical theory?
[Duhem calls #2 (above) "Natural Classification," and places great emphasis on it. He seems to think it is most lasting. See page 25.]
"Physical theory never gives us the explanation of experimental laws; it never reveals reality; but the more complete it becomes, the more we apprehend that the logical order in which theory orders experimental laws is the reflection of an ontological order, the more we suspect that the relations it establishes among the data of observation correspond to real relations among things, and the more we feel theory tends to be a natural classification."
[Duhem is cognizant of the psychological attractiveness of feeling that the theory is some reflection of reality.] "The scientist cannot compel himself to belief that a system capable of ordering so simply and so easily a vast number of laws, so disparte at first encounter, should be a purely artificial system." (p. 27).