text
stringlengths
0
1.71k
then we could never be in a position to judge whether speakers of X
have concepts or beliefs radically different from our own.
P22. If we could never be in a position to judge whether speakers of X have
concepts or beliefs radically different from our own, then we cannot
make sense of the idea of there being partial failure of translation.
P23. It is a reasonable assumption that there will be sentences uttered by
speakers of language X that those speakers reject as truths, and we are
to interpret those rejected sentences.
168 George Wrisley
C3. We cannot make sense of the idea of partial failure of translation
(hypothetical syllogism of P18 – P22, and modus ponens , P22, P23).
The Unintelligibility of the Very Idea of a
Conceptual Scheme
C4. The idea of different conceptual schemes is not intelligible (destructive
dilemma, P1, C2, C4).
P24. If there is only one conceptual scheme, then it is false that there are
different conceptual schemes.
P25. If the idea of different conceptual schemes is not intelligible, then it is
not false that there are different conceptual schemes.
P26. It is not false that there are different conceptual schemes ( modus
ponens , C4, P25).
P27. There is not only one conceptual scheme ( modus tollens , P24, P26).
P28. If the idea of different conceptual schemes is not intelligible and there
is not only one conceptual scheme, then the very idea of a conceptual
scheme is unintelligible.
P29. The idea of different conceptual schemes is not intelligible and there
is not only one conceptual scheme (conjunction, P27, C4).
C4. The very idea of a conceptual scheme is unintelligible ( modus ponens ,
P28, P29).
44
Quine ’ s Two Dogmas of
Empiricism
Robert Sinclair
Quine , W. V. β€œ Two Dogmas of Empiricism , ” in From a Logical Point of
View , 20 – 46 . Cambridge, MA : Harvard University Press , 1981 .
Originally published in Philosophical Review 60 (1951): 20 – 43.
Hylton , Peter . Quine . New York : Routledge , 2007 .
Kemp , Gary . Quine: A Guide for the Perplexed . New York : Continuum ,
2006 .
Russell , Gillian . β€œ The Analytic/Synthetic Distinction . ” Philosophy Compass
2 ( 2007 ): 712 – 29 .
There appears to be an intuitive difference between these two claims:
(1) All bachelors are unmarried.
(2) All bachelors are less than 15 feet tall.
While both of these statements are true, the way in which they are taken
to be true highlights what many philosophers have seen as a signifi cant
difference. The fi rst is an β€œ analytic ” truth, whose truth is determined solely
through the meanings of the terms involved and independently of any
empirical fact. The second β€œ synthetic ” truth is true because of empirical
facts about the world. In his famous and widely read article, β€œ Two Dogmas
of Empiricism, ” W. V. Quine declared that the use of this distinction in
modern empiricism was an unsupported dogma, and he further argued that
Just the Arguments: 100 of the Most Important Arguments in Western Philosophy,
First Edition. Edited by Michael Bruce and Steven Barbone.
Β© 2011 Blackwell Publishing Ltd. Published 2011 by Blackwell Publishing Ltd.
170 Robert Sinclair
what he calls β€œ reductionism, ” roughly, the view that theoretical statements
can be logically reduced to statements about experience, is a second dogma
that should also be rejected. These criticisms target the views of Rudolf
Carnap, C. I. Lewis, and others who used analyticity to make sense of the
a priori elements of human knowledge and, more specifi cally, advocated its
importance in clarifying and understanding the language of science.
In β€œ Two Dogmas, ” Quine ’ s main concern is with clearly explicating the
distinction in question, and he argues that there is no such sharp division
between analytic truths and synthetic truths. His argument has been usefully
described as analogous to the kind one might fi nd offered in the physical
sciences (Kemp, 19 – 20). A scientist might reject a type of physical phenomena
because it cannot be explained in ways that do not already assume its
existence. It might be further argued that the evidence cited in support of
such phenomena can be accounted for in other ways without them. In
general, it is this type of attitude that informs the structure of Quine ’ s
overall argument, where he begins by surveying a number of attempts to
explain the concept of analyticity and fi nds them all uninformative. Here,
he appeals to what has been called the β€œ circularity argument, ” where analyticity
is defi ned in terms of sameness of meaning or synonymy (Russell,
718).
Two expressions are synonymous when sentences containing them
remain true when one is substituted for the other, what is here described as
interchangebility salva veritate . When applied to necessity statements in
English, this view seems to work, since the sentence β€˜ Necessarily, every
unmarried man is unmarried ’ and β€˜ Necessarily, every bachelor is unmarried ’
is a case where truth is preserved when we switch β€˜ unmarried man ’ for
β€˜ bachelor, ’ and these terms are also synonyms. The problem is that such
sentences are understood as true in virtue of being analytic. The attempt to
explain analyticity by an appeal to synonymy is then circular.
Quine criticizes the second dogma of reductionism by claiming that theoretical
sentences have connections to experience only as a collective body
and not when isolated from each other. This then prevents the type of
phenomenalist reduction of science to experience advocated by the logical
empiricists and further prevents us from defi ning synthetic statements as
true when confi rmed by sets of experience and analytic truths as those
confi rmed by any experience whatsoever. With each of these attempts to
clarify analytic truth found wanting, Quine claims that the reasonable thing
to conclude is that the distinction itself is an unempirical dogma. In the last
section of his paper, he outlines his alternative view of empiricism, often
described as β€œ epistemological holism, ” which is further developed in his
later work. Here, he indicates how the alleged a priori necessity of mathematics
and logic can be explained by its deep entrenchment within our
overarching system of theoretical commitments rather than by an appeal to
Quine’s Two Dogmas of Empiricism 171