Journal of Philosophy, Inc.

On Russell's Phenomenological Constructionism

Author(s): Carl G. Hempel
Source: The Journal of Philosophy, Vol. 63, No. 21, American Philosophical Association
Eastern Division Sixty-Third Annual Meeting (Nov. 10, 1966), pp. 668-670
Published by: Journal of Philosophy, Inc.
Stable URL: http://www.jstor.org/stable/2024158
Accessed: 19/07/2010 22:49

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=jphil.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Journal of Philosophy, Inc. is collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Philosophy.

http://www.jstor.org
668 THE JOURNAL OF PHILOSOPHY

ON RUSSELL'S PHENOMENOLOGICAL
CONSTRUCTIONISM *
IN his examination of Russell 's ontological development, Pro-
fessor Quine comments briefly on Russell's project of construct-
ing the external world out of sense data by means of the logical
methods that had been so brilliantly employed in Principia
Mathematica to construct mathematics from a small array of logi-
cal and set-theoretical concepts and principles.
Quine sees no hope for a realization of Russell's bold reduction-
ist vision, and he suggests that the worst obstacle can be located by
reference to Carnap's Der logische Aufbau der Welt (1928),
namely, at the stage where colors and other qualities are ascribed
to physical space-time points; for the ascription is effected in a
manner that allows for revision in the light of later experience and,
hence, cannot be reduced to definition.
This last inference, however, does not seem compelling. Sup-
pose that colors were assigned to space points in the manner of
this definition:
Point x is blue if and only if all phenomenal
aspects of x are blue.
Ascription of blueness to a point would then always remain open
to revision, and yet it would be definitionally reduced to phe-
nomenalistic terms. Carnap's procedure is more complex than the
one just considered; but the differences between the two do not,
as far as I can see, affect the point at issue.
Yet, serious difficulties do arise for the constructionist program
precisely at the stage indicated by Quine. For colors and other
qualitative and quantitative characteristics of locations or of physi-
cal bodies will have to be construed, broadly speaking, as dis-
positional tendencies of certain kinds of sense data to occur under
suitable conditions of observation; and it seems quite clear that
the concept of disposition, which is closely linked to the concepts
of law and of nomic connection, cannot be expressed within the
extensional framework of logic and set theory.
Russell 's own constructional sketches rely explicitly on laws
and nomic connections to select "from a world of helter-skelter
sense-data" 1 those subclasses whose elements are to count as as-
* Abstract of an APA symposium paper, commenting on W. V. Quine,

"Russell's Ontological Development," this JOURNAL, 63, 21 (Nov. 11, 1966):

657-667.
1 Our Knowledge of the External World (New York: Norton, 1929),
p. 114.
SYMPOSIUM: PHILOSOPHY OF BERTRAND RUSSELL 669

pects of, and to provide a constructional definition for, individual

physical things. Russell proposes to base the selection on require-
ments of continuity and of conformity with physical laws, and he
defines physical things as "those series of aspects which obey the
laws of physics" (117). He uses essentially the same idea again
in supplementing the class of actually observed aspects of a physical
thing by what he calls "ideal appearances" of it, i.e., aspects
that would be, or would have been, observed but for the lack of a
suitably located observer.
In this reliance on physical laws, however, Russell could be
assumed to limit himself to the laws provided by some specified
physical theory (for example, contemporary physical theory),
thus eschewing reference to the general concepts of law, nomic
connection, and disposition. But still, any specific law invoked
in a Russellian construction would be formulated in terms of char-
acteristics of physical things, not in terms of aspects or sense data.
Hence, the application of such laws to aspects can at best be
only indirect and will presuppose a clarification of the relationship
between the physical characteristics of things on one hand and
aspects or appearances on the other. And here, the difficulty of
accounting for dispositions and nomic connections arises again.
One consideration that made the constructionist program seem
so attractive-and even philosophically imperative-to Russell was
his idea that all statements about the physical world are verified
by the occurrence of appropriate sense data, and that therefore, in
so far as physics is verifiable at all, it should be construable in
terms of actual sense data alone, rather than in terms of inferred
or postulated entities.2 But a theory is stronger than the class of
all actual and potential observational data that conform to it;
and, in a somewhat similar sense, the concepts of a theory may be
said to be stronger than those serving to describe the relevant
observational evidence. For, first, the "meaning" of a theoretical
term depends not only on the observational criteria for its applica-
tion, but also on the theoretical principles in which it functions.
Second, while there seems to be no clear instance of a definition
for a theoretical term by means of observational or phenonienalistic
terms, there are at least some plausible examples of definitions
that go in the opposite direction, as witness the characterization of
colors (as observable physical characteristics) in terms of the
emission or reflection of light whose frequency falls within certain
2 See, for example, op. cit., p. 86, and "The Relation of Sense-data to
Physics, " in Russell 's Mysticism and Logic (Baltimore: Penguin Books,
1953), especially pp. 148-151.
670 THE JOURNAL OF PHILOSOPHY

intervals, or the definition of such terms as 'water' or 'copper'

by means of theoretical expressions referring to molecular or atomic
structures. But these examples, and others that could be added,
all stay within the realm of physics; for the observational terms
they define stand for public physical characteristics. A phe-
nomenalistic construction of the physical world, however, would
require general statements of biconditional form colinecting a
physical term with an expression containing phenomenalistic terms
only; and currently available theories seem to offer no plausible
examples of any such bridge statements. There does indeed appear
to be no hope for the realizability of Russell's program.
CARL G. HEMPEL
PRINCETON UNIVERSITY

RUSSELL AND PHILOSOPHY *

USSELL has said so many different things and is so close

to most of the major trends in twentieth-century Anglo-
American philosophy, that I have chosen to speculate on the
whole scene and discuss merely one technical point (on classes and
ontic commitments). Elsewherel1 I have written about Russell's
logic and shall not repeat myself.
From about 1898 on, Russell joined G. E. Moore in a revolt
against Bradley: refuting idealism and defending common sense.
For both, time is real, physical objects are real, Platonic ideas
are real. A preoccupation with foundations of mathematics led
Russell to his theory of descriptions and to a claim that not only
descriptions but also classes and relations can be handled by
incomplete symbols eliminable in contexts. The belief that classes
can be had at no extra expense undoubtedly lent plausibility to
the principle of extensive abstraction which Russell credited to,
and shared with, Whitehead. Following this principle, a physical
object is, for example, said to be a class of classes of events. Con-
currently with these more specific proposals, Russell and Wittgen-
stein were said to have founded logical atomism, although there
are significant differences between their views.
The logical positivists were strongly under the influence of
Mach, Russell, and the Tractatus. A deceptively simple picture
of human knowledge emerged. Only very gradually did it become
* Abstract of an APA symposium paper ' commenting on W. V. Quine,

''Russell's Ontological Development," this JOURNAL, 63, 21 (Nov. 10, 1966):

657-667.
1 ''Russell and His Logic," Ratio, 7, 1 (June 1965): 1-34.