VOL. I8 July, 1951 NO.
PROFESSOR MARGENAU AND THE PROBLEM OF
PHYSICAL REALITY
W. H. WERKMEISTER
I. A publication by Professor Margenau is always of interest to persons con-
cerned with philosophy of science. This is especially true, however, of his re-
cently published book, The Nature of Physical Reality;' for this book, dealing
with basic epistemological problems arising from the development of modem
quantum mechanics, is the most comprehensive and most systematic formula-
tion of its author's philosophical position and is at the same time conceived as a
"challenge" to "uncritical realism, unadorned operationalism, and radical em-
piricism"-to points of view, that is, which Professor Margenau regards as
"outmoded and in disharmony with the successful phases of contemporary
physics" (v).
Professor Margenau's competende in physics and philosophy of science is,
of course, too well known to require emphasis. As usual, his style is clear, direct,
and forceful. The organization of his book-moving from preliminary epistemo-
logical considerations through point mechanics, field theory, thermodynamics,
probability, quantum mechanics, etc., to a final outline of the "contours of
reality"-is logical and integrative. The argument-for the presentation as a
whole is essentially one argument-is effectively keyed to a minimum of mathe-
matical explanations and demonstrations. It is a pleasure to read this book, and
students in the appropriate courses should find it most helpful. A few typographi-
cal errors and other minor blemishes can easily be eliminated in a second printing.
As far as "the reality of the objects of our daily lives" is concerned, Professor
Margenau finds that his own "constructional view" "gets precisely the same
factual consequences as realism, be it naive, reformed, or critical" (455). This is
so, however, not because the realism of common sense, as realism, provides an
adequate criterion of reality, but because this realism is itself "a first approxi-
mation" to, or "a streamlined version" of, the methodology of science and is
"satisfactory as a working hypothesis" (457). The view of the world of common
sense, like that of science, rests upon "acts of reification" which "impart signifi-
cance to primitive cognition" (455), and thus depends upon "construction" and
"rules of correspondence" in the same sense in which science depends upon them.
There is only this difference: In the case of common sense realism "the rules of
correspondence are so obvious, the passage performed with their aid is so short"
(455) that this simple version of constructionalism is "unable to represent cor-
McGraw-Hill Book Company, New York, 1950; xiii, 479 pages. $6.50.
183
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
184 W. H. WERKMEISTER
rectly the parts of scientific experience made significant by 'long' rules of cor-
respondence" (547). In so far as realism involves the metaphysical thesis that
"the physical world is outside experience," Professor Margenau explicitly re-
pudiates it (455); for such a thesis, he holds, "relinquishes control over reality"
and "leaves science without defense against fairies, ghosts, and goblins" (457).
Unless reality be found in "the counterplay of construction and verification,
there is no available criterion to give reality its warrant and to set it apart from
the unreal" (457).
Professor Margenau's opposition to "radical empiricism" is closely connected
with his distinction between "correlational and theoretic procedures within
science" (28). The former are essentially inductive and/or statistical; the latter
are deductive and involve construction. "What can be attained by induction,"
Professor Margenau points out, "is the probablevalidity of the statement: This
rose will appear red to me whenever I look, but not the factual statement in-
tended by the physicist with all its unadorned simplicity: There is a rose, and
it is red. This requires what we have called construction" (229). Historically all
sciences may start at the correlational level but they evolve progressively toward
the theoretic stage; for when a theory is born "the contingency or correlation"9
gives way to "logical necessity" (28). It is possible, of course, to take the position
that a law of nature is "the mere statement of a reasonably invariant connection
between phenomena" and that "science is the establishment of a universal
catalogue of correlation coefficients between all perceptible phenomena"-not
an uncommon attitude among statisticians. Professor Margenau maintains, how-
ever, and I think rightly, that such an attitude "does violence to certain sciences"
(27-28). "Investigators bent on basic explanation are never satisfied with a
statement of correlation coefficients. . . They feel the urge to probe more deeply,
to derive this strange uniformity of experience from principles not immediately
given" (28); and they hail with acclaim the discovery of "subsurface connections"
which go far beyond "augmenting and organizing empirical knowledge" (29).
In view of the position thus taken by Professor Margenau it is not astonishing
to find that he repudiates as inadequate Carnap's empiricistic interpretation of
the formation of physical concepts (226-231) and Bridgman's operational defi-
nitions (231-232). Both approaches fail to do justice to what Professor Margenau
calls "constitutive definitions" and regards as the very core of theoretic science.
However, Bridgman's operational definitions are, in part, admitted as "rules
of correspondence" connecting "constructs" with the data of sense experience.
The fact that Professor Margenau repudiates also Reichenbach's attempt to
interpret quantum mechanics in terms of a three-valued logic need here be men-
tioned only in passing (461-462). "An equation like p + q = r is in principle
either true or false," and so long as quantum mechanics goes ultimately back
to a limited set of mathematical operations (such as the law of adding integers,
for which tertium non datur) any departure from two-valued logic "does violence
to the straightforward habits of that branch of science" (462).
Professor Margenau's anti-Humean position is epitomized in his interpreta-
tion of causality. To begin with, he points out that the "predictability doctrine"
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
PROFESSOR MARGENAU AND THE PROBLEM OF PHYSICAL REALITY 185
of causality puts the cart before the horse. "That prediction has much to do
with causality is of course not an accident, for if causality holds, then prediction
is always possible." But the fact that "prophetic prediction" is at least conceiv-
able "cautions us not to state this relation the other way" (401).
For Professor Margenau the principle of causality is a stipulation concerning
the kind of laws acceptable in theoretic science: "Causality holds if the laws of
nature (differential equations) governing closed systems do not contain the time
variable in explicit form" (405). The force of the principle of causality, thus
understood, is methodological. It simply means that "we have found a way of
describing our experience which renders the relation between two states, sepa-
rated in time, unique (one implies the other) and invariable (the same impli-
cation always holds)" (407). Since it is at least conceivable that "the so-called
constants of nature" may change in time, the principle of causality is "not tauto-
logical and can be contradicted by scientific experience" (408).
If causality is taken to be a relation of necessity between two or more immedi-
ate sense impressions, then "strict causality has been lost in modern physics"
(419) and Hume was right in treating it as illusory. Sensationalism and/or radical
empiricism fail utterly in their attempts to disclose a causal nexus within im-
mediate experience. Professor Margenau, however, would agree, I am sure, that
their failure is an argument against sensationalism and/or radical empiricism
rather than against the principle of causality as needed in theoretic science. It
was Kant's great achievement, Professor Margenau points out, to have destroyed
the belief that laws of nature directly involve or concern immediate observation;
and if Kant was right here, then "causality reigns in quantum physics as it did
in the classical theory of nature."
Professor Margenau finds that his own analysis of scientific method, his con-
ception of "constructs" and "verifacts," and his use of "rules of correspondence"
clearly commit him to the thesis that "Humean causality is quite definitely
gone" (420) and that Kant was right (420). The spirit of the causal principle-
if causality is understood as a stipulation concerning the kinds of equations that
may serve as laws of nature-asserted itself with renewed vigor in the demands
of the theory of relativity. "The special theory of relativity maintains invariance
[of the laws] relative to inertial systems; the general theory attempts to preserve
it for all space-time transformations. In the deepest sense, therefore, the theory
of relativity is a natural complement of the principle of causality and indeed its
ultimate fulfillment" (411); and if equations of first and second order are ac-
ceptable as laws in theoretic science, then the range of cauisality widens suffi-
ciently to include practically all branches of physics (409).
II. As must be evident to all readers of my own relevant publications2 I find
myself in basic agreement with Professor Margenau's interpretations of the
nature of scientific laws and the principle of causality. However, my sympathy
with his point of view extends even further than this.
2 See, for example, The Basis and Structure of Knowledge,New York, 1948; "Science, Its
Concepts and Laws," The Journal of Philosophy, XLVI, July 1949;and "An Epistemological
Basis for Quantum Physics," Philosophy of Science, XVII, January 1950.
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
186 W. H. WERKMEISTER
It is needless to say that I am in substantial agreement with Professor
Margenau when be says: (a) that epistemology acknowledges as given only " 'ex-
perience,' which presumably means my own first-person experience" (298); (b)
that the reflecting self, along with the objects of common sense or scientific
knowledge, "is initially a construct to be verified" (455; 456; 47); (c) that "reality
is constructed from the elements of [first-person] experience and is still part of
it when fully formed" (298); (d) that "a complete severance of entities from
theories is impossible" (296); (e) that "constructs are not valid because they
refer to something real" but that they "denote something real because they have
been found valid" (292); (f) that the "problem of externality"-the problem of
an external world, that is-is really the "problem of objectivity" (48); (g) that
a construct denotes something real only if there is "at hand a rational context
into which it is or can be integrated" (292); (h) that "other persons" exist for
me in precisely the same sense as do "verifacts" in the rational context of ex-
perience (298-299); and (i) that past "occurrences" are constructs within the
integrative context of "a conjectured temporal course of an objective universe"
(297). Nevertheless, I find in Professor Margenau's book certain lines of reason-
ing which, in my opinion, require close scrutiny and, perhaps, a basic reformu-
lation.
Professor Margenau has set himself the specific goal of determining "the mean-
ing of physical reality" (1; 3; 290) and has restricted his considerations of the
data of experience so as to exclude from discussion but not to deny any other
meaning of reality (289; 462). He realizes, for example, that, "being part of
experience, physical reality cannot function as the why of experience" and that,
as far as physical reality is concerned, the fact of experience itself "remains an
unfathomable mystery" (458; 289). "Science will tell us what things are real
but will refuse to say what is reality" (12).
In his search for the meaning of physical reality Professor Margenau starts
with the "immediately given" as encountered "within experience" (46) rather
than with "the ontological premise characterizing the spectator-spectacle dis-
tinction" (47), and proceeds on the understanding that only "if experience
should not provide a stable basis for an objective world from within itself by
immanent procedures"3should we be "forced to undertake the initial metaphysical
plunge" of realism (47). Professor Margenau hopes to show that such a plunge
is "unnecessary" (47).
As Professor Margenau sees it, the real is (a) "enduring," (b) "thing-like,"
and (c) "efficacious" (1). But we must be most careful in accepting such broad
and loosely defined criteria. Professor Margenau states-to use but one illustra-
tion-that "we want reality to be more permanent than our fleeting sense im-
pressions" (4). I agree-but only conditionally. In view of the fact that the half-
time of certain radio-active elements are very short indeed compared even with
the shortest durations of our most fleeting sense impressions, Professor
Margenau's statement, unless properly qualified, provides no criterion of reality
adequate to modern physics. The "enduring," it seems to me, which alone
I My italics.
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
PROFESSOR MARGENAU AND THE PROBLEM OF PHYSICAL REALITY 187
deserves to be called "physically real" turns out, in the end, to be a system of
energy exchanges rather than some "entities" which, individually and in isola-
tion, are "more permanent" than are our particular sense impressions.
To be sure, Professor Margenau's argument does not depend upon the criteria
of reality mentioned above-at least not exclusively. He explicitly asserts that
"it is the methodologyof science that defines physical reality" (13; 16); and he
views methodology, in this sense, as "a part of metaphysics" (13)-as a set of
''assumptions or rules of procedure which are not dictated by sensory evidence
as such" but "whose application endows a collection of facts with internal
organization and coherence, makes them simple, makes a theory elegant and
acceptable" (13; 81).
The methodological or "metaphysical" rules here in question "first emerge in
the stream of experience as tentative expedients, grow into implicit beliefs . . ..
and finally, strengthened by repeated success, pervade the entire texture of our
theories about the world" (81). Professor Margenau explicitly formulates six
such rules: (I) "Constructs shall be so formulated as to permit logical manipula-
tions"; they shall "obey logical laws" (81). (II) "The character of the connec-
tions which constructs may enter ... may be of two types, formal and epistemic."
"A formal connection is one which sets a construct in a purely logical relation
with another construct; an epistemic connection is equivalent to and arises from
a rule of correspondence which links the construct with data" (84). (III) "A
construct remains what it is so long as the premises of the theory are accepted"
(88). (IV) "Scientists judge the quality, and ultimately the correctness, of a
given theory by its range of application, taking the generality of a system as a
measure not only of its usefulness but of its credibility." In other words, they
adhere to the belief that "ultimately a single theory will render an adequate
account of all experience" (9). (V) Causality is a methodological requirement,
"a relation between constructs," a property, if you wish, of laws rather than of
observations (95). The principle of causality "demands that constructs shall
be so chosen as to generatecausal laws" (96)-and "causal laws" are defined as
first and second order equations (409). (VI) "When two theories present them-
selves as competent explanations of a given complex of sensory experience,
science decides in favor of the 'simpler' one" (96-97).
It is obvious, I believe, that all six stipulations pertain in one way or another
to matters of theory, that they are meta-theoretical (if that term be permissible)
and have little bearing upon the question of reality as such. To be sure, stipula-
tion (II) provides that, through appropriate rules of correspondence, at least
some constructs "stand in uniform correlation with immediate experience" (99),
and that, through continual tests against immediate experience, a high degree
of trustworthiness is conferred upon them. But even this stipulation does not
change the status of constructs. They are, and remain, logical bridges connecting
in various indirect ways the sense data of immediate experience. What Professor
Margenau calls "valid constructs, or verifacts" (105) are but constructs which
serve well as logical bridges. They are integral parts of theories which success-
fully link sense data to sense data but do not transcend the logico-sensory realm
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
188 W. H. WERKMEISTER
in which a radical positivism (of Petzoldt's type4) is at home-a positivism re-
stricted to logical manipulations of concepts correlated with sense impressions
and with the coincidences of sense impressions. I am sure, however, that some-
thing more than this is involved in the idea of physical reality; and Professor
Margenau, I believe, will agree. The question is, What is this "more" and how
can we obtain it? In our deeper quest for the meaning of physical reality the
six meta-theoretical stipulations given above are of no help; and it is at this
point that, as I see it, a basic ambiguity creeps into Professor Margenau's argu-
ment.
The key to Professor Margenau's position is his idea of constructs and of the
r6le which constructsplay in cognitions (292); and since the ambiguity referred to
above involves this key idea, it permeates most of Professor Margenau's argu-
ment pertaining to "the nature of physical reality."
Although physics starts from sense data, returns to sense data for confirma-
tory evidence, and, on the basis of elaborate theories, predicts the occurrence of
sense data, "modern physics is an indictment of the universal adequacy of
common sense'2 and cautions us "against too glib an acceptance of the so-called
'deliverances' of our senses" (44). "What we encounter in purely sensory expe-
rience is ineffably complex and multiple" (51), and language-even the language
of physics-is inadequate to express the full richness of "spontaneous expe-
rience" (57). Every description or interpretation of that experience, therefore,
depends upon rules or procedures of selection and integration "which are not
themselves furnished by sense" (50; 54) but which make the given manifold of
sense experience logically manageable and comprehensible. The employment of
such rules or procedures, according to Professor Margenau, involves recourse to
the use of constructsof various types.
In order to understand clearly what constructs are, let us consider briefly the
difference between a "seen tree" and the "physical object, tree" (59). In the case
of the "seen tree," the visual, tactile, and kinesthetic impressions given in our
actual sense experience are supplemented by sensory qualities, remembered
and/or anticipated. That is to say, a multitude of actual, remembered, and
anticipated sense data are integratedto form the "seen tree." Such a complexus
of "integrative properties," however, does not constitute the "physical object,
tree." The latter involves "more than an abstraction from or integration of
sensory perception" (59-60); it involves "permanence, or continuity of exis-
tence".-a property "which could never be abstracted from data" (60). Accord-
ing to Professor Margenau, therefore, the "act of reification" requires "more than
integrations"; it requires construction-"construction in accordance with rules"
(60).
It seems to me that, despite a difference in terminology, what Professor
Margenau here calls "construction in accordance with rules" I have described at
some length in a previous publication.5 Our agreement on this point becomes
I See, for example, Petzoldt, Joseph, Die Stellung der Relativitatstheoriein der geistigen
Entwicklung der Menschheit, 2nd edition, Leipzig 1923.
6 The Basis and Structure of Knowledge, Chapter 3.
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
PROFESSOR MARGENAU AND THE PROBLEM OF PHYSICAL REALITY 189
quite evident when we consider that, according to Professor Margenau, the
simplest application of the rules of construction is "the act of postulating a
thing in the face of certain sensory evidence" (64) but that such a "postulated
thing" cannot be certified as an external objectby the rules of construction alone;
that such a certification requires rather a "documentation which refers to the
coherence of our entire experience" (64).6 Beyond this point of agreement,
however, I encounter certain difficulties with Professor Margenau's interpreta-
tion and use of the idea of constructs.
A construct, Professor Margenau says, "has many of the qualities of an
invention" (70). Trees, masses, and electrons, therefore, "are more than mere
constructs." "They are valid constructs" or "verifacts" (70; 99; 105; 450)-as
distinguished from "ghosts, mirages, and the luminiferous ether," which are
mere constructs (70). The term 'construct,' therefore, as here employed and as
used without qualification, merely assigns to trees, electrons, ghosts, and devils
their "correct genetic status in experience" (70).7 It does not ascribe to them a
reality outside the experiential situation. Moreover, "there exists a large area of
discourse in which the word construct... is wholly synonymous with concept"
(70-71). Nevertheless, Professor Margenau maintains that "the tree is the con-
struct"; that "there is not a tree and my construct of it" (70); and that "the
electron, as an external object,8is the construct" (71).
The juxtaposition of the two ideas-(a) that constructs merely assign to
various objects their "correct genetic status in experience" and (b) that the
constructs themselves are the "external objects"-disturbs me.9 I find these two
ideas irreconcilable; and Professor Margenau's further discussion of the matter
only adds to my difficulties.
"Tree, molecules, electrons, genes," Professor Margenau says, "are not com-
pounded from sense impressions, past and present, alone. They also contain
rational elements which point beyond all the aspects of immediacy that go into
their making" (71). I am puzzled by such a statement. To be sure, if by trees,
molecules, electrons, and genes we mean ideational complexes in terms of which
we interpret experience, then it may be permissible to say that they are "com-
pounded from sense impressions, past and present," and certain "rational ele-
ments." But if trees, molecules, electrons, and genes are real things in a real
external world, then either they are not compounded from sense impressions and
rational elements or the whole of the external world-this "choir of heaven
and furniture of the earth"-is reduced to a Berkeleyan "idea." Professor
Margenau, I am sure, would repudiate the thesis that the "external world" is
but a Berkeleyan "idea." But does not such repudiation require the repudiation
of the premise which entails the thesis?
6 See my own discussion of the significance of the coherence of experience as a criterion
of reality. Op. cit., 101-115.
7 My italics.
8 My italics.
9 Professor Margenau willingly accepts it because he feels that in the idea of the con-
struct we have a "union" of particulars and universals (72; 71; 303) which avoids the nomi-
nalism-realism issue. I shall not discuss this problem in the present article.
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
190 W. H. WERKMEISTER
I submit that if the term 'construct' were used to designate the ideational
complexes referred to above, one could then regard the real trees, molecules,
electrons, and genes as inferred entities and could thus save the realism implied
in the orderliness of experience and demanded by science. The Berkeleyan thesis
could then be avoided, but constructs could not be directly identified with things.
At times Professor Margenau argues as if he accepted the distinction just made;
but even his clearest statements along this line retain a certain ambiguity. "By
virtue of a largely unanalyzed rule of correspondence," he says, "we transcend
the realm of immediacy and construct the external object, flower. The postula-
tion of an external objectis the first phase of the cognitive act" (172). I find no
evidence in human experience that we ever constructa real flower. Flowers grow
and come into being as integral parts of the "pattern of things." However, I
agree that on the basis of sensory and rational elements in experience we do
construct the ideational complex "flower"-a complex, that is, which, in combina-
tion with other evidence inherent in our experience, leads us to the postulation
of an external object flower as an inferred entity. It is possible that Professor
Margenau has a similar interpretation in mind when he says that "external
objects" are "constructional entities" which function in the "substantival r6le"
as "carriers of observable properties", and which he calls "physical systems"
(172; 176); but I am by no means sure about Professor Margenau's meaning
because the term "constructional entities" raises once more the specter of an
ultimate identification of constructs and external objects.
I encounter the same ambiguity in Professor Margenau's contention that the
physicist "constructselectrons, atoms, and so forth, talks about them as [physical]
systems, and uses them as the carriers of observables" (229); for the reference
to "physical systems" as "carriers of observables" might well point to inferred
entities. But Professor Margenau states specifically that although he has spoken
at times as if "verifacts were to 'refer to' and to 'denote' something real," he
wants "to correct that usage, lest it give rise to some spectral reality behind
valid constructs"; and he re-affirms the thesis that "verifacts, as part of expe-
rience, are reality" (296)-even though they may not be the whole of reality.
In view of this explicit statement we have no alternative but must accept as
central to Professor Margenau's theory the proposition that constructs which
have been confirmed in experience-the so-called verifacts-are literally the
real and external objects investigated by physicists. What are the implications
of such a proposition?
If the constructflower is the real flower, does the construct wilt when frost
destroys the flower? Or is it the constructfrost that destroys the constructflower?
These questions are not as outrageous as they may seem. Professor Margenau,
it must be remembered, maintains that atoms, electrons, photons, mesons, and
so forth, are constructsand that these constructs-misleadingly called "physical
systems"-are what is meant by real objects. Yet, in measuring a photon's
position the physicist usually destroys the photon, as he destroys mesons and
neutrinos when he subjects them to measurement (373). "Perfectly good meas-
urements in atomic physics ... may 'kill' or annihilate a [physical] system"
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
PROFESSOR MARGENAU AND THE PROBLEM OF PHYSICAL REALITY 191
(377). Are we to take literally the implied assertion that measurements of the
type here referred to actually destroy constructs compounded from sensory
and rational elements of experience? Or are we to escape such consistency be-
cause of the term "physical system" which, though in a strict sense but a con-
struct of a certain type, suggests, by associations and overtones, an inferred
entity?
Professor Margenau himself, when facing this issue, prefers consistency with,
rather than a revision of, his initial commitment to the proposition that con-
structs are the external objects. In answer to the question, "How can a construct
be real before it is constructed?," he says that "the acceptance of a construct as
valid, even if temporary, must ... bring with it an imputation of its permanence
and hence project its reality into the past. When this is recognized, it is no
longer paradoxical to say: After a construct is validated, it must be said to have
been real before it was formed" (294). More specifically, Professor Margenau
admits that "reality does change as discovery proceeds' and that he "can see
nothing basically wrong with a real world which undergoes modifications along
with the flux of experience" (295); that "reality changes when new discoveries
are made" (459) and that this change indicates "the dynamic quality of physical
reality" (459).
If I understand these passages correctly, Professor Margenau is saying that
prior to the development of relativity theory the physically real world (not our
picture of it) was Newtonian; that prior to. the development of modern wave
mechanics the real atomsactually were as Bohr described them in his now classical
theory; that reality (not our interpretation of it) changed when the first radio-
active elements were discovered, and that it changed again when photons,
mesons, and neutrinos were also discovered. To me such a conception of the
"dynamic quality of physical reality" seems to be not only contrary to the most
reasonable view of the history of science (for it makes that history an ever
changing creation and re-creation of reality) but self-contradictory as well. If,
when validated, a construct was "real before it was formed," then the physically
real world must have been relativistic and Newtonian at the same time; the
real atoms must have been as Bohr described them and as modern wave mechanics
interprets them; and reality must have existed at the same time with and with-
out photons, mesons, and neutrinos. Finally, if constructs are "compounded
from sensory and rational elements of experience," what does it mean to say
that they were real before they were formed? Would not an acceptance of the
idea of inferred entities and of a reality distinct from the constructseliminate the
difficulties here alluded to?
One more point may deserve consideration. According to Professor Margenau,
the so-called properties of external objects are also constructs (172) and are
"joined to sensations by rules of correspondence" (172). Some properties-such
as "the number of stamens and pistils, the color, shape and composition of
petals" of a flower-Professor Margenau calls "observables" (176); others-
such as the properties of an electron-he regards as "latent observables" (176).
His terminology, however, is not completely consistent and it may be truer to
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
192 W. H. WERKMEISTER
his general point of view to regard all properties with which objects functioning
as their carriers are invested as latent observables.10The important point is that
in assigning properties to external objects we remain entirely within the field of
constructs (173).
If we now examine the constructs which are of special interest to the physicist-
such as mass points, electric fields, electrons, and photons-we find that they
are created by "constitutive definition" (237),11 i.e., by a "postulated grouping"
of latent observables (240). This means, however, that the "substantival" con-
structs-which, according to Professor Margenau, are the real external things-
can and do possess, as "latent observables," only those properties specifically
assigned to them in a constitutive definition. In classical mechanics, the con-
struct mass point is thus established when, in addition to its "inherent property,"
mass, it is invested by constitutive definition with five properties, as latent
observables, involving position and velocity (178; 179). By definition, therefore,
the mass point of classical mechanics has six properties.
But why do we invest it with six properties and not with more? The answer
is that the pertinent laws of classical mechanics are "self-sufficient with respect
to states defined in terms of these six quantities" (180). This, however, is but
another way of saying that the constitutive definitions of such constructs as
mass points, electric fields, electrons, and photons are determined by the system
of laws in which they are incorporated, and thus by the theory accepted at any
given time. The constructs, being systemically integral, can possess only those
properties which are required by the logical coherence of the theory. But since,
according to Professor Margenau, these same constructs are also the real external
objects, it is difficult to see how reality can ever be more than what is created by
constitutive definition, or how experiments can ever disclose new properties of
the physically real. I fear that by eliminating inferred entities and a reality
"beyond" sense impressions and the rational elements of experience-a reality
which is other than our systemic constructions and re-constructions-Professor
Margenau has demolished the realism basic to all experimental science. Although
his book contains most illuminating chapters on modern physical theory, he has
not solved the problem of physical reality.
University of Nebraska
10 See the statements: "The property, blue, with which I invest a substantival flower
is other than the sensation I have of it in a particular perception" (172). And: "The observ-
able is a kind of abstract quality, assigned as a latent attribute to objects .... It is as
though the flower had a latent color" (175).
11 "For the statement, an electron, or an atom, or a mass point, or a field, is that which
has such and such properties, is a constitutive definition" (237).
This content downloaded from 144.173.006.094 on June 27, 2020 14:01:31 PM
All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
