<< . .

( 29)

. . >>

of analysis.
How can contradiction bypass theory and appear directly at the level
of observation? Easy cases are two thermometers, whether of the same
or different construction, placed in the same medium but reading differ-
ent temperatures, or an object that shows different weights when placed
on different scales, or on the same scale, one minute from the next. So,
you say, something must be wrong with at least one of the thermome-
ters, or with the scale. Apparently they are not good measures of tem-
perature and weight. But the only evidence we have that there are such
objective properties as temperature and weight at all is that there exist
ways of making thermometers match consistently and ways of making
scales that weigh consistently. We can, of course, turn to our theories
about the causal properties of temperature and weight to explain why
thermometers and scales agree when they do and why they don™t agree
when they don™t. We also may turn to theories when it is necessary to
repair or calibrate our thermometers and scales. But having adjusted the
thermometers and scales, whether by employing a theory, or by trial
and error tinkering, or by sheer accident, no theories are implicated in
the use of these instruments to confirm the objective adequacy of our
concepts of temperature and weight. Similarly, making a prediction that
a certain proposition will come true as a result of performing an exper-
iment and later judging perceptually that it has indeed come true is a
way of testing not only the theory but the objectivity of the concepts
involved in the judgment. And these two are independent tests.That the
concepts have been reaffirmed as objective does not depend on the the-
ory itself being true. The method of prediction used may work for a
reason independent of the particular theory, as ancient astronomical
predictions often proved accurate for good reasons but not for the rea-
sons the astronomers thought at the time.

Most people purchase their thermometers and weight scales know-
ing nothing of the principles of their construction and operation. This
does not make these people™s concepts of temperature and weight less
well epistemologically grounded than those of the scientist. It does not
give them less reason to place trust in the objective meanings of these
concepts. Similarly, all of us were natively endowed with perceptual sys-
tems whose principles of operation scientists are barely beginning to
fathom. Trust in the objective reference of judgments made using these
systems is warranted in so far as we agree each with ourselves in these
judgments: if p then p.What feels cubical looks cubical, and continues to
look cubical from different angles and distances. What sounds as if in
front of me looks to be in front of me and can be attained by reaching
in front of me “ like measuring temperature with a mercury ther-
mometer, a gas thermometer, and a bimetallic strip. Nor are even the
most basic perceptual self-agreements logically necessary. Müller-Lyer
arrows measure the same length but look different lengths. After watch-
ing a waterfall closely and continuously for a minute or two, if the eyes
are then fixed on a stationary object it will appear to be at once mov-
ing and stationary (Crane 1989). There is a way of focusing one™s eyes
on a pair of spots, one red and one green, such that there appears to be
only one spot that is both red and green all over. A and B can appear to
be the same color, B and C can also appear to be the same color, while
A and C appear to be different colors (Goodman 1966). That the affir-
mative and negative stimulus meanings of any perceptual judgment are
mutually exclusive is not a necessary truth but a matter of experience,
and a continual reaffirmation of the objective meaningfulness of the
empirical concepts used in making the judgment.


That concepts are tested and honed in ways that do not entangle them
with theories does not imply, in general, that they can be tested singly
or one by one. Adequacy in concepts is tested by whether their em-
ployment makes stable judgment possible. But no judgment employs
only one concept. To make the same judgment again, one must recog-
nize its subject or subjects as being the same again and also the proper-
ties or relations it attributes as being the same. Both subject and predi-
cate terms must be adequate if stable judgment is to result. Equally
important, that a judgment is stable implies that it might have been un-
stable, that one might have fallen into contradiction instead. Adequacy

in concepts can be tested only if one can recognize contradiction in
judgment. And this requires the capacity to recognize the complements
or contraries of predicates. Let me explain.
Consider Quine™s observation that many stimulations may be ex-
pected to belong neither to the affirmative nor negative stimulus mean-
ing of an occasion sentence. One important reason is that the mere ab-
sence of affirmative stimulation does not constitute negative
stimulation. Most obvious, I cannot make either an affirmative or a neg-
ative perceptual judgment if I fail to recognize its subject. If I don™t see
the rabbit at all, I can™t judge it to be white or not to be white. Less ob-
viously, failing to perceive that the predicate of a proposition applies to
its perceived subject does not warrant judging it not to apply. I may feel
the apple in the dark, know it is an apple, even know which apple it is,
but I cannot judge its color by feeling. I may strike a match and look at
the apple, but still not be able to see its color clearly or at all. Not ob-
serving that the apple is red does not equal observing that it is not red.
To tell that it is not red I must be able to tell what other color it is in-
stead, that it is some contrary of red or, more generally, that it is non-
red, the complement of red. Having concepts of the contraries and
complements of predicates is required if negation in judgment is to be
possible, hence if contradiction in judgment is to be possible. To judge
that it is not blue, you must be able to judge that what you are seeing
is its being grey, not, say, its being in shadow. To judge that it is not
round you must be able to judge that what you are seeing is its being
elliptical, not, say, its being at an angle.
It follows that subject concepts can be tested and honed only along
with at least some applicable predicate concepts and also complements of
these. It also follows that Quine was at least close to right about the em-
pirical status of at least one law of logic, the law of noncontradiction ap-
plied to empirical judgment. It is an empirical matter that we can carve
out concepts of objects along with concepts of properties and their con-
traries such that the object concepts are suitable to be subject terms for
empirical judgment, each consistently taking just one contrary from each
of a series of predicate contrary spaces. Just as it is an empirical matter
whether anything real has Euclidean structure, it is an empirical matter
that there exist objects to judge about that have properties discernable as
stable over a variety of perspectives. It is an empirical matter, that is, that
there exist any “substances” as these were described in Chapter 2.
In Sellars™ famous myth of the necktie shop, Jim teaches shopkeeper
John to use the language of “looks” and “seems” after the installation of

electric lights has caused John to misjudge the color of one of his

“But it isn™t green,” says Jim, and takes John outside.
“Well,” says John, “it was green in there, but now it is blue.”
“No,” says Jim, “you know that neckties don™t change their colors merely as
a result of being taken from place to place.”
“But perhaps electricity changes their color, and they change back in day-
“That would be a queer kind of change, wouldn™t it?” says Jim.
“I suppose so,” says bewildered John.
(Sellars 1956, pp. 270“1)

Here Jim convinces John to recalibrate his ways of judging color con-
traries, of making negative color judgments, by appealing to stability of
judgment across change in perspectives and conditions as an ideal. Still,
his argument for misperception rather than change of color seems
rather weak. What really is the evidence that the necktie does not itself
change when placed under incandescent light, that the distal stimulus is
constant despite the proximal variation? Isn™t that a matter of stability in
distal causal properties, hence a matter of law, hence of theory? But the
evidence against distal change need not digress through theory. Unless
other ways can be found of observing this supposed change, unless
other perspectives can also reveal it, there is no evidence for it™s reality.
Evidence for the objectivity of objects and properties can only be ob-
tained by triangulation, triangulation in that there is variety in the kinds
of evidence for them. I have argued that triangulation can be achieved
through variety in perception taken alone, and it can also be achieved
by the use of theory. Presumably neither route is possible in the case of
the necktie™s change of color.
The myth of the necktie shop raises another and broader question,
familiar to us from the earlier discussion of conceptual development
(Section 5.7). How does one know what kinds of properties can be ex-
pected to be stable over what kinds of perspectives and for what cate-
gories of objects? Animals regularly change their shapes over short
stretches of time whereas most other physical objects do not. The
material gold, as discovered in different places, has any of innumer-
able shapes and sizes, but is stable with respect, for example, to density,
color, malleability and resistance to corrosion.The frog species Rana pip-
iens, as observed in different places and different times, is quite stable
with respect to adult size, with respect to the variety and placement of

its inner organs, and pretty much all of its behavioral dispositions, but
not, say, with respect to shape (postural attitude) or the contents of its
stomach. Acquiring concepts of these various substances must involve
some understanding of which predicate contrary spaces are correlative
to them, that is, of the “substance templates” under which they fall (Sec-
tions 1.8 and 2.7), such as person, animal, animal species, plant, plant
species, mineral, and so forth. Thus the claim that theories need not be
involved in the development and testing of empirical concepts does not
imply that no concepts are interdependent.
On the other hand, the ability to reidentify substances is required to
guide practical as well as theoretical activities. The practical use of the
capacity to reidentify important individuals, kinds, and stuffs probably
long predates theoretical conception, ontogenically as well as phyloge-
netically. In order to accumulate knowledge over time of how to deal
with any individual or kind or stuff, also in order to apply what has
been learned, an animal must be able to reidentify it over various en-
counters under a wide variety of circumstances. Practical tests of the ad-
equacy of substance concepts are independent of other concepts, thus
providing a certain sort of foundationalist base for the conceptual abil-
ities later employed in theoretical knowing.

Content and Vehicle in Perception1


I have tried to show that the ability to reidentify things that are ob-
jectively the same when we encounter them in perception is the
most central cognitive ability that we possess. It is an extremely dif-
ficult task, deserving careful study by psychologists and neuroscien-
tists as well philosophers. But in order to study how a task is per-
formed one must begin, of course, with some understanding of what
that task is. We have not yet asked in what the act of reidentifying
The question is made more difficult by a tradition we have all been
trained in, philosophers and psychologists alike, that takes the answer to
be obvious. Answers to various other questions have then been con-
structed on this implicit foundation, so that challenges to it have be-
come both hard to understand and anxiety producing. This traditional
answer is that reidentifying an object or property in either perception
or thought consists in being able to discriminate it, and that this ability
is manifested in sameness of one™s reaction to the object, or sameness of
one™s treatment of it, or sameness of the mental term or concept one
applies to it. That is, reidentifying is repeating some kind of response.
Call this “the repetition view of reidentifying.”
One familiar doctrine constructed on the repetition view is that
when sameness in the referential content of two perceptions or
thoughts fails to be transparent to the thinker, this is because the con-
tent is not thought of in the same way both times. It is because one

1 Portions of this chapter were revised from “Perceptual Content and Fregean Myth” (Mil-
likan 1991), with the kind permission of Oxford University Press.

does not repeat one™s way of thinking of it, because the referential con-
tent is not thought of under the same mode of presentation. To fully
describe the content of a person™s thought thus requires indicating in
what way, under what modes of presentation, the various objects of
their thought are grasped. A second familiar thesis is that wherever
identity of referential content fails to be transparent, this identity can
only be grasped by making an identity judgment correlating the two
modes of presentation. In the following chapters, I will try to show that
these views are mistaken. I will argue for another view of the act of
identifying, and supply other tools with which to understand the phe-
nomena that modes of presentation and judgments of identity were in-
troduced to explain.
The point to be made about grasping sameness is a very abstract ap-
plication of a more general point that pertains to all varieties of mental
representation. It will be easiest to explain using quite concrete examples
taken from the realm of perception. In Chapter 5, I tried to show how
perception of substance sameness was in certain ways similar to or even
continuous with cognitive understanding of substance sameness.The ba-
sic lesson to be learned about cognitive grasp of identity, also, is applica-
ble to the theory of perception. So here I will temporarily broaden the
focus, beginning with points that may at first appear to concern percep-
tion alone, only later applying them to cognition. The chapter will be
mainly negative. It is no help to introduce a new theory of what grasp-
ing sameness consists in unless a need for it has been shown.


In its most general form, the confusion that produces the repetition
theory of identifying is found also in classical representational theories
of perception. It consists in a confusion or mingling of the intentional
contents of a representation with attributes of the vehicle of representa-
tion. For a starting intuition, compare Kant™s suggestion in the Analogies
that Hume had confused a succession of perceptions with a perception
of succession. In the case of the repetition view of reidentifying, I will
later argue, the error consists in confusing sameness in the vehicle of
representation with a representation of sameness.
Classical representational theories of perception typically were moti-
vated by an argument from illusion. Verbs of perception all are, in the
first instance, achievement verbs. In the primary sense of “see” you can-
not see what is not there to be seen, you cannot touch what is not there

to be touched, and so forth. If there is perception at all, there must be
a real object that is perceived. Add to this the fact that perceptual illu-
sion is possible. Straight oars look bent in the water, and the same
bucket of water may be perceived as cold by one hand and hot by the
other if one hand is first heated and the other first cooled. A simple step
takes us to the conclusion that what is directly perceived is never the
real world, but merely an inner representation or picture of it. The rep-
resentation really is bent, or cold at one hand and hot at the other.
In part, the temptation to make this move results from missing words
in the language. In the realm of conception we have the term “know,” an
achievement verb, but we also have another term “believe,” which is a
verb only of trying. If you know something, it has to be true, but if you
merely believe it, it may or may not be true. Missing are verbs of trying
that contrast in this way with the achievement verbs of perception. Sup-
pose then that we introduce a general term for what stands to perceiving
as believing stands to knowing. I will coin the term “visaging” for this
purpose. Let it stand for apparent hearings and touchings and smellings
and so forth, as well as for apparent seeings. There is little temptation to
conclude from the fact that you can undergo illusions of knowing that
what is known is never what the world is like but only what one™s repre-
sentations of the world are like.The parallel conclusion in the case of per-
ceptual illusion is equally easy to refuse if we allow ourselves to speak of
visaging things we are not actually seeing. Believing wrongly about things
in the world is not knowing about an inner realm that mediates between
me and the world.Visaging things in the world wrongly is not perceiving
an inner realm that mediates between me and the world either.
But classical theories of perception claimed otherwise. Clearly there
is nothing that we know when we believe falsely. But according to clas-
sical theories of perception, there is something that we perceive when
we visage falsely. The intentional object of a visaging is always some-
thing real, but not, of course, something in the ordinary world. Just as
primitive peoples take dreams to be knowledge, though knowledge of
another realm, classical theories of perception take illusory visagings to
be knowledge, though knowledge of another kind of object.
Visagings were taken to be graspings of, awarenesses of, a realm of
representations, and representations, on nearly all the classical views, are
likenesses. Visagings were taken to involve items appearing before the
mind that are similar to what they represent, hence that have the prop-
erties that they represent. The properties claimed by visagings to char-
acterize the world exist in “objective reality” (Descartes), or they, or

doubles of them, are true of sense data, or percepts, or phenomenal ob-
jects, or visual fields, and so forth. Not exactly the same properties, per-
haps, but at least properties having something like the same “logical
form.”When the world resembles the inner picture, then the visaging is
veridical, showing how things really are. But like pictures drawn with
the purpose of showing how things are, visagings can also misrepresent.
Gareth Evans calls this sort of move “the sense datum fallacy,” and
then says, “[i]t might better be called ˜the homunculus fallacy™ . . . when
one attempts to explain what is involved in a subject™s being related to
objects in the external world by appealing to the existence of an inner
situation which recapitulates the essential features of the original situa-
tion to be explained . . . by introducing a relation between the subject
and inner objects of essentially the same kind as the relation existing
between the subject and outer objects” (1985a, p. 397). He thus suggests
that the main problem with this sort of view is that it invokes a regress.
How will the inner eye then perceive the inner picture? In the same
way that the outer eye does?
I think this is a mistaken analysis, that regressiveness is not really the
problem. Nothing forces a regressive answer to the question how the
inner eye works. After all, the purpose of introducing inner representa-
tions was to account for error, but there seems no reason to suppose
that the inner eye would have the problems the outer eye does of some-
times misperceiving what was there before it. So there would be no
need to suppose that it must use additional still-more-inner representa-
tions in order to see. What is wrong with this classical view, I submit, is
the story that it does tell about how the inner eye works. Having pro-
jected the visaged properties into the direct presence of the mind, the
classical assumption is that there can be no problem about how these
properties manage to move the mind so as to constitute its grasp of
what they represent.Their mere reclining in or before the mind consti-
tutes the mind™s visaging of them and their contents. They are before
the mind, hence the mind is aware of them, hence of the properties
they embody and represent. That™s all there is to the story. Call this the
“passive picture theory” of inner representation.
The passive picture theory produces a facade of understanding that
overlooks the need to give any account at all of the way the inner un-
derstander works, any account of the mechanics of inner representation,
any account of what kind of reacting is comprehending. Clearly it must
be the mind™s reaction that constitutes its understanding of the content of
an inner representation.The mere being of the representation cannot by

itself constitute an appreciation of it. Rather, the inner eye or mind
must understand the representations before it by reacting to them ap-
propriately, by being guided by them appropriately for purposes of
thought and action. But once you see that it must be the mind™s reac-
tion that constitutes understanding of an inner representation, you see
that the picture part of the passive picture theory is also suspect. Why
would a picture be needed to move the mind appropriately? At least,
wouldn™t something more abstractly isomorphic do as well?2
Perhaps no philosopher explicitly holds quite the passive picture of
perception today. But there are vestiges of this way of thinking in many
modern discussions of perception. The passive picture theory has left its
mark in arguments that implicitly move from the fact that certain prop-
erties are visaged to the conclusion that the vehicle of the visaging must
also have these properties. Or they move from the assumption that the
vehicle of visaging must have certain properties to the conclusion that
these properties must be ones that are visaged. Let us look in detail at
some of these moves.


The passive picture theory projects properties claimed in or by the vis-
aging onto the inner vehicle of the visaging. Call this move “content
internalizing.” It also projects properties of the vehicle of the visaging
into the visaging™s content. Call this move “content externalizing.” The
illusion is thus created both that one directly apprehends aspects of the
nature of the vehicle of perception in apprehending the visaged object,
and also the reverse, that one can argue from the nature of the vehicle
of perception to what must be being visaged.
One result of these moves is to make it appear problematic how in-
consistencies could occur in the content of a visaging. Inconsistencies in
content would have to correspond, per impossible, to inconsistencies in
the actual structure of the representation™s vehicle. We can call this the
“demand for consistency” in content. The demand for consistency in a

2 I have taken the position that thinking and perception probably both involve inner repre-
sentation and that representation involves abstract mappings by which representation are
projected onto representeds (Millikan 1984 and elsewhere). See also Sections 14.2“4 be-
low. But this claim does not entail that any particular concrete properties and relations are
shared by representation and represented. Nor is it likely to be open to merely philosoph-
ical demonstration which abstract mathematical relations are shared.

<< . .

( 29)

. . >>

Copyright Design by: Sunlight webdesign