<< . .

( 29)

. . >>

low that they simply have not concerned meanings. Conception is def-
initely one of the things that “meaning” can mean. But these studies
need to be reinterpreted as studies of conceptions if their value is to be
secured. For example, their relevance to understanding word meanings
will be problematic should it be common for different speakers of the
same language to have quite different conceptions of the same substance
yet use the same word for it. And indeed, beginning in Chapter 6, I will
argue that radical divergence in speaker conceptions of exactly this sort
are the rule rather than the exception.


If substance concepts are not just classifiers, so that conceptions of sub-
stances are not what determines the extensions of their corresponding
concepts, how then are the extensions of these concepts determined? To
argue that the extensions of substance words are not determined by
conceptions and to explain how these words do hook onto their ex-
tensions instead was, of course, supposed to be part of Kripke™s project
in Naming and Necessity (1972) and Putnam™s project in “The Meaning
of ˜Meaning™ ” (Putnam 1975). If Kripke was right, “Bill Clinton” does
not attach to Bill Clinton by means of speakers associating with “Bill
Clinton” any particular properties or relations, ipso facto not by means
of associating with it any inner or outer causes, any essence, any partic-
ular kind of ontological ground. A proper name is not, as I have been
putting it, a classifier but an identifier. Similarly, Putnam argued that to
call a thing “water” or “elm” is not to describe it. Natural kind terms
do not work by being associated with properties. Rather, the extensions
of “water” and “elm,” like the extent of “Bill Clinton,” are natural units
in nature, units to which the concepts water and elm do something like
pointing, and to which they can continue to point despite large
changes in the properties the thinker represents these units as having.
Taking Lakoff ™s example discussed in Section 1.7, large changes can oc-
cur in the manner in which a child identifies cats, hence in the things
the child is willing to call “kitty,” without affecting the extension of the
child™s word “kitty.” The difficulty, of course, is to cash out the
metaphor of “pointing.” Speaking literally, what determines which sub-
stance is the extension of a given substance term or concept on the
present view?

Putnam spoke of “indexicality” rather than “pointing,” and Kripke
suggested (without actually endorsing it) a causal theory of the refer-
ence of proper names. As mentioned earlier, a difficulty with both of
these suggestions was that they tended to collapse into more compli-
cated descriptionist views (see Fumerton 1989).The situation was exac-
erbated by the fact that Kripke and Putnam both used a form of argu-
ment that seems to assume the very thing it is trying to disprove. Their
arguments proceeded by offering examples, taken either from this world
(Kripke) or from other possible worlds (Putnam), in which our intu-
itions cry out that wrong results follow from some particular classical
view of the kind of description that determines reference. But if our in-
tuitions are really the final judges here, that would certainly suggest that
we have in mind what determines reference, and this brings us back ei-
ther to a descriptionist view, or at least to a conceptionist view. Com-
pare Russell™s early view of demonstratives, where “This is a book” un-
packs into something like “the thing at which I am pointing is a book.”
And compare Ned Block™s view (1986), according to which the way we
would make decisions about its extension in possible cases determines
what sort of Kaplan-style character, what sort of function from possible
worlds to extensions, a concept expresses, for example, whether its ref-
erent is indexically determined as its cause, or as bearing some other re-
lation to the concept.
Still, what else could possibly determine the extensions of our con-
cepts if not our own intentions or dispositions? And if something else
does determine the extension, what determines that this is what deter-
mines it, if not our prior intentions or dispositions? After all, what is an
extension anyway? What is it for something to have an “extension”? Isn™t
an extension made into an extension by the fact that we apply the word
“extension” in this way and not that? Aren™t extensions determined by
how we intend or are disposed to apply the term “extension”?
Is there a way to stop pulling at our bootstraps?
Consider: No one supposes that the function of vision is determined
by the intentions of the individuals who happen to have eyes. Similarly,
why should functions of the developmental processes responsible for
concept formation and the functions of the concepts these processes
shape be determined by the intentions of the individuals in which these
processes happen to occur? It is not the purposes of individuals, but the
biological functions “ the unconscious purposes “ of their inborn con-
cept-tuning mechanisms that connect substance concepts with certain

I have proposed a theory telling what the most general function of
substance concepts is. It is their job to make it possible to utilize sub-
stances as these are objectively defined in nature for purposes of gather-
ing and applying information. In order to do this, they must include
skills in reidentifying substances. Only in so far as they succeed in this
task can they help us to proceed with successful inductions. My claim
will then be that the extension of a substance concept is whatever sub-
stance in the world it is the job of that particular concept, given its par-
ticular phylogenetic and ontogenetic history of development, to be rei-
dentifying or conceptually “tracking.” Many mechanisms are involved in
the development of a substance concept, and different kinds of mecha-
nisms effect the development of concepts of different kinds of sub-
stances. Some speculations about these mechanisms, and about their
particular phylogenetic and/or ontogenetic origins, will be offered in
Chapter 5. But the rough idea is that the specific functions of the
mechanisms or abilities responsible for originating a particular substance
concept, whether these functions or abilities originated through evolu-
tion of the species or through individual learning, determine whether
the concept is of anything definite and if so of what. This claim will
need to be filled out a great deal before it will be plausible (or clearly
understandable). The core of the project will be an analysis of what an
ability, hence an ability to identify, is that does not equate abilities with
dispositions (Chapter 4), and that explains how it is determined what a
particular ability is an ability to do, even though the ability may be cur-
rently operating under conditions that fail to express it properly (Chap-
ters 4 and 14).
You can call whatever the conception filling out a certain person™s
concept happens to corral, that is, whatever the person takes to be part
of the concept™s extension, by the name “the concept™s extension” if you
like. Humpty Dumpty was right about that. But then “extension” be-
comes a notion with very little interest, and we will need to coin an-
other term for the thing it was the real purpose of that particular con-
ception to capture. A parallel would be to label whatever a frog happens
to snap up with its tongue reflex “ for example, beebees “ as its “prey,”
and then have to coin another term to designate the things its reflex
snap was designed to capture.

The Nature of Abilities:
How Is Extension Determined?


The conception you have of a substance does not determine the exten-
sion of your concept. The extension is the extent of a certain substance
in nature, not whatever you would identify as part of the extension. But
the extent of which substance? That question is crucial. What deter-
mines, in the particular case, what particular substance one™s perhaps
stumbling, sketchy, and inadequate conception is aiming at? This chap-
ter will make some progress toward answering that question. Further
pieces of the puzzle will be added in Chapter 5, and the last pieces will
finally settle into place at the end of Chapter 14.
Substance concepts are abilities of a certain kind. They are, in part,
abilities to reidentify their assigned substances. How are these substances
assigned? It is not a function of the cognitive systems as handed down
by natural selection to identify any particular substance. Natural selec-
tion did not endow me with the ability to identify either 1969 Ply-
mouth Valiants, or gasoline, or my husband. What I was endowed with
was the capacity to acquire these abilities. Thus the general form of the
question what determines the reference of a certain substance concept
is: What determines what a learned ability is an ability to do? It will
help to tackle the matter in this entirely general form.
The question of what abilities are deserves a lot of attention that it
hasn™t gotten. The modern philosophical tradition has unreflectively as-
similated abilities to capacities and capacities to dispositions.This affords
a slippery slope. A “capacity” can be either a living thing™s abilities (the
capacity of a camel to go without water for weeks) or a nonliving

thing™s dispositions (the capacity of gold to resist corrosion) but abilities
and dispositions, in these contexts, are not the same. An ability is very
much more than a disposition in one way, I will argue, and less in an-
In the usual philosophical sense, for something to have a “disposi-
tion” to behave in a certain way, say to do A, is for there to be circum-
stances under which it will do A in accordance with natural law. But to
say merely that there are circumstances under which a thing will do A
is nearly vacuous. If put in the right circumstances “ for example, if
hooked up to the right gadgets and so forth “ any kind of thing can
probably be made to supply a contributing cause to any kind of out-
come you please. So usually one has in mind a disposition to do A in
some definite circumstances C. To speak without restriction of a dispo-
sition to perform in a certain way must be either (1) to imply that the
actual conditions, or the most likely conditions, are such as to realize
the disposition, or such that it is often realized or (2) to assume or im-
plicitly to refer to some specific sort of circumstances in which it will
be realized. It is in the first way that people are said, for example, to
have bad dispositions or sunny dispositions, or to be well disposed to-
ward one another, and so forth. It is in the second way that salt is said
to have a disposition to dissolve.
Sense (1) certainly is not the same as the notion of having an ability.
Sunny dispositions are not, as such, abilities. Moreover, we all have dis-
positions, in sense (1), to depress the carpet on which we walk, to at-
tract nearby mosquitoes and frighten nearby mice, and also to slip on
ice. None of these are abilities. Moreover, we have many abilities that
we have no dispositions at all, in sense (1), to realize. I have, for example,
the ability to kill cats, to stand on my head on the commuter train, and
to play bebop on the violin, but I have no sense (1) disposition to do
any of these things. At the very least, something else must be added to
a sense (1) disposition to make it into an ability. And it has to be sub-
tracted that one is or is likely to be in conditions that realize it.
Perhaps abilities are dispositions in sense (2) if we fill in the condi-
tions in the implicit antecedent of the conditionals correctly. An ability
to do A is a disposition to do A if “ what? The obvious suggestion is,
“if one tries.”
The first thing to notice about this answer is that it will require us
to unpack the notion trying, and that this is not easy. We cannot do it,
for example, by making reference to an intention to do A, if intending to
do x requires having a concept of A. That would probably leave many

animals with no abilities at all, and it would certainly leave us without
most of our cognitive abilities. Especially, it would leave us without the
ability to identify substances.We cannot suppose that in order to have a
substance concept one must first have a prior concept of the substance
one is trying to identify so that one can intend to identify it.We would
have to unpack “trying” by reference instead to some kind of purpose
more primitive than that involved in explicitly intending, in terms, for
example, of biological purpose. I will come back to this later.
But no matter how we unpack trying, an ability is not the same as a
disposition to succeed when one tries. My abilities often fail me. I have
the ability to walk, but also, under certain circumstances, the disposition
to slip or trip, and I do this exactly when trying to walk. I know how
to cook, but I may still burn the dinner tonight. Many people with the
ability to swim have drowned, presumably when they were trying to
swim. We got into the question what abilities are by noting that the
ability to reidentify a substance is fallible. Sometimes I misidentify
things, but I would not do that if it were not, in some sense, my pur-
pose to identify them. It seems that something else has to go into the
antecedent besides “if I try,” if abilities are to be understood as disposi-
tions. If I try under what conditions?
There is a standard reflex answer to this sort of question:“under nor-
mal conditions.” This combines disposition in sense (1) with disposition
in sense (2). We saw that abilities can™t just be what one has a disposi-
tion to do under the circumstances one is in or is likely to be in. Nor
are they just what one has a disposition to do when one tries. Are they,
then, what one has a disposition to do if one tries under the circum-
stances one is in or is likely to be in “ under circumstances that are
“normal”? That is, are abilities dispositions for the most part to succeed
if one tries? The reference to likelihood would explain why, for ex-
ample, although lots of people try and succeed in catching fish on hook
and line, we do not say that a person knows how to catch fish, but only
that they know how to fish. Reasonably, this is because they do not reg-
ularly succeed when they try to catch fish. It might also help to explain
how different people can have abilities to do the same thing, but some
can be better at it than others. Some are more likely to succeed when
they try than others. So let us explore this answer.
I have an ability to swim granted I am likely to succeed if I try. I am
now sitting at my desk in a dry sunny room. I now have the ability to
swim, indeed I have had that ability ever since I was six. But if I try to
swim sitting at my desk in a dry sunny room will I be likely to succeed?

Not likely. I know how to swim, but I am not in a position at the mo-
ment to swim. Indeed, most of the time I am not in a position to swim.
I go down to the lake only once in a while. So it can™t be that know-
ing how to swim is the same as being likely to succeed if I try “ what
has gone wrong?


Focus on the antecedent, “if I try.” Is the idea that if I just tried, I could
do it? “ that all that is needed for me to accomplish swimming is to add
in a trying? That, we have just seen, cannot be the right idea. Certainly
I could not immediately swim just by trying. It might be, of course, that
if I tried I could bring about a situation in which I would then be in a
position to swim. But this is not true for all abilities either. There is a
difference, a surprisingly wide one, between being able to and having
the ability to. I may know exactly how to invest a thousand to make a
million, except that I haven™t got any money at all. I have the required
ability, but I am not able to exercise it, nor am I necessarily in a posi-
tion to get myself into a position to invest a thousand to make a mil-
lion. I may know how to make a marvelous gourmet dish for which
the ingredients are completely unavailable. If there was a time when
people knew how to make tasty dodo stew, they didn™t suddenly stop
knowing how on the expiration date of the last dodo.They still had the
ability, but they were no longer able to apply it.True, in ordinary speech
the word “ability” may be a little fuzzy around this edge. But let us set-
tle on using it in this unwavering manner. Abilities don™t disappear just
because the world is uncooperative in supplying the necessary condi-
tions for their exercise.
You may have noticed that I have been slipping back and forth be-
tween “knowing how to” and “having the ability to,” whereas there are
some differences in how these terms are typically used. One difference
is exactly that “having the ability to” tends to slide a bit toward actually
being able, whereas “knowing how to” does not.Whether I am in a po-
sition to do a thing virtually never bears on whether I “know how” to
do it. Also, “knowing how” is more likely to be used for learned skills,
“having the ability” for innate skills. I know how to do sums and ride
a bicycle. I don™t know how to see. Instead, I have the ability to see.
I will continue to ignore this distinction. It won™t be pertinent for
our purposes.


Here is another way to interpret “if I tried.” Usually I will not try
something unless I believe I have a fair chance of succeeding. For ex-
ample, you could not possibly induce me to try to swim while sitting at
my desk in a dry sunlit room.You might induce me to pretend to try,
but not really to try. Perhaps what we need then is this. One™s abilities
are what one has a disposition to do if one tries under the circum-
stances one would likely be in were one to try. Let us see how far this
definition will take us.
Abilities can be more or less well developed.You can know very well
how to read or swim or ride a bicycle, or you can know these things
but not very well. I know a small child who replied when asked “Are
you learning how to play the violin?” with a denial: “No, I already
know how to play the violin; I™m learning how to play it better.” What
does learning how to do something better consist in? In the case of vi-
olin playing, of course, it consists in learning how to turn out a more
polished product. But there are two other dimensions in which abilities
can improve.
You can learn to do the same thing under more circumstances. For
example, you might learn to drive a car safely even on ice, or how to
ride a bike without slipping even over muddy paths, or how to row a
straight line even in a crosswind, or how to peel potatoes neatly with
only an ordinary knife available rather than a potato peeler. If abilities
are dispositions to do something when you try in circumstances you are
then likely to be in, then it would be wrong to think of learning these
various things as acquiring new abilities. It is the same thing you are
aiming at, whether the wind blows crosswise or from the stern, whether
the bike path turns out muddy or dry.You are trying for the same thing;
it is just that the circumstances are different.You are not aiming at the
circumstances. They were just the ones that happened to come along.
(Of course it also is possible to aim, specifically, at rowing-straight-in-a-
crosswind. You purposefully go out in a wind, and then purposefully
row across it while trying to keep a straight course. That is a different
thing to do. For example, in that case you don™t succeed if you don™t
find a crosswind.)
Notice that learning to do the same thing under a wider variety of
conditions need not produce a more reliable disposition to do that
thing when you try. So long as throughout your learning you are pretty

good at distinguishing conditions, given your current ability level, un-
der which you would succeed from conditions under which you would
fail, and so long as you don™t try unless reasonably sure you will suc-
ceed, no change in the frequency of your failures necessarily occurs. So
this might seem to be a way to get better at something, to improve an
ability, without becoming more likely to succeed when you try. On the
other hand, the harder it is to see the ground ahead with a project, the
more difficult it is to see at the start whether conditions necessary for
success lie ahead, hence whether to begin trying. Then learning how
under more conditions will increase the probability of succeeding when
you try. So far so good for our proposed definition.
This brings us to the second way in which you can improve an abil-
ity. You can learn to recognize better the circumstances in which you
will succeed if you try. Adults fall down less often than children, and
have to be rescued from drowning less often. This is not just because
they are better coordinated. They are better at recognizing risky situa-
tions, better at knowing when not to try, or not to try this way but
rather to try some other way. It is not just that adults are more cautious.
They know better when to be cautious.This way of improving an abil-
ity necessarily makes it more reliable. It improves the chances of suc-
ceeding when you try, because it improves the chances of trying only at
times when you will succeed.
Granted these ways that an ability can improve, notice that we can™t
interpret “normal conditions” for exercise of an ability as conditions
people generally are in when they try that activity. Different people may
know how to do the same thing in quite different ways, under quite
different conditions, so that the conditions under which they would try
are quite different. Perhaps neither would be able to do it under the
conditions the other finds most suitable. This sort of effect is especially
evident and especially important in the area we are most interested in.
Different people can have entirely different conceptions of the same
substance, entirely different ways of recognizing it, so that neither would
recognize it under the conditions the other would. People also can have
skills that no one else has, skills developed under conditions, perhaps, in
which no other human has been. Consider, for example, skills devel-
oped by space walkers engaged in specialized tasks. Similarly, a person
can have a concept of a substance of which no other person has a con-
cept. Any reference to “normal conditions,” if it were to help in defin-
ing abilities would have to be made not just relative to the particular
ability but also to the particular person who is able.


Is it true that my ability to do A is a disposition that I have to do A un-
der conditions I would probably be in if I tried? How is a probability
concerning only one person to be defined? In fact it is a very tough
question how to interpret a probability of this kind, and merely waving
one™s hands toward something called “individual propensities,” as some
now do in the philosophy of biology, is certainly a step in the wrong di-
rection. But I will let these issues lie. Whatever sense might be made of
the notion of an individual™s own personalized probability of being in
certain conditions when they try, still an ability can™t be defined in terms
of an individual™s probability of succeeding should they try. Here™s why.
At the start of this discussion, I argued that an ability can™t be merely
a disposition one has under conditions one is likely to be in, citing as
counterinstances those dispositions that make up one™s temperament,
one™s disposition to depress the carpet on which one stands, and to slip
on ice. A reason to reject these dispositions as abilities is that so far as I,
the organism, am concerned, these dispositions are entirely accidental,

<< . .

( 29)

. . >>

Copyright Design by: Sunlight webdesign