concepts

To put it simply... Basic

Massimo Piattelli-Palmarini
too specific. Lunch is a basic concept, but so
concepts
are bread, spoon and banana, all possible
ry a simple experiment. Stand beside parts of that lunch. The desire to decompose

T a window, take a look at the scene

outside and then sit down and quickly
write a half-page report of what you have
such concepts into real basics has proved to
be almost irresistible. A famous example is
the (alleged) decomposition of ‘kill’ as ‘to
Why has cognitive science struggled
to find an explanation for the terms
that we use every day?
seen. It is a safe bet that you will mention cause to become not alive’. It is claimed that
people walking, cars, buses, streets and in our mental lexicon, kill is just shorthand example, as no one would introduce a child
buildings, or (depending on the chosen for the latter composite expression. But to the concept ‘bird’ by displaying a penguin,
window) grass, trees, hills, rivers and birds consider situations in which the cause is or to the concept ‘tree’by displaying a bonsai.
flying. It’s extremely unlikely the report separated from its outcome by a long and/or There is every reason to suppose that the
will mention blue sport utility vehicles, anomalous chain of intervening events, human mind is natively equipped with the
belted kingfishers, or instances of strutting, for example,the suicide of a government capacity to lock onto the salient property
ambling or swaggering — even if your auto- advisor after his involvement in a media row after an encounter, as long as it is supplied
motive, avian or human-locomotion exper- had been made public the previous week. with the correct mode of presentation and
tise allows you to identify things and events at It is clear that we will consider this an the appropriate situation. The capacity to
that level of detail. At the other extreme, it did instance of ‘causing to become not alive’, generalize instantly and competently from
not cross your mind to write expressions but not an instance of ‘killing’. Therefore, it a good example, while retaining the corre-
such as “intentionally cause their body to seems that if any such decomposition is sponding verbal label, remains awesome,
move horizontally”, or “self-propelled wheeled to hold water, we need an additional compo- and at present mainly mysterious.
vehicle”, which some philosophers might use. nent X, so that: kill4CAUSE&BECOME It is tempting to make direct connections
Interestingly, your report will contain only NOT ALIVE&X. That would be acceptable between the meaning of a concept and the
terms that are the first to be learned by a if X were both general (in the same league most obvious inferences that the person
child, are usually expressed by a single word as CAUSE), and sufficient. But it turns who possesses it is disposed to make — for
in most languages, are remembered best out that X needs to be as specific as ‘kill’. example, the inference going from ‘bird’ to
and are preferentially used when we ‘talk to So there is no gain in understanding — ‘animal’ and from ‘water’ to ‘wet’. This move
ourselves’. They are, in the cognitive-sciences and therefore no explanatory use — for goes under the name of inferential role
jargon,‘basic concepts’. any such decomposition. semantics. The counter to this move is that
Ideally, in a theory of concepts and in It has been widely assumed in the philoso- the content of a concept cannot ultimately
lexical semantics, the term ‘basic’ ought to phy of mind,psychology and lexical semantics consist of any kind of readiness to do some-
cover the topmost level of abstraction that basic concepts must be homogeneous thing, not even a ‘disposition’ to draw infer-
(something like the undifferentiated essen- under some interesting description or other. ences. It is rather a mental particular that
tial furniture of the world). Or, at the other For example, they may have a characteristic applies to things, notably to the standard
extreme — in a tradition that extends back role to play in concept acquisition, percep- instances of the category for which the
to the philosopher David Hume — they have tion, memory or thought. However, there concept stands. But even failure to identify
roots in our most direct access to unadorned is no serious evidence for any such claim, marginal exemplars (is a whale a fish? Is
sense impressions (a green splotch here and and when combined with a thorough mercury a metal?) does not count as failure
now in front of me). Unfortunately, ‘basic account of concept possession, this discom- to possess that concept fully.
concepts’ sit comfortably at an intermediate forting fact appears to lead to implications As much of our mental life consists of
level — they are neither too general nor that are utterly implausible. For instance, it applying concepts to things, it may come as
suggests that similarities among things a disappointment that the most plausible
A. BELOV, SONOMA MORNING/CORBIS

imply similarities among concepts of those attempts at an explanation (decomposition,

things. A rope may be similar to a snake inferential power, strategies of verification)
because of their shape, but our concept of a did not work. Halfway between an avowal
rope is not at all similar to our concept of a of impotence and an incitement to do better,
snake. A leading thinker in this domain,Jerry the theory of concepts seems to be a
Fodor, concludes that basic concepts have signal place where, in spite of a voluminous
nothing interesting in common except their literature, cognitive science “went wrong”—
basicness — basic concepts are boring. But, at least until now. ■
he hastens to add, the fact that they are isn’t! Massimo Piattelli-Palmarini is in the Department
Fodor champions a different tack of Linguistics, University of Arizona,
altogether, treating basic concepts as ‘atoms’ 1103 East University Boulevard, Tucson,
that cannot be decomposed, expressing Arizona 85721-0025, USA.
exactly the property that they express (for
example, the property of killing) — no less FURTHER READING
and no more. There are two core compo- Fodor, J. A. Concepts: Where Cognitive Science Went
nents of this atomistic approach: first, a Wrong (Oxford Univ. Press, Oxford, 1998).
causal link between our mind and the Margolis, E. & Laurence, S. (eds) Concepts:
property being exemplified, or evoked in Core Readings (MIT Press, Cambridge,
discourse; and second, some efficient way of Massachusetts, 1999).
Taking in the view: we tend to use simple
presenting a good example of that very prop- Murphy, G. L. The Big Book of Concepts (MIT Press,
language when asked to describe a scene.
erty. Better still would be a prototypical Cambridge, Massachusetts, 2002).

NATURE | VOL 426 | 11 DECEMBER 2003 | www.nature.com/nature © 2003 Nature Publishing Group 607