Philosophy of Science, 71 (December 2004) pp. 892–900. 0031-8248/2004/7105-0020$10.00
Copyright 2004 by the Philosophy of Science Association. All rights reserved.

892

Human Kinds and Biological Kinds:
Some Similarities and Differences

John Dupré†

This paper compares human diversity with biological diversity generally. Drawing on
the pluralistic perspective on biological species defended in earlier work (2002, chs. 3
and 4), I argue that there are useful parallels to be drawn between human and animal
kinds, as there are between their respective sources in cultural evolution and evolution
generally. This view is developed in opposition to the insistence by sociobiologists and
their successors on minimizing the significance of culture. The paper concludes with a
discussion of the relation between cultural difference and individual difference, and
the relation of the latter to conceptions of human freedom.

Hard-headed sociobiologists are inclined to insist that humans are just
one species like any other, and just like any other species, they have a set
of genetically determined dispositions to behavior. It will, of course, be
admitted that the complexity of such dispositions will vary from species
to species. The dispositions of a dolphin, say, are far more subtly attuned
to the environment in which the animal finds itself, than are those of a
flatworm. The dispositions of the dolphin will surely include details that
are sensitive to the particularities of individual development. Years ago,
Konrad Lorenz observed that the behavior of baby geese following their
parents depended on the geese being exposed at birth to the right animal.
If they were exposed to a middle-aged bearded ethologist rather than a
mother goose, they would follow him as happily as they would a more
appropriate leader. The sensitivity to upbringing of a dolphin or a chim-
panzee is presumably a good deal greater. Still, all of this may be conceived
of as happening within an extremely complex set of genetically determined
parameters that dispose the creature to one kind of behavior rather than
another. A proper account of such parameters for ants, elephants, or

†To contact the author, please write to: ESRC Centre for Genomics in Society, De-
partment of Sociology and Philosophy, Amory Building, University of Exeter, EX4
4RJ, UK; e-mail: J.A.Dupre@exeter.ac.uk.



COMPARISON OF HUMAN AND BIOLOGICAL KINDS 893

humans gives us an account of ant nature, elephantine nature, or human
nature.

Less hard-headed (perhaps) opponents insist that there is a more fun-
damental difference than this suggests, between human nature and the
natures even of our most sophisticated nonhuman relatives. Often this
intuition is articulated in terms of the importance of culture. Culture is
important because there are many different cultures. And if this is correct,
then the generic identification of an organism as possessing human nature
is only the beginning of an understanding of what kind of thing it is. We
shall also want to know the culture, or cultures, to which it belongs.

The hard-headed (let us, for convenience, call them sociobiologists) are
scornful of this move. Culture, they retort, is merely an evolved device,
and would surely not have evolved if it did not conduce to the fitness of
the individuals that took part in it; and it would hardly do this if it were
not quite tightly constrained by the biological natures of individuals. At
this point, the culturalists (as we may call the other participants in the
debate, among whom I admit to counting myself) reply that this is un-
warranted adaptationism—Panglossianism even—and the empirical fact
is that culture has produced all kinds of behavior—celibate monasticism,
suicide bombing, and so on—that is plainly not adaptive in a biological
sense. Biology does not determine culture any more than (as I shall men-
tion later) culture determines the behavior of individuals.

Another obvious reason for concluding that biology does not determine
culture is simply to observe that there are many different cultures. Here
we encounter a characteristically frustrating aspect of the debate, for the
sociobiologists, while not exactly denying this, insist that cultures are not
nearly as different as the culturalists pretend. Much cited, for example,
is Donald Brown’s list of several hundred human universals, said to be
found in every culture (Brown 1991; an updated list is offered in Pinker
2002). This list, and its underlying presumptions, could be debated in-
terminably. Some items on the list are rather unsurprising—false beliefs,
for instance. Classification, and a dozen particular kinds of things clas-
sified, seem to add nothing to the fact that humans have languages, some-
thing that nobody, I suppose, would deny was a biological fact about us.
Speaking as an Englishman, it is hard to imagine that any speaking crea-
tures could fail to classify kinds of weather. At the other extreme there
are some highly dubious universals. There is a reasonable doubt, I should
think, whether any cultures exhibit the Oedipus complex, though perhaps
if any do, all do. But the important point is that it is not even remotely
surprising that such a list can be produced. We are—and this is not in
dispute—all human, which is to say, all part of a single biological species,
and it is to be expected that we will have very large numbers of features
in common. In just the same way, it would be possible to list a very large



894 JOHN DUPRÉ

number of features common to all mammals (some of these would appear
on Brown’s list). This hardly shows that there are no distinct species of
mammals. So my point is that whether or not one accepts all or some of
the cultural universals on the list has little to do with whether there may
be legitimate kinds to be distinguished within the human species. If there
are such kinds, this fact would point to one way in which a full biological
account of what it is to be human fails to exhaust the classificatory moves
relevant to understanding the nature of a particular human, and thus
points to a respect in which the sociobiologists’ claim misses something
that the culturalist can grasp.

The culturalist objection to the picture sketched in my opening para-
graph, then, might be that the human species was in some respects more
like a genus encompassing a range of distinct cultural species. While being
aware that there are obvious dangers in pushing this analogy too far—
there are political advantages in stressing the unity of the human species—
there are respects in which this analogy seems to me to be a useful one,
not only theoretically, but also politically. There are also dangers in ex-
aggerating the homogeneity of the human species, namely, that such an
exaggeration gives license to the assumption (by powerful groups that
have the time to reflect on human universals) that the characteristics they
see around them are universal characteristics of humankind. Just as it is
widely agreed that biological diversity is a thing of great value that should
be preserved as far as possible, we should consider the possibility that
cultural diversity may have comparable value. In fact, I think the parallel
here is in many respects much closer than is generally supposed.

At any rate, my aim in the present paper is to explore the theoretical
value of the analogy between biological diversity and human cultural
diversity and to suggest some important similarities as well as some im-
portant differences. The place to begin this comparison is with some
variably controversial views on the nature of biological kinds (the views
summarized here I defend in greater detail in Dupré 2002, chs. 3 and 4).

The facts of biological diversity do not determine, though they very
substantially constrain, the possible modes of classification. (By possible,
here, I also mean sensible.) There are, of course, more or less controversial
ways of expanding on this thesis. Almost entirely without controversy is
the notion that different systems of classifications can draw boundaries
in somewhat different places. In many cases, for example in the classifi-
cation of many kinds of flowering plants, there is a fairly smooth con-
tinuum of intermediate forms; if classification succeeds in assigning every
organism to a kind, it can only do so with a degree of arbitrariness. But
of course, only a degree of arbitrariness. No one will be tempted to classify
aardvarks in the same species with zebras, or even roses with lilies.

A more controversial thesis is that a unique principle suitable for clas-



COMPARISON OF HUMAN AND BIOLOGICAL KINDS 895

sifying all organisms does not exist. The dominant tendency in contem-
porary systematics is to suppose that taxonomy should reflect phylogeny,
or evolutionary history. Given the common assumption that evolutionary
history can be depicted as a diverging tree, mapping relations of descent,
then it is suggested that every named taxon should represent an entire
branch (or twig) of this phylogenetic tree. There are several difficulties
with this proposal. One fundamental one is that it assumes that phylogeny
can be represented by an only divergent, never convergent, tree. Conver-
gence makes it ambiguous as to what should count as a branch. But it
is increasingly clear that convergence is common. It need not consist of
the merging of whole species—though that can happen—but merely re-
flects the fact that genetic material appears often to be transferred between
taxa that may be at some phylogenetic distance one from another. The
ancestors of a particular taxon may then be a motley crew. In the case
of bacteria, in particular, the phylogenetic map seems likely to be highly
reticulated, and phylogeny may give little guidance for classification. In
practice, bacteria are classified on the basis of various morphological
criteria, often pragmatically chosen on the basis of their relevance (path-
ological, agricultural, ecological, etc.) to human interests. This seems gen-
erally a good practice.

Despite this somewhat skeptical perspective, it is important to remem-
ber that there are also many species that show a striking coherence and
persistence of form. One thing that contributes to this coherence is the
ability of species to preserve reproductive isolation from other species,
and there are a range of subtle mechanisms by which members of a species
distinguish suitable sexual partners—members of the same species—
thereby ensuring reproductive closure. These two observations ground
familiar species concepts: the biological species concept, which until re-
cently dominated systematic theory, and the mate recognition species con-
cept. But it seems likely that there are other mechanisms—ecological pres-
sures, developmental homeostasis, for instance—by which species
maintain coherence, and this has led to the more catholic coherence species
concept, based on the idea that species should be defined by whatever
factors maintain their coherence. The full path to pluralism requires the
recognition only that there are many areas of the biosphere in which
species are not that coherent, and in which only purely morphological
criteria are likely to be useful.

We have a general interest in classifying organisms, in being able to
ask and answer the question of any biological individual, What kind does
this belong to? The reliable availability of classifications is a prerequisite
for the possibility of storing and retrieving information about the char-
acteristics of particular kinds of organisms. And this general interest un-
derlies the possibility that addressing the need to classify organisms may



896 JOHN DUPRÉ

have to draw on different resources in different cases. The current pop-
ularity of pluralism about species derives from the fact that the concept
of a species answers to two only partially reconcilable demands. On the
one hand, we have evolutionary theory, which is inclined to assume that
a proper answer to the question, “What is a species?,” will provide us
with the basic unit of evolutionary process. And on the other hand, we
have the general classificatory interests. Experience suggests that any an-
swer to the theoretical question will be applicable only to some subset of
the actual biological diversity we discover. Consequently, the classificatory
interest forces us to accept a variety of theoretical conceptions of what
it is to be a species.

Let me now turn to human classifications, beginning with what I think
is the most important difference from the general biological case. We have
no general interest in classifying people. The question, “What kind of
human is that?,” without some much more specific context, is senseless.
We classify people with regard to nation, gender, occupation, and some-
times race, but we don’t just classify them, period. There is, therefore, no
concept analogous to the species which answers to a general interest in
classification. And this, finally, makes pluralism much less problematic.
Our general interest in classification causes some discomfort with the idea
that an organism might belong to two non-coextensive kinds, both of
which lay claim to the concept of a species. We are obliged to choose one
or other kind in the interest of classificatory uniqueness, but this will have
the consequence that the same theoretical concept, when applied to one
kind of organism will be a species, and when applied to another will not.
This kind of problem has led me to argue that we should sharply differ-
entiate biological classification from theory, and reserve the term species
for its original use as a classificatory concept (Dupré 2002, ch. 4). But no
such problems arise for humans because there is no concept applying to
human kinds in the way species applies generally to biological kinds.
Perhaps it was once thought that race had such a role, but this is certainly
not so today. Without such a connection there is, of course, no problem
in acknowledging that an individual is an Estonian, a male to female
transsexual, and a dealer in exotic reptiles, without supposing that these
categories are coextensive.

So much for the disanalogy, but what about the positive analogy? I
suggest that, like biological kinds, human kinds often have some degree
of coherence; that this coherence is maintained in ways comparable to
the methods by which species coherence is maintained; and that, like
human kinds, biological kinds must often be understood historically. Like
biological evolution, cultural evolution often flows down more or less
well-defined channels.

What is cultural evolution? For the purposes of the present argument



COMPARISON OF HUMAN AND BIOLOGICAL KINDS 897

it is best to be as noncommittal as possible. Culture consists of the phe-
nomena studied by social scientists (and, in certain important respects,
humanists) and its evolution is the process studied by history. I don’t want
to make any assumption about how strong an analogy can be drawn
between natural selection and processes of cultural evolution, and I most
certainly don’t want to make any commitment about the ontology of
cultural evolutionary processes as, for instance, is provided by the so-
called science of memetics. All I need to insist on, rather noncontrover-
sially I take it, is that innovations occur within culture, and innovations
spread unevenly through human populations. New ideas and techniques
in rap music, for instance, tend to occur among different groups of people,
and affect the future behavior of different groups of people, from those
affected by changes in the practice of military game theory. Society is
segmented in multitudinous ways that determine the probability that cer-
tain kinds of cultural innovation will be transmitted.

There are two importantly different kinds of cases, and, no doubt, a
spectrum of cases in between. At one extreme, a part of the human species
will form a cultural kind most closely analogous to a biological species
if it is very strongly isolated from the rest of humankind. This perhaps
still applies to some “primitive” cultures and has been attempted by some
totalitarian cultures. In the reasonably open societies in which most of
us live, on the other hand, there is rather a very large number of mi-
crospecies, and any individual will typically belong to many of these. The
fairly banal consequence is just that if you want to predict the behavioral
phenotype of a human it won’t get you very far to know just that they
are human. As you discover more of the microspecies to which the in-
dividual belongs—Chinese, stamp collector, philosophy professor, and so
on—you can make increasingly reliable predictions of behavioral dispo-
sition. (I won’t address here the very interesting case—perhaps the most
interesting case—of gender classifications (but see Dupré 2002, chs. 8 and
9). Here it is hotly disputed how much there are biologically grounded
differences in behavioral dispositions, though it is surely noncontroversial
that there are also culturally grounded differences.) The reason it is pos-
sible to make such predictions is that the various cultural streams distin-
guished by these classifications flow through, so to speak, this particular
individual. And though something like this may be true to a very limited
sense in other species that have rudimentary culture, in no species is it
true to anything like the extent to which it is in our own. And to add
one more banality, surely a large part of the basis for this unique feature
of our species is the complexity of the languages that enable us to transmit
complex cultures.

I have said that much of this is banal. Its significance is also widely
minimized. To bring back the combatants with whom this talk began, the



898 JOHN DUPRÉ

sociobiologist will of course admit that there are culturally transmitted
differences in behavior, but he or she will insist that they are superficial.
What may initially strike us as fundamentally different cultures will turn
out, on closer inspection, to be no more than slight variants on the cultural
universals mentioned earlier. The culturalist, on the other hand, will insist
that the specificities of culture are fundamental to understanding human
life and, insofar as the concept is even allowed, human nature. In con-
trasting these views it is necessary to bring in one further ingredient.
Sociobiologists deny that there are important differences between mem-
bers of different cultures per se, but they do not conclude that there are
no significant differences between any humans. There is also individual
variation within cultural groups (if any). It is not uncommon for the
defenders of the more fundamental importance of biology than culture
also to be very enthusiastic about the importance of individual difference.
(I can’t help thinking here of Margaret “there is no society” Thatcher. A
slightly more intellectual version can be found in Stephen Pinker’s The
Blank Slate (2002).) Of course the culturalist does not deny that there is
individual difference. But what I do want to argue, and this will be my
final point in this paper, is that the culturalist has much more adequate
resources for understanding these individual differences. And this I offer
as one further compelling reason to favor the perspective of culturalism.

What, first of all, can the sociobiologist say about individual difference?
On good days, anyhow, we can all agree that different environments will
affect the development even of identical genotypes. So one explanation
of individual difference is just environmental luck. Then there is genetic
luck. Everyone agrees on any day that there is genetic variation between
individuals, and if one believes that genes determine human nature, then
it is to be expected that difference in genes will determine differences in
nature—though perhaps only small differences if the unfolding of the
human essence is not to be drastically derailed. So there is environmental
luck and genetic luck. On reflection, this analysis may seem rather dis-
couraging, but it should hardly be surprising that it is a widely held view.
I have said myself that we should see the genome and the environment
as interacting in the production of an individual, but in that case, good
or bad outcomes in human development are presumably to be attributed
to the genome, the environment, or a bit of both. Since people never
choose their genomes and only choose their environments when it is largely
too late, these determinants of good and bad outcomes can only be seen
as good or bad luck.

Philosophers will notice that at this point we have run straight into a
familiar aspect of the free-will problem, and it is likely to seem unim-
portant from the point of view of this problem whether we are inclined
to emphasize genetic or environmental causes of human behavior, or a



COMPARISON OF HUMAN AND BIOLOGICAL KINDS 899

mixture of both. Although I cannot fully elaborate the idea here, I suggest
that the interaction between the biological and the cultural is, in fact,
exactly the place to look for a better explication of free will. (The following
ideas are developed further in Dupré 2001, ch. 11.) As Kant explained,
there is a crucial difference between action in response to a desire and
action according to a principle. Although it is important, now following
Hume, to distinguish a situation where we are free to act in accordance
with our desires from those where, for reasons of constraint, incapacity,
and so on, we are not, action following desire seems in no important way
distinguished from the broader causal category that includes the instinctive
behavior of lower animals and even the mechanical behavior of inanimate
objects. Kant observed that accordance with a principle, especially if some
objective standing can be granted to that principle, offers a different kind
of grounding to action from this quasi-mechanical embedding in our par-
ticular desires.

A reductionist cast of mind, wedded to bottom-up explanation, will
immediately respond that it makes no difference whether what is encoded
in the individual brain is a desire or the commitment to a principle. I may
refuse the glass of wine because I desire to avoid the pain of a hangover
or because I have signed a temperance declaration; either way, something
in my brain causes me to refuse the drink. Even if, for the sake of ar-
gument, I accept this much of the reductionist’s position, there is a per-
spective from which these cases are quite different. A principle is a social
object. Even a secretly embraced principle solely for my own conduct is—
arguably, and I would argue this—something that is only possible through
the medium of language. But leaving this contentious issue, most prin-
ciples are public, and people come to embrace them as a consequence of
public exposure to them. Thus, principles of conduct are devices whereby
what exists socially can determine how individuals act. A widely accepted
culture of politeness, for instance, will result in countless individual acts
of politeness.

If we now consider differences between people we can see another way
of distinguishing between two sources of these. There are the differences
in taste so beloved of economists: you like snails, I prefer oysters. But
there are also differences of principle. The pursuit of world peace or the
preparation for the second coming structures my actions in systematic
ways that make me affect the world in distinctive and quite different ways.
Hence the existence of social structures imbues me with causal powers
that I would not possess without the appropriate relationship to those
structures. And finally, an individual, in the course of implementing or
debating the principles on which she acts, may thereby bring about
changes to these social structures, and thereby to the powers of herself
and other individuals provided by those structures. At the risk of ro-



900 JOHN DUPRÉ

manticizing intellectual work, this is perhaps the purest from of auton-
omous individual action.

It is characteristic of the free-will problem that it is always possible to
push the hard questions back, so that part of the art of addressing the
problem must always be knowing where to stop. Hence, someone will
inevitably ask, “What determines the principles I decide to embrace?”
And if that is not autonomously decided, then how can my actions, guided
by principle, be held to be truly autonomous? The Kantian answer, that
I am autonomous if I choose the objectively correct principles, is enor-
mously appealing but, I fear, untenable. I think it is more promising to
offer an answer to the question of where to stop, namely, this: we cannot
decide who to be. I do not choose my genes and I do not choose my
environment. Hence I do not ultimately choose who I am. But being who
I am makes a particular kind of difference to the world I live in, and a
difference that is teleologically structured by features of the society in
which I exist. This is already something quite different from what is pos-
sible for a creature that lacks a society with any comparable structure.
When, finally, we notice that acting in the ways characteristic for the
particular individual I am may include influencing in some ways the con-
ceptual structures that make me and others what we are, we have, I think,
as much free will as is worth wanting.

Thus, finally, back to human kinds. Human kinds, I have said, reflect
the streams of cultural evolution and create human diversity. But these
streams of culture also provide the most interesting grounds for human
action and human freedom. The human kinds to which people belong
determine, in essential part, what they can do—not just in the sense of
emitting behavior but in the sense of acting in the world and making a
difference to it.

REFERENCES

Brown, Donald E. (1991), Human Universals. New York: McGraw-Hill.
Dupré, John (2001), Human Nature and the Limits of Science. Oxford: Oxford University

Press.
——— (2002), Humans and Other Animals. Oxford: Oxford University Press.
Pinker, Stephen (2002), The Blank Slate: The Modern Denial of Human Nature. New York:

Viking.