Two apparently distinct ontological distinctions: vs

"After the development of set theory, however, a distinction of the
scholastics between intension, of sets that were circumscribed by
definitions, and extension, by member inclusion, was revived, and the
logical tradition of species was held to be a matter of intensional
definition. In a seminal summary of the traditional pre-set theoretic
logic of diairesis, or division, from the most general to the most
specific, H. W. Joseph (1916) made a clear distinction, as Whately had
90 years earlier in 1826 (see below), between logical species and
“natural” species, but the developers of the essentialism story failed
to pick this up, and read him as saying that species of living things
were the same as the logical species (as discussed in Chung 2003,
Winsor 2001, 2003, 2006a)."

In a discussion of what counted as a kind in natural history, William
Whewell in his Philosophy of the Inductive Sciences (1840) gave a good
account of natural kinds as being types from which there were
deviations, although he treated species themselves as simple objects.
John Stuart Mill disagreed and asserted that natural kinds were
definable and had universal (causal) properties for all members (Mill
2006).1 For Whewell, the type of the taxonomist was a natural kind;
for Mill, it was the element and the compound of chemistry and
physics. Still, I believe the weight of venerable history is on
Whewell’s side, not Mill’s, or to put it another way, that Mill’s
conception of natural kinds is not something that applies well to
historical sciences that are restricted to specific domains, like
natural history or taxonomy. And despite what we might think based on
the discussion of logic from Frege onwards, as late as the early 20th
century, for instance with John Venn (1866) and others, a natural kind
was indeed typically thought to be a kind of living beings, caused by
generation (Hacking 1991).

So much of the confusion about essences can be resolved if we do not
adopt the view that Mill introduced, that a real Kind must have a set
of necessary and sufficient properties. For Mill, a species would be a
natural kind (a phrase introduced by Venn, although he did not adopt
the Millian view regarding it; Mill just used the word Kind) if it had
some set of universally shared properties that made each organism a
member of it, rather like having a certain number of electrons,
positrons and neutrons makes each atom of an element that element. By
contrast, for Whewell, and for those taxonomists who he was accurately
describing in the Philosophy of the Inductive Sciences, to be a member
of some group, a species, a genus, and so on, is to be mostly like the
typical form, and to be something that can be supposed to share a
genealogy with that form. This sense of “essence” is something that, I
believe, is quite consistent with our present understanding of
genetics and populations; most members of most species share most
genes, but there are multiple genetic controls over some typical
traits, and some species have major genomic varieties. David Hull has
said that there is nothing so unusual or absurd in biology that some
species doesn’t have it somewhere or somewhen; I call this Hull’s
Rule. To be an essentialist in the world of Hull’s Rule means that you
cannot insist that taxa are going to always have some set of genetic
or other causes, but it doesn’t mean that you can’t say that taxa
mostly share causes. To be a taxon, I think, is to have some set of
general properties, the bulk of which any member will share.2 This is
sometimes called in philosophy the “family resemblance predicate”,
after Ludwig Wittgenstein’s example in the Philosophical
Investigations (Pigliucci 2003, Wittgenstein 1968), but over a century
before Wittgenstein, Whewell made just this case. Families resemble
each other because they share generative histories and hence
generative causes, but they share them typically.

The thesis known as the Individuality Thesis (Gayon 1996, Ghiselin
1997, Hull 1978), in which species are considered to be not kinds, but
named objects that have a historical location, is a defense of
biology, and especially genetics, against the encroachment of Mill’s
notion of a Natural Kind. Sure, say the individualists, species and
other biological taxa are not Natural Kinds (as defined by Mill). The
only other metaphysical notion open to philosophers of taxonomy is
that of an Individual, a thing that exists in one time and one place
or region, and has a start and an ending. Hence, species are
Individuals. I cannot fault this logic – species clearly aren’t the
kinds of Kinds that Mill required, and they actually are historical
objects, so I have no objection to their being called Individuals; but
I do think they have “essences”, or, rather, typical developmental
systems and responses to typical environments; to preclude confusion,
let us call these developmental types.

There is a recondite argument in metaphysics as to whether individuals
in the metaphysical sense can have an essence – some say not, others
say they can. I won’t rehearse it here, as it is of interest only to a
certain kind of philosopher. However, there’s a much older sense,
Aristotle’s, that I think works well for individuals. Aristotle did
not have a technical term for essence, especially not for natural
objects, but instead he used the phrase “the-what-it-is-to-be”3. We
can usefully ask what it is to be a member of a species. Generally, it
is to share a genetic, developmental and ecological process. To be a
member of Homo sapiens is to have some majority of typical properties:
a certain number of chromosomes, genes, cell types, and ecological
resources that reliably result, most of the time, in the morphology of
the typical members of the species (Pigliucci 2003, Pigliucci and
Kaplan 2006). We can investigate this and list these properties
(although we must be careful not to call the first genome, such as
Craig Ventner’s, the “typical” genome); this is a matter of empirical
research. It is not a matter of definitions, it is a matter of finding
out what the developmental types are by sampling populations and

No philosopher of language who is inclined to the Millian kind of
Kinds or to linguistic or logical essentialism is going to be happy
with this, just as family resemblance predicates haven’t been greatly
liked by such philosophers in general (but see Gasking 1960), but it
is a sense of the essence of species that is at once compatible with
the biology, and consonant with the original history of the notion of
essences. Species can be metaphysical Individuals, and yet have
something mostly shared between its members, which can be granted the
honorific of “essence”, the what-it-is-to-be a member of that species,
its developmental type."

— John S. Wilkins, "What is a species? Essences and generation"

You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
To post to this group, send email to
Visit this group at
For more options, visit

Reply via email to