286 Speciestime, and species do.^13 Kitcher, on the other hand, argued that sets can, and do,
evolve.^14
When people say, as do Kitcher and Ruse,^15 that species are classes, they typically
mean that species are natural kinds in a Kripkean sense of having some intensional
definition or set of properties.^16 On this account, species are things that share some
set of criteria or microstructures that function as the essence of the species. It might
be a genome, a wild-type, or just some simple morphological criterion, but all mem-
bers of the species must have that set of properties. Hence, says Ruse, we can regen-
erate Tyrannosaurus rex because if we have the genome and a suitable egg, we have
the essence of that terrible lizard. According to Hull, though, if we have recreated
a dodo by breeding it from its ancestral pigeon stock, it is not the same species as
Dodo ineptus, destroyed by sailors on the Indian Ocean islands where it once lived.
Species, like people, can only be born once—everything else is a knockoff, like a
Tiffany-style lamp is a knockoff of an actual Tiffany lamp.^17
A problem for Hull is that species have been observed, under certain conditions,
to evolve more than once—in plants which can repeatedly duplicate their chromo-
somes to “instantly” form new species, or in fish which independently evolve deep-
water morphs in different lakes, and which will breed preferentially with similar
morphs in other lakes.^18 Let us call this the Respeciation Problem, after Turner’s
title. Moreover, the new individuals formed this way are interfertile with their prede-
cessors, and fold their genetic complement into that of the established lineages. How
can Hull deal with this?
One way might be to take the Hennigian approach and say that the two new
species are extinguished when they merge and form a new one. This is a purely
nomenclatural solution to an ontological problem. Yes, the new species has a some-
what different character (unless the modifications are infinitesimally distinct from
each other), but if, in order to save the individuality thesis one has to say that every
time a change occurs you have a new individual, the thesis is bordering on banality.
Anyway, it is not Hull’s solution; he wants to be able to restrict species in time, but
not to reject their mutability. In a discussion using the Hobbesian Ship of Theseus
example, which being rebuilt ends up entirely replaced and then duplicated from the
older parts, Ghiselin, the originator of the species as individuals thesis, proposed that
there is no fact of the matter which is the “same” ship, but that this does not mean
Theseus’ ship is a universal.^19
So perhaps here, Hull might wish to treat species as having vague boundaries in
time—there is a point x at which there is no species A, and a point y where there
is, and in between we have a vague border where it is ambiguous whether or not A
(^13) Hull 1980, Ghiselin 1997.
(^14) Kitcher 1984. Kitcher’s argument was based on the extensionality of sets—if the composition of the
set changed over time, so too did the set change (to a new species).
(^15) Ruse 1987. See also LaPorte 2004.
(^16) Mellor 1977 discusses this sense of natural kind, but see Hacking 2007b, 2007a for discussion of the
modern essentialist tradition of natural kinds, and Slater 2016 for further elaboration.
(^17) With a difference in the price tag to match, Hull 1988, 78.
(^18) Turner 2002, Martin et al. 2015.
(^19) Ghiselin 1997, 52. See also Hull’s discussion of theoretical traditions [Hull 2002].