316 SpeciesQuine is wondering how words in one language or idiolect can be translated
into another and imagines a field anthropologist trying to work out what the locals
mean by their words purely empirically, in terms of the stimuli that elicit the verbal
responses. A rabbit scurries by and the local says, “Gavagai!” What does “gavagai”
mean? According to Quine it might mean “rabbit,” “rabbit part,” “extended rabbit,”
and so forth. Since the stimulus in each instance of the word so far encountered is to
the anthropologist’s eyes identical, how could we ever know?
The idea that species names are anchored in the biological world in terms of the
type specimen, or the initial experience of naturalists and taxonomists, is commonly
held, at least implicitly, and goes by the name the “baptismal” account. Why else
do specialists apply the name to the holotype, and go to pains to assign a replace-
ment specimen (lectotype) if the holotype is lost?^95 The problem with this is that the
causal account underdetermines the extension of the term. Devitt and Sterelny call
this the Qua Problem.^96 How, they ask, do we know what it is that we are referring to
when we name a general kind term like Tiger? Is it the striped fur (if so, what about
albinos), or the teeth (then what about toothless old and circus tigers?), and so forth.
What is it that makes us refer to the species Leo tigris qua tiger?
The Qua Problem is known to taxonomists. There are numerous cases in which
the extension of a named taxon, particularly species but also genera, turns out to
be other than the biologist expected, sometimes requiring that the species name be
subsumed to a prior designation, sometimes requiring that the taxon be divided into
separate taxa. There are cases where taxon names have been given independently to
male and female forms in birds, and where the females of a genus are so similar that
they were put into one taxon while the gaudy males, each individually distinguish-
able by humans, have been placed into different taxa, and the mistake not realized
until the two morphs were observed actually mating.^97
As a response to essentialism, the baptismal account has something to recom-
mend it; general kind terms do not have to have definienda, since the commonality is
provided by the causal relations between the initial specimen and the other members
of the kind. However, it does not manage to release us from the problems of knowing
that other specimens are in the same kind.
In order to resolve the Qua Problem, we would have to range too far from the
present subject, and so I shall merely note here that this problem occurs for any
and all general terms, whether they are terms that cover only an extensional set, a
causal lineage, or an abstract class. The more interesting problem in biology is how
we know that a holotype “refers” to all other members of the species.^98 As always,
there is the issue of what is, and the issue of what we can know about it. I take it that
the idea of species as interlocking lineages (the Hennigian notion of tokogenetic, or
(^95) Levine 2001.
(^96) Devitt and Sterelny 1987.
(^97) For example, the male superb warbler (Malurus cyaneus, now called the superb fairywren) was
described in The voyage of Governor Phillip to Botany Bay [Phillip et al. 1789], and the female then
given its own entry after the initial description of the male had been reported, based on later studies:
(^) Except from the size and shape, this bird would not be suspected at first sight to belong to the
same species as the male: the epithet of superb applies very ill to the female.
(^98) Levine 2001, LaPorte 2003, but see Colless 2006.