47have names which possess certain properties, while point 5 makes a strong claim
regarding the conceptualisation and content of taxa of the same rank. As noted
earlier, Berlin also places great signifi cance on the scientifi c concept of the genus,
asserting that genera tend to be perceptually salient, and that monotypic genera , in
particular, played a central role in ethnobiological classifi cation s. This is surpris-
ing, for although the concept of rank is deeply embedded in modern systematics,
professional taxonomists would be extremely wary of claiming that members of a
particular rank were somehow equivalent across taxa. Stevens [ 142 , 146 ] has the
following to say on the topic of hollow curve s :
...for such curves to represent anything about biology, the units should be comparable; there
must be a rank of genus (for example), all members of which have the same properties.
Biological classifi cations cannot readily be compared in this way, even within fl owering plants,
even within a family. Antoine-Laurent de Jussieu, Bentham and Hooker, and Charles Bessey,
to name just three prominent botanical examples, made classifi cations that, when interpreted as
hierarchies, cannot be converted directly even to the authors' own ideas of relationships.
In any case, the Australian languages discussed above clearly violate Berlin ’s
rules, as “primary names” are used to label both plants and animals right through the
folk taxonomy. The Lardil taxon bararun ‘turtle’, for instance, which might be con-
sidered a folk generic due to its inclusion in the superordinate kendabal ‘dugong/
turtle’, contains the subordinate taxa yuburu ‘leatherback turtle’, lebulbul ‘hawks-
bill turtle’ and dunkumudin ‘greenback turtle’, which bear no resemblance to each
other, or to their category label bararun. An analysis of the familiar English folk
taxon ‘dog’ is also illuminating in this respect. In Berlin’s scheme, labels such as
‘dog’, ‘cat’ and ‘horse’ would be classed as folk generics , by virtue of their primary
names, and the fact that they label perceptually salient ‘natural kinds’. While the
subordinate groups ‘sheep dog’, and ‘bulldog’ behave like good subgeneric (spe-
cifi c) taxa, in that they possess secondary names which make reference to the super-
ordinate taxon, the vast majority of dog breeds are labelled by primary names, such
as ‘terrier’, ‘beagle’, ‘mastiff’ and ‘boxer’. Of course, some of these specifi c taxa
can be further subdivided into ‘varietal’ taxa, such as ‘fox terrier’, but the specifi c
taxa themselves are mostly primary names. Berlin’s Point 11 makes provision for
culturally signifi cant folk species-level organisms to be named by primary names,
but is a ‘great Dane’ (complex primary name) really more culturally signifi cant than
a ‘sheep dog’ (secondary name)? Furthermore, the only way to resolve the even
more serious objections, that can be raised when comparing taxa of the same rank
from widely separated groups, is to put forward explanations that are so ad hoc as
to be completely meaningless. Could one ever convincingly argue that ‘rainbow
trout’, ‘Dover sole’ and ‘red mullet’ have had secondary specifi c names thrust upon
them because they are somehow less culturally salient than ‘beagle’? Or that a
‘Rhode Island Red’, with its complex primary name is more highly valued, cultur-
ally speaking, than a ‘fox terrier’?
In the following section, I attempt to demonstrate that at a very general level,
Solega shows signifi cant departures from the predictions made by Berlin ’s model.
More specifi c instances of departures are presented in Chaps. 3 and 4.
2.6 Folk Genera, Rank and Nomenclature