528 
Telopea Vol. 6(4): 1996 
From the 20:20 perspective of hindsight, how do our views look today? On which 
issues did we turn out to be right? On which issues do we now seem to have missed 
the mark? More importantly, what effects did these two periods of heightened activity 
in taxonomic philosophy and methodology have on systematics? In the following 
pages 1 deal with Johnson's early evaluations of numerical taxonomy as well as his 
later criticisms of cladistics. 
What's in a name? 
Robert Sokal, Peter Sneath, Paul Ehrlich and others decided in the late 1950s that 
systematics, as it had been practiced for generations, was too subjective, impressionistic 
and downright messy. They decided that taxonomic method had to be made more 
explicit, quantitative, objective and repeatable. In short, they wanted to eliminate what 
Simpson (1961) termed the 'art' in systematics. The founders of this school of systematics 
preferred to call themselves 'numerical taxonomists,' to emphasize the increased role 
that they saw for computers and various mathematical techniques in systematics, but 
they also formulated a general theory about proper scientific method, the sort of 
methodology that they thought necessary if systematics was to become quantitative. 
In particular they repeatedly warned about allowing a priori speculation to enter into 
the early stages of classification, and among the most dangerous sorts of a priori 
speculations were those that concerned the evolutionary process or possible 
phylogenetic relationships. Systematists should limit themselves to observable features 
of organisms, at least in the early stages of classification. As a result of their emphasis 
on phenotypic characters and their anti-theoretic stance, their opponents termed them 
'pheneticists' or more accurately 'numerical pheneticists.' 
As my use of quotation marks in the preceding discussion indicates, I have some 
doubt as to exactly what the school of systematics initiated by Sokal and Sneath 
should be termed — numerical taxonomy, phenetic taxonomy, or numerical phenetics. 
In the past decade or so, several schools of thought have arisen in the humanities in 
which concern with language swamps any interest in the non-linguistic world. For 
these folks, it seems that the term used to refer to the AIDS virus is vastly more 
important than the development of a vaccine or treatment for this terrible disease. 
These 'postmodernists' seem to think that we can chat our way to solutions to the 
world's problems. Once we rework language to eliminate sexist, racist, homophobic, 
etc. connotations and implications, all will be well. Socially sensitive and politically 
correct language will rule the day. 
Numerical taxonomy and phenetics 
As can easily be inferred from my characterization of these various schools of thought 
in the humanities, I am not especially taken with them. Even so, need I convince 
systematists that names do make a difference? For example, Sneath (1995: 281) sees 
an 'ambiguity in the meaning of the term "numerical taxonomist" between the original 
broad sense of those who use any quantitative computer methods and the narrow 
sense of a discernable group of systematists who practice numerical phenetics.' If 
numerical taxonomy is defined in terms of the use of quantitative techniques, then 
it has been extremely successful, because the taxonomic literature is now filled with 
algorithms and computer programs. However, even though Camin and Sokal (1965) 
published one of the earliest papers on using computers to infer phylogeny, numerical 
cladistics is not seen as a branch of numerical taxonomy. Instead, fair or not, it is 
viewed by most systematists as a branch of cladistics. Numerical taxonomists would 
