536 
Telopea Vol. 6(4): 1996 
Operationism revisited 
Johnson agrees that scientific concepts must be made as operational as possible. 
Operationism as a philosophy of meaning can be shown to be impoverished (Hull 
1968). It cannot begin to do what it is supposed to do, even in those areas of science 
where it arose — relativity theory and behavioral psychology. If operationism won't 
work in any other area of science, what makes biologists think that it can work in 
biology? However, a more limited sense of operationism is absolutely central to 
science. In looking back on my own work, I am most uneasy about my blanket 
rejection of operationism. Sure, as a general philosophy of meaning, it won't do, but 
did I really think that the pheneticists were attempting to produce a general theory 
of meanmg? If philosophers really want to contribute to our understanding of science, 
we might use our time more wisely by studying how scientists actually do 
operationalize their concepts rather than working out the logical consequences of 
operationism as a philosophy of meaning. The trouble is that very little can be said 
about this process that is in the least bit general, and philosophers are concerned to 
uncover generalizations about science. 
Johnson is in a much stronger position than philosophers in this respect. He rejected 
operationism as a philosophical program but could exemplify in his own work the 
ways in which scientific terms can be operationalized. For example, Johnson (1972:18) 
coined the term 'umbellaster' to refer to the basic unit of inflorescence in Eucalypts. 
He then went on to explain how to recognize such units, in part through a diagram. 
Does it follow that all attempts to operationalize scientific concepts must include 
diagrams? Obviously not. The problem with saying anything general about how 
scientists operationalize their concepts is that this process is highly particularized. 
You can list example after example, and that is about it. Perhaps the methods that 
scientists use to operationalize their concepts exhibit some interesting regularities. 
So far I have not been able to discern any. 
Did the pheneticists really think from the start that all systems of classification, 
including their own, depend on v^arious theories, including theories of homology, 
theories of information-rich groupings, theories of general causes, and theories of 
history (Sneath 1995: 285)? I think not. At the very least, their repeated protests 
against 'theories' and 'speculation' served as a rallying cry for the pheneticists. 
'Your work rests on the shifting sands of theories and speculations, while our work 
rests solidly on observable fact.' If the pheneticists had from the start emphasized 
that their views differed from those of evolutionary systematists only in degree, not 
in kind, they are not likely to have caught the imagination of young systematists. 
Cladists carry on in this time-honored tradition. 
Conclusion 
As sceptical as Johnson is about the long-term usefulness of all the methodological and 
philosophical discussions that have been carried on primarily in the pages of Systematic 
Biology (formerly Systematic Zoology), I would like to say a few words in their defense. 
For one thing, they have shown that some ideas that initially looked very appealing 
just don't pan out. For example, the notion that out there in nature there is something 
termed 'overall similarity' has repeatedly seemed quite plausible to a lot of people, 
including scientists. If the numerical taxonomists have done anything, they have shown 
that this belief is illusory. Several of us have argued against this position, but the actual 
failure of so many bright, hardworking systematists to deliver on the promise of a 
general-purpose classification based on overall similarity is even more convincing. 
If after forty years, they have not been able to do it, then probably it cannot be done. 
