The Identical Rivals Response to Underdetermination

Frost-Arnold, Greg and Magnus, P.D. (2009) The Identical Rivals Response to Underdetermination. In [2009] Underdetermination in Science (Pittsburgh, 21-22, March 2009).

Abstract

The underdetermination of theory by data obtains when, inescapably,
evidence is insufficient to allow scientists to decide responsibly
between rival theories. One response to would-be underdetermination
is to deny that the rival theories are distinct theories at all, insisting
instead that they are just different formulations of the same underlying
theory; we call this the identical rivals response. An argument adapted
from John Norton suggests that the response is presumptively always
appropriate, while another from Larry Laudan and Jarrett Leplin
suggests that the response is never appropriate. Arguments from
Einstein for the special and general theories of relativity may fruitfully
be seen as instances of the identical rivals response; since Einstein’s
arguments are generally accepted, the response is at least sometimes
appropriate. But when is it appropriate? We attempt to steer a middle
course between Norton’s view and that of Laudan and Leplin: the
identical rivals response is appropriate when there is good reason for
adopting a parsimonious ontology. Although in simple cases the
identical rivals response need not involve any ontological difference
between the theories, in actual scientific cases it typically requires
treating apparent posits of the various theories as mere verbal
ornaments or computational conveniences. Since these would-be posits
are not now detectable, there is no perfectly reliable way to decide
whether we should eliminate them or not. As such, there is no rule for
deciding whether the identical rivals response is appropriate or not.
Nevertheless, there are considerations that suggest for and against the
response; we conclude by suggesting two of them.