incoherence, self-referential See SELF-REFERENTIAL. INCOHERENC. incommensurability, in the philosophy of science, the property exhibited by two scientific theories provided that, even though they may not logically contradict one another, they have reference to no common body of data. Positivist and logical empiricist philosophers of science like Carnap had long sought an adequate account of a theoryneutral language to serve as the basis for testing competing theories. The predicates of this language were thought to refer to observables; the observation language described the observable world or (in the case of theoretical terms) could do so in principle. This view is alleged to suffer from two major defects. First, observation is infected with theory – what else could specify the meanings of observation terms except the relevant theory? Even to perceive is to interpret, to conceptualize, what is perceived. And what about observations made by instruments? Are these not completely constrained by theory? Second, studies by Kuhn, Paul Feyerabend, and others argued that in periods of revolutionary change in science the adoption of a new theory includes acceptance of a completely new conceptual scheme that is incommensurable with the older, now rejected, theory. The two theories are incommensurable because their constituent terms cannot have reference to a theory-neutral set of observations; there is no overlap of observational meaning between the competitor theories; even the data to be explained are different. Thus, when Galileo overthrew the physics of Aristotle he replaced his conceptual scheme – his ‘paradigm’ – with one that is not logically incompatible with Aristotle’s, but is incommensurable with it because in a sense it is about a different world (or the world conceived entirely differently). Aristotle’s account of the motion of bodies relied upon occult qualities like natural tendencies; Galileo’s relied heavily upon contrived experimental situations in which variable factors could be mathematically calculated. Feyerabend’s even more radical view is that unless scientists introduce new theories incommensurable with older ones, science cannot possibly progress, because falsehoods will never be uncovered. It is an important implication of these views about incommensurability that acceptance of theories has to do not only with observable evidence, but also with subjective factors, social pressures, and expectations of the scientific community. Such acceptance appears to threaten the very possibility of developing a coherent methodology for science. See also PARADIGM , PHILOSOPHY OF SCIENCE , THEORETICAL TER. R.E.B.