Rejection of the personal aspect of knowing, or the subjective element, leads to an exaggerated emphasis on objectivity -- a desire for a level of certainty that eventually becomes self-defeating.
Objectivity is based on the assumption that synthetic statements are possible and necessary. The observer is not independent of what is observed; truth is impersonal and theory-barren. But the ground for believing in the general reliability of sense perception is not itself something sensed; it is a belief. Senses cannot always be trusted. Things are not what they seem.
But if truth is conceived as purely independent of all commitment and impersonal, then one's own choice of true/false cannot be acknowledged. This dilemma "has long haunted philosophy in the guise of the 'correspondence theory of truth.' Bertrand Russell, for example, defines truth as a coincidence between one's subjective belief and the actual facts (The Problems of Philosophy, 4th ed., 1919, p. 202); yet it is impossible, in terms which Russell would allow, to say how the two could ever coincide.
Positivism, as promoted energetically by Ernst Mach in his book Die Mechanik (1883), rejected the claims of scientific theories of any inherent persuasive power, which he considered metaphysical and mystical. Scientific theories served only as convenient summaries of direct experience. Positivism culminated in the "verification principle", which stated that no proposition could be considered true, or even meaningful, unless it could at least in principle be verified by observation.
Trouble was, the verification principle itself could not be verified by direct observation. It was in fact a "metaphysical" proposition. Since there appeared to be no escape from this situation, by the 1960's positivism collapsed. It is no longer seriously advocated by philosophers.
It is now common to refer to all observations as "theory-laden". There is no general formula for the demarcation of theory and observation.
Return to dialogic diagram