The aim of the paper is to demonstratethat Special Relativity and the Early Quantum Theory were created within the same programme of statisticalmechanics, thermodynamics and maxwellianelectrodynamics reconciliation. I shall try to explainwhy classical mechanics and classicalelectrodynamics were ``refuted'''' almost simultaneouslyor, in more suitable terms for the present congress,why did the quantum revolution and the relativisticone both took place at the beginning of the 20-thcentury. I shall argue that the quantum andrelativistic revolutions were simultaneous since theyhad a common origin -- the (...) clash between thefundamental theories of the second half of the 19-thcentury that constituted the ``body'''' of ClassicalPhysics. The revolution''s most dramatic pointwas Einstein''s 1905 photon paper that laid thefoundations of both Special Relativity and OldQuantum Theory. Hence the dialectic of the oldtheories is crucial for theory change. Modern physicsbegan with Einstein''s reconciliation ofelectrodynamics, mechanics and thermodynamics in 1905and his unification of Special Relativity andNewtonian Theory of Gravity. Or, in a more generalsocial context: progressive scientific change can bedescribed not in Weberian terms of zweckrationalaction forcing out all the other forms of action onlybut in terms of Habermas''s communicative rationalityencouraging the establishment of mutual understandingbetween the various scientific communities also.Einstein''s programme constituted a progressive stepwith respect to its rivals not because it couldexplain more ``facts'''' or was more ``mathematical''''. Itwas better than its rivals because it constituted abasis of communication and interpenetration betweenthree main paradigms of 19-th century physics. Ofcourse in the long run it resulted in empiricalsuccesses. (shrink)

: In this paper I argue that belief in the greater confirmatory value of prediction over accommodation can best be understood as a function of the practice rather than the logic of science. Attempts to account for this asymmetry within the logic of science have revealed no non-arbitrary way to address the problem of underdetermination as it applies to prediction and thus have failed to account for the preference for prediction over accommodation on logical grounds. Instead, I propose a model (...) that not only justifies and explains this preference, but allows for a richer taxonomy of the types of evidential confirmation that are employed in scientific reasoning. (shrink)