摘要：The paper deals with the so-called Preference Theory developed in the works of Theo Vennemann and Robert Murray within the scope of historical phonology. The first part of the paper examines the constituting assumptions and claims of the theory. The goal of the preference-based historical phonology – uncovering the motivation for sound changes which the Neogrammarian methodology can merely describe – will be achieved only if the universal preferences are reliably established. It is shown that the procedures which are employed to extract the universal preferences from empirical data do not lead to reliable results. The reason for this is the failure of the Preference Theory to distinguish in a non-arbitrary way between the alleged universally preferred structures and the mere by-products of sound changes with different or unknown motivation. The second part of the paper examines a recently suggested modification of the traditional notion of the exceptionlessness of sound changes. According to Vennemann, the traditional exceptionless sound changes are in fact to be considered as non-exclusive tendencies towards universally more preferred phonological structures. The paper shows that this position is neither based on the core assumptions of the Preference Theory nor supported by the adduced empirical evidence.
关键词：historical phonology; preference laws for syllable structure;
sound laws; regularity of sound change; Neogrammarian controversy
