The paper aims at reconsidering the problem of “practical knowledge” at a proper level of generality, and at showing the role that personal abilities play in it. The notion of “practical knowledge” has for long been the focus of debates both in philosophy and related areas in psychology. It has been wholly captured by debates about ‘knowledge’ and has more recently being challenged in its philosophical foundations as targeting a specific attitude of ‘knowing-how’. But what are the basic facts accounted in the “knowing-how” debate? The problem is much more fundamental than knowledge: it addresses the need for an explanation of intelligent or guided behaviour, that could account for some distinctive aspects involved in the performance, but without positing too much beyond the observable actions. This is what I call the problem of “practical mastery” (PM). PM raises three questions: what kind of behaviour require such an explanation? What is distinctive about practical mastery? What does it consist in: a form of knowledge, or something else? I argue here that the notion of ability offers a less restrictive, though no less powerful answer to these three questions. It can offer an independent objective grasp on the subjects of attribution. I conclude that the notion is central both to account for common-sense psychology and to understand what experimental psychology actually measures and tests for.

In this paper, we discuss contextualism, a philosophical position that some pragmatists have endorsed as a result of the philosophical reflection on pragmatics as a science. In particular, we challenge, from the results on phrasal pragmatics, the contextualist approach on incomplete definite descriptions and referential metonymy according to which optional pragmatic processes of interpretation are required (an optional pragmatic process of recovering unarticulated constituents for incompleteness and an optional pragmatic process of transfer for metonymy). By contrast, we argue from the standpoint of phrasal pragmatics that what is descriptively referred to depends, in both cases, on truth-conditionally mandatory pragmatic processes of recovery of unarticulated constituents.