All accounts of causality that presuppose the propagation or transfer or some physical stuff to be an essential part of the causal relation rely for the force of their causal claims on a principle of conservation for that stuff. General Relativity does not permit the rigorous formulation of appropriate conservation principles. Consequently, in so far as General Relativity is considered and fundamental physical theory, such accounts of causality cannot be considered fundamental. The continued use of such accounts of causality ought not be proscribed, but justification is due from those who would use them.

I defend the conserved quantity theory of causation against two objections: firstly, that to tie the notion of “cause” to conservation laws is impossible, circular or metaphysically counterintuitive; and secondly, that the conserved quantity theory entails an undesired notion of identity through time. My defence makes use of an important meta-philosophical distinction between empirical analysis and conceptual analysis. My claim is that the conserved quantity theory of causation must be understood primarily as an empirical, not a conceptual, analysis of causation.

The correlations between distant systems in typical quantum situations, such as Einstein-Podolosky-Rosen experiments, strongly suggest that the quantum realm involves curious types of non-Iocal influences. In this paper, I study in detail the nature of these non-Iocal influences, as depicted by various quantum theories. I show how different quantum theories realise non-Iocality in different ways, whichreflect different ontological settings.

We can distinguish different senses in which a formal language can be said to have been provided with an interpretation. We focus on two: (i) We provide a model (or structure) and a definition of satisfaction and truth in the standard way (ii) We provide a translation into a natural language. We argue that the sentences of a formal language interpreted as in (i) do not have meaning. A formal language interpreted as in (i) models the way the truth of a sentence would be affected by two factors: the interpretation as in (ii) of the language, and a way the world might be. Viewing in this way the relation between interpreting a formal language as in (i) and as in (ii) allows us to justify the conceptual adequacy of the standard model-theoretic definitions of the properties of logical truth and logical consequence.

The standard argument against the compositionality of adjective-noun compounds containing "red" says that "red" does not make the same semantic contribution because a red car has to be red outside whereas a red watermelon has to be red inside. Fodor's reply to that argument is that the inside/outside feature is semantically irrelevant because "red F" just means F which is red for Fs. That account agrees with our intuitions concerning analyticity; but it seems to be in conflict with a central test for understanding: a person who knows nothing else about these expressions than what is offered by Fodor is far from applying them successfully.

Michael Dummett holds that the sense of a natural language proper name is part of its linguistic meaning. I argue that this view sits uncomfortably with Frege's observation that the sense of a natural language proper name varies from speaker to speaker. Moreover, the thesis under discussion is not supported by Frege's views on communication. Recently Richard Heck has tried to develop an argument which is intended to show that assertoric communication with sentences containing proper names is only possible if Dummett's thesis or a version of it is true. I will challenge this argument and argue that it does not support Dummett's thesis.

There are many reasons to assume that the contents expressible by declarative sentences are generally truth-evaluable (reasons stemming from semantics, logic and considerations about truth). This assumption of global truth-evaluability, however, appears to conflict with the view that the contents of some sentences do not admit of truth or falsehood for lack of objectivity of their subject matter. Could there be a notion of truth on which the truth-evaluability of a content does not rule out the non-objectivity of its subject matter?In this paper, I discuss Crispin Wright's criterion of Cognitive Command as a criterion for objectivity. This criterion faces the Problem of A Priori Error. I reject Wright's response to that problem and propose to solve the problem by relativising truth. This move allows for the possibility of contents that are truth-evaluable yet non-objective.