Anton Alexandrov

This paper argues that the concept of human mindedness may be employed as a Carnapian explicatum that enables progress to be made in the debate about strong AI, that is, the thesis that ‘the appropriately programmed computer really is a mind’. It is shown that a holistic notion of human mindedness – conceived as all of what may typically occur in the mental life of a human being – yields greater precision without sacrificing simplicity or similarity to ordinary psychological concepts, including consciousness and intelligence, which were formerly central to the debate. The alternative approach is assessed by reference to Carnap’s four criteria of adequacy – similarity, precision, fruitfulness, and simplicity – and defended against Strawson’s charge of changing the subject. The paper concludes that explication provides a methodologically superior framework for testing strong AI and promises progress on further issues in the philosophy of AI.

ABSTRACT In recent debates about conceptual engineering (broadly understood), it appears that the internalist has an explanatory advantage when it comes to accounting for meaning change and conceptual change. In this paper, I argue against this impression. I show how two different varieties of externalism, originalism and anti-individualism, can coherently explain various cases of meaning change, irrespective of whether they involve proper names or kind terms; and also irrespective of whether they occur in everyday, legal, or scientific contexts. I point out which mechanisms and interpretational strategies the originalist and anti-individualist can invoke, relative to their respective background assumptions about the nature of concepts. Furthermore, I offer an argument in favour of the claim that externalism as such has no implications concerning whether concepts can undergo change. I establish this by showing that originalism allows for conceptual change while anti-individualism insists on conceptual stability.