The Extended Mind Advanced Topics in Mind and Knowledge Lecture 6 Varieties of Externalism I • The debate over semantic externalism questions whether mental content is individuated purely according to internal factors (ones in our heads/brains/bodies). • But what about the thoughts which possess this content? Both sides of the semantic debate have tacitly assumed that internalism regarding thoughts is true. • Is this the case? Varieties of Externalism II • First, let’s clarify the question a little. Hurley (1998) distinguishes between ‘contents’ and their ‘vehicles’. So far we’ve concentrated on contents, but what are vehicles? • Vehicles are the mental processes which subserve, or make possible, particular thoughts. • The content of a thought tells us what it is. Its vehicle tells us how it happens (Hurley 2006). The Extended Mind I • Clark and Chalmers (1998) argue for externalism regarding mental vehicles, using two thought experiments: tetris and Otto. • Tetris involves rotating shapes to fit them together without leaving any gaps. Intuitively, there are two ways we could do this: 1. Using our imagination to visualise a rotation. 2. Pressing a button to produce a rotation. • The first case looks like a mental process, and the second not. Tetris, Cyborgs and Martians • But what about these further cases: 3. A cyborg has the computational machinery from (2) implanted in his head. • Would we then classify the rotation as a mental process? Maybe, maybe not. 4. A Martian whose natural brain includes the machinery in (2) and (3). • Surely in this case we have a mental process? The Extended Mind II • The extended mind hypothesis invites us to treat (2), (3) and (4) as equivalent – the only difference being whether the mental process occurs inside the head or not. • But isn’t this a significant difference? – Perhaps it isn’t. According to functionalism, what matters for cognition is a particular role being filled, not what fills that role. Why should we think that what fills the role needs to be in our heads? Otto’s Memory I • But image rotation is a relatively peripheral mental process – maybe externalism is true in such cases, but not for central cognitive processes involving beliefs and desires etc. • Clark and Chalmers’ second example suggests not. • Otto has mild Alzheimer’s and so a poor memory – he carries a notebook in which he records information. Otto’s Memory II • On hearing of an exhibition, Otto checks the address (53rd St.) in his notebook and goes there. • Compare Inga, who remembered where the exhibition was and went. • Clark & Chalmers: both had the same belief (that the exhibition was on 53rd St.) even before Otto checked his notebook. • So Otto’s long-term beliefs aren’t all in his head. • And if Otto’s aren’t, then neither are ours! Objections and Replies • The Extended Mind seems to polarise opinions – it’s often regarded as obviously false, or obviously true. • There have been a number of objections made, many of which run parallel to responses to the arguments for semantic externalism. We’ll consider a couple here. The Control Objection • Loosely parallels narrow content: • Even if external objects and processes are sometimes involved in cognition, isn’t it internal biological processes which control the others, and have the ‘final say’? – Reminiscent of ‘Cartesian theatre’ in which a homunculus observes the internal workings of the body, and does all the ‘real work’. – Why identify the control mechanism with the agent? (Clark 2003, forthcoming). – There may be no processes which have the ‘final say’ (Dennett 1995). Compatibilism • Similarly to two-factor theories of content, it may be that mental processes are both internal and external. • Stalnaker (forthcoming) observes that external processes (e.g. diaries) are publicly available and susceptible to sabotage, so we tend to be circumspect in using them. This is quite unlike access to internal resources. – Clark (forthcoming) replies that we do not always use external resources cautiously, and that is enough to support vehicle externalism.