There are many problems with externalism.
For example, if narrow content doesn't completely encapsulate meaning, and we have to take into account our causal interactions with the world, then perhaps our beliefs will not fully cause our behaviour and instead those very causal interactions will do so instead. As John Heil writes:
"… although Wayne and Duane’s beliefs play a part in determining what they do, their being beliefs – that is, propositional attitudes with definite contents – is causally irrelevant to their behaviour. Given the same stimuli, both will behave identically." (425)If the meaning or the mental content is dependent on the world, or on causal interactions with the world, and meaning determines behaviour, then not only belief will determine behaviour – our causal interactions and the world will also do so.
Again, we may not be fully aware (or aware at all) of all our causal interactions with the world. Does this means that we don't fully know what we mean or even what we actually believe? Similarly, does this mean that we will not fully know (or know at all) how we will behave given a set of circumstances?
If how we behave is fully determined only by stimuli or our causal interactions with the world, then aren't we like machines or computers in that they too don't know the meanings of their input; just as, on this picture, we don't know fully the meanings of the sensory stimulations we receive and therefore the meanings of our mental contents or beliefs. Surely this would mean that we sometimes act in ways that we didn't know or predict we would act. We would be like thermometers which rise or fall without knowing why they rise and fall.
Heil puts the internalist objection to this position in this way:
"To the extent that an object’s behaviour is affected by its current state, how the object came to be in that state is irrelevant to how it comes to behave." (426)What matters is not how we came to have our current mental state, or what causal interactions with the world brought it about; but the nature of the state itself. This is the internalist or Cartesian point. All that actually matters is the mental state as it is now, not how it was caused and not what caused it. The mental state itself is all we need to know and need to have access to. Only that, on this internalist picture, matters when it comes to our behaviour.
For a start, perhaps a vast array of things in the world might have brought about our current mental state. Not only that: perhaps historical factors dating back in time (perhaps indefinitely) were also partly responsible for my current mental state. Clearly we can't know all that is to be known about the aetiology of our current mental state. Perhaps we don’t need to know anything about any of the causes of it. That is the Cartesian position. However, if one is an externalist, it's important to know what causally determined our beliefs and therefore our behaviour otherwise we can't be in complete control of either our beliefs or our behaviour. Heil concludes:
"If you take externalism seriously, however, you will regard the contents of agents’ beliefs as being determined by historical factors. In that case, the contents of beliefs could make no difference to how agents’ behave." (426)To the internalist, only narrow content matters. The externalist, on the other hand, believes that only wide content matters. To the internalist, the agent can indeed be completely unhooked from the world. To the externalist, we are thoroughly stuck (as it were) to the world. Or, more accurately, we are part of the world. There is no ‘external reality’. To know our beliefs or the mind itself is to know the world. To know the world is to know our minds or our beliefs.
If one is a naturalist, it will come as no surprise to say that we are part of the world. Of course we are! If we a part of the world, and there is no true ‘external reality’, then it's the case that scepticism about the existence of the world, or even about its nature, is thoroughly vanquished. Cartesianism or internalism itself creates the sceptical dilemma by completely separating mind from world. It does this, effectively, by proposing substance dualism. If internalism is rejected, then, so too is substance dualism. Thus agents can take there place as ordinary inhabitants of the world. Mind and world are brought back together. Indeed mind and body are also brought back together. We don't have souls after all.