Thursday, 26 May 2011

The Assumptions of Representationalism

It had seemed to me a fact of the most obvious kind that we are fundamentally related to the world as subjects perceiving objects. I’ve made sweeping epistemic and ontological decisions based on this assumption and on account of its seeming so obvious I never stopped to question it. I am me, this is this – end of story. When I began to take philosophy more seriously I found myself drawn to subjectivist or sceptical positions on account of this veil of perception which I believed occluded me to the proper nature of the things I encountered.

When I began to come across non-representationalist accounts of what had typically been known to me as “mental” phenomena, of course I was interested. This was a claim which was so fundamentally alien to me that I just had to have a peek… The idea is that we are not first and foremost thinking subjects related to objects by means of mental representations, but are rather beings with a way of unthinkingly engaging with a world in which we are wholly absorbed. The representationalist picture wanted to demonstrate how one’s world was revealed to them by the medium of “internal” representations, however what philosophers have noticed is that most of our dealings are in fact non-representational. Walking, for instance, or our use of language or equipment – all pass over us transparently, without our explicit awareness. Even our awareness of ourselves is rendered transparent in the midst of what Heidegger calls “coping” (Drefyus, 1991).

Our most primary mode of engagement with the world is not through a canvas of conscious deliberation, calculation, belief, desire, and ultimately action. We can, and more often than not do, engage with the world unthinkingly. It was recently discovered that the brain plays a much more insignificant role in movement than was previously thought, that the spinal column and legs control movement themselves, while the brain simply sets up the nervous system to anticipate movement (sounds hauntingly like Merleau-Ponty, no?) So the act of conscious volition is not even neurologically grounded in the case of movement - it simply takes place in the body, responding to its environment accordingly, devoid of representation or any feeling of acting.

Heidegger separates two modes of dealing with things, the practical (zuhandenheit) and the theoretical (vorhandenheit). The practical mode of being involves engaged participation: one understands a piece of equipment through using it, through its having a purpose corresponding to a possibility of one’s being, e.g. the hammer is something “with-which” you hammer nails, “in-order-to” build the house, “for-the-sake-of” being a carpenter (King, 2001). The theoretical way of looking at things, however, involves detached inspection – when all using and producing, etc., has ceased and all that remains is understanding things in their “outward appearance” (Heidegger, 2010).

That we must have a practical understanding of our world (taken not to denote the totality of things in existence, but rather the inter-referential nexus in which things play a part: one cannot make sense of the paper without the pen, and neither are intelligible without the possibility of writing or drawing) before we can have detached, theoretical, mental representations is not obvious. Strong AI theorists, for instance, believe it possible to program in a list of purely objective, factual data to a machine which does not share our world, and have it demonstrate consciousness purely through such factual representations. To establish the necessity of being “worlded” in order to have such representations, I’ll borrow two examples from Hubert Dreyfus:

(1) The first example involves a chair. One could be told “this is a chair” and shown an image. The factual data gleaned from this “merely looking” at the image would allow one to point out things which have similar objective properties and say “that is a chair”. However, this is not sufficient for properly understanding what a chair is. If you had one who was from a culture unfamiliar to seating, they could reproduce the expression “chair” whenever they saw something with the same properties, but they could not know what a chair was without knowing what it is for. This understanding cannot be given over by objective, factual, representational, knowledge. One must come to understand the chair as an object of use with a purpose by engaging with it.

(2) The second example involves a broken radio. If we were to say to someone “if it is not making sound, it is broken” it would be a simple enough test for them to check if it was broken, right? So they approach the radio, fiddle around with it for a bit, try turning it on and no sound comes out. It’s broken, there – easier than we would have expected. But try doing this without world. By having a world in which things make sense, expressions like “broken” come to have significance – how? Only on the basis of an item’s being for something can it fail to do what it is supposed to do. And only by being so constituted as to reach out to things in this practical manner can we understand something’s not doing what it’s supposed to do. Our engagement, our being tangled up in things, is what is required for understanding.

The mistaken assumption is that one can simply program understanding of brokenness into a computer by giving it some list of data through which it is said to know when something is broken. However, because no previous understanding of brokenness stands with the machine, we’re left with a merely behavioural mimicry. The machine was never reaching out to the radio on the basis of a purposive aspect, and so cannot possibly grasp how something can be broken, a notion which is from top to bottom couched in purposiveness You can look at the objective properties of a broken radio all you like, “rectangular”, “weighs 2 pounds”, “silver”, “produces no sound”, but without the background of world it’s just a list of facts stripped of their significance.

So when we come to have explicit mental representations like “this is broken” or “that is a chair” we can only do so on the basis of a non-cognitive, non-representational background understanding of how things fit together in the world. We cannot hope to gain an understanding of expressions like “this is broken” or “that is a chair” without engaging with them, and being tangled up with them in this inter-referential world through which things come to have meaning.


Drefyus, H. (1991). Being-in-the-World. MIT Press.

Heidegger, M. (2010). Being and Time. SUNY.

King, M. (2001). A Guide to Heidegger's Being and Time. SUNY.