© 2005 Graham Little
The notion of the homunculus has been an issue in discussions of consciousness for hundreds of years. In short, the argument went as follows: we can imagine consciousness as like a little TV screen inside one’s mind, with ‘us’ then watching the screen. But if so, then who is watching the person watching, and who is watching them…it leads to an infinite regress with no resolution of ‘who is watching?’ so consciousness left hanging.
It was in part issues of this sort that guided various people to seek alternatives, like Roger Penrose who wrote several books arguing consciousness was rooted in quantum variability. Unfortunately his discussion tends to overlook so many core issues as to be hardly worth seeking to refute, and this is not the paper to do such a refutation. Suffice, for example, no where does Penrose deal with the question ‘what is cause?’ So everything he says presumes that the analysis of causality will no have impact on it, so he can ignore causality, and speculate away. Now given that the whole of quantum mechanics presumes to be causal, and if it is not then why do people bother to ‘interpret it,’ this seems a pretty remarkable position for a Nobel Prize winner and presumed scientific thinker to adopt since an analysis of causality could decidedly impact how we view quantum mechanics (for example, see the notes on the issue at this web site).
With the model of causality of human conduct developed at this site, a very different perspective emerges. Most notably the system is parallel and is not hierarchical as offered in the original homunculus argument. So it is quite possible for some part of the system to observe or to take in as input events in from other part of the system and for yet another part of the system to accept all of the first systems as input, etc, and for none of this to entail or imply any hierarchy, hence no infinite regress, no problem of the homunculus. The original homunculus argument was terminated by an infinite regress, and this caused the issue. So what terminates the argument with parallel systems?
Within the model ‘self’ is a critical notion embodied in a quite finite number of critical sub-units of the overall system of mental sets. In the absence of involvement of any of this part of this ‘self structure’, then we will not feel involved; the ‘observer’ will not be felt to be present.
The self can also be divided, but because self exists in a set of sub-systems that are very finite, this division cannot go on indefinitely, eventually ‘self’ can be divided no further without losing the very sense of ‘self’ being involved.
We can now put the issues in the model together and state a significantly limited form of the homunculus does in fact exist, but this form is limited by the finite structure of self and not by any infinite regress. In fact any regress of self watching self watching … is a deliberately constructed circumstance, since all such systems that would naturally occur are parallel not hierarchical, and would typically reflect a person contemplating and monitoring several issues and tasks at once. For any and all forms of division of self, the extent of the division is limited to the finite resources that in fact constitute ‘self’ within the system of mental sets that form our psyche.
Given this structure, to explain what we sense and see and feel, etc, a theory of consciousness does not need to go beyond self-knowledge, self-understanding, self-awareness, self-monitoring, and goal setting, all occurring within an extremely complex and highly interactive structure. Self as known and understood, self as experienced, self as seen by others, self as sensed and not conceptualised, self as felt, self as a the sense of values and rightness, none need go beyond that which we have and can readily understand without need of extreme or remarkable speculation.