The Nature of Error and the Role of Free Will

For a metaphysics asserting the existence of an infinitely-good God, a perfect being in itself incapable of error, it is necessary for its coherence to account for the apparent existence of imperfection - error - in the world. Descartes meets that demand by postulating a wholistic approach to the perfection of - as opposed to in - the world, and finds the human share of perfection in their free will instead of the sum of their knowledge.

It may be that Descartes's choice of a very Christian metaphysics already presumes the existence of a free will and explains the ease with which he uses it to account for the presence of error. It may also be that my, very subjective, defence of Descartes presented below falls into the same trap. However, I would rather view the necessity of a choice a philosopher makes when defending an argument, an evidence for the necessity of a free will, as much as doubts form an evidence for the necessity of our existence. Free will, with error as its direct consequence as Descartes shows, belongs to the same ontological category as mind, which, although it can be asserted objectively, is intrinsic only to the human existence, and should therefore be investigated without external references.

To investigate error, defined in its broadest sense as a partial lack of perfection, is to investigate the nature of perfection itself, although Descartes is careful to emphasise the complementary nature of each. The naive approach is to identify error with the sum of the imperfection in the world. This, however, is made implausible by the realization that we would like to see perfection as the sum of reality - atheists because there is nothing beyond the actualized world, and therefore nothing more perfect than it; theists because of the world's perfect creator - and would consequently find no place for error.

Because Descartes feels no need to question the perfection of the sum of reality, and, further, he is required by his faith to assert the perfection of people as created in God's image, he amends the naive interpretation of error as the lack of perfection. He notes that for God to create utter perfection, and thus eliminate error, would be for him to reproduce Himself. The difference between God and the world is that the later can be partitioned into objects which, each in themselves, are capable of error.

At this stage it must be noted that, so far, we have assumed a mysterious universality of error which allegedly penetrates every area of existence. However, such a generalisation is certainly a nonsense, as there are aspects of human life in which truth and falsity does not play a role (appreciation of an aesthetic beauty is but the most obvious example). Descartes proceeds to reason that, if out understanding is not perfect, we must posses some other capacity which is (a more detached approach would be to say “may be’.) The most obvious candidate is the concept which has traditionally been used to account for error - the free will.

It may seem paradoxical that Descartes, who insists on certainty obtained through rigorous reasoning, seems to find a concept superior to knowledge. By accepting the alleged infinity of the free will, he does not reject the importance of understanding, but rather names its agent. Free will, for Descartes, is the medium through which our rational thoughts are actualized as mental or physical events.

We can now proceed to state a new definition of error as our choice to affirm ideas that are in fact wrong or false. I find this definition very straightforward and very traditional. Nevertheless, the concept of free will as both the sum of our perfection and the source of our errors is an appealing one even though serious objections to it have been pronounced.

So far, our discussion has been mostly hypothetical. Now, we should consolidate it with reality through a more structured argument.

Before I present an argument for the existence of free will, I will investigate its logical consequences. Free will, defined for our purposes as our ability to assert propositions regardless of their truth or falsity, implies that, when A is faced with a proposition P, she has three choices. A may elect not to further her knowledge, by pronouncing her ignorance of P. However, she may also assert or reject P. In the later case, she enters one of two states, either that of understanding, had she asserted true P or rejected a false one, or that of an error, when she asserts a false P or rejects it while true.

According to this model, the state of error is complementary to the union of the states of a conscious ignorance and of an understanding (whether conscious or not). Therefore, error defines the bounds of our rationality in that it serves as an all-inclusive set of the remaining reality. As such, error is an essential part of reality, and not necessarily contrary to its perfection, provided that perfection does not imply an infinite capacity for understanding.

I would like now to postulate an a posteriori knowledge of our free will. This can be done by considering the difference between two possible worlds, F actualized with the free will, and N actualized such that everything happens out of necessity. Assuming a world with multiple individuals, and further restricting ourselves to rational individuals asserting rationally obtained propositions (which we can do without loss of generality, as otherwise the cause enforcing our “choice” could be attributed to, for example, a mental illness), we must conclude that in N, all individuals would be necessarily led into the same conclusions about the same facts. That is to say, in N, different causes cannot influence different individuals forcing them into a spectrum of conclusions, for that would not only impair, but deny our capacity for knowledge. Such a restriction is hardly acceptable. We cannot even qualify the situation in N so that some propositions are necessary true and evident for everyone, and some are necessarily confused amongst individuals. For the proposition in question, namely the question of existence of free will, would certainly belong to the later category, and, having just lost any sensible meaning, this discussion would loose its subject matter.

I believe that the very existence of controversial topics (such as the alleged existence of a free will), demonstrates the necessity for unanswered - if not unanswerable - questions. The status of these questions poses serious problems for the thesis of the non-existence of a free will, for it would imply that anyone taking a stand about an as-yet unanswered topic is either doing so irrationally, or is classifying the topic as subjective. Clearly that would be a very serious objection to every axiomatic science.

I am thus left to conclude that the only world that could be plausibly actualized is F, and we are led to accept the existence of free will, and attribute to it not only our errors, but also our understanding and our conscious ignorance.

Having thus accepted the existence of free will, the remainder of Descartes's model of error can be justified.

By definition, free will allows us a more or less unlimited choice when accepting a truth of a proposition. In particular, our understanding is independent of the reality of the world, in that an assertion of P is disjoint from P itself. In particular, we are free to assert a false proposition, in which case we are falling into an error.

Note that by the above I do not claim that there is absolutely no causal chain between P and our comprehension of P. On the contrary, such a causal chain, as asserted by any Realist, exists and is well-defined. However, that causal chain is not enforcible in that we are free to disregard the evidence for P. The very existence of unanswerable questions of the sort “are there objects undetectable by the senses?” gives us a room for belief. And while our assertion of a proposition does not make its truth, not even does it assure our knowledge, the fact remains that we are capable of accepting, rejecting and withholding judgement. Necessary, we are thus capable of making false assertions, which are otherwise known as error.

In the second part of Ethics (III B(i)), Spinoza replies to Descartes's Meditations: “I do not see why the faculty of willing should be called infinite, when the faculty of sensing is not.” While I believe to have already replied adequately to his objections as to the separation of the will and the understanding, the problems outlined in the above statement are worth analysing. The confused concept of infinity as used by Descartes with reference to the will certainly justifies such objections. However, the “infinity” of will lies not in the utter lack of constrains, as we would like to imagine the infinity of knowledge, but rather in the lack of bounds of a sort. Knowledge is bounded by two factors, the reality and our finite resources. As we have seen, the bounds of reality do not apply to the free will. I am also unable to imagine why would we, in principle, be unable to will for anything, even when we do not comprehend it. Our, often more blind then it should be, faith in mathematical axioms is the most obvious example of the later. To summarise, from the trifold-choice model presented above, it is obvious that there is nothing parallel between our faculty of willing and our faculty of understanding, as the earlier is an constraint on the later, and in fact on all the areas of our thought. So much as we limit ourselves to that description, our free will is infinite, while our understanding cannot be.

Error - our ability to assert propositions regardless of their truth, is an important part of the metaphysics of a world regardless of its level of perfection. It does not impair the perfection of the world insomuch as our understanding does not have to be the ultimate measure for perfection, and the limits imposed on knowledge do not necessary limit the perfection of the totality of facts. Free will actualises our capacity for error, but also serves to further our understanding by providing the scope and the boundaries for choice and for the critical rejection and assertion of doubt. Further, free will is unbounded in that is serves as bounds for our comprehension, ignorance and understanding. Whatever else may be said about the free will which has not been said in this discussion, and despite the many questions left unanswered or unconcluded above, it is evident that the existence of free will can be asserted with logical certainty, and used to model and account for the error limiting our knowledge.

Patryk Zadarnowski, Sydney 1997