Causation and determination are closely related but distinct concepts. I will not pretend that defining these concepts is not itself problematic and even controversial. Nevertheless, causality is relational in nature: the relation of cause to effect involves the cause being in some way prior to the effect and either partially or wholly responsible for the effect. Determination is generally less focused upon the relation of cause and effect as entities as much as it is focused on the relation of particular qualities of the cause and dependent qualities of the effect. Or, in other words, X causes Y tells me that X is prior to Y and is either partially or wholly responsible for Y; thus, also, Y is contingent upon X, either in whole or in part. If X is a necessary cause of Y, I can state unambiguously that if Y were not, X would not be. If I said, on the other hand, X determines Y, I would be stating that some set of Y's qualities are "fixed" by X; or, again, that Y by virtue of its causal relationship to X also has a similar but derivative causal relationship to X's qualities. I would argue that this distinction is uncontroversial, although I understand if some of my readers would argue that it is nevertheless completely artificial.
Yet another way to discuss "determination" is to link it to the concept of predictability. by predictability, I mean that before X's causal interaction with Y, it is possible to accurately know how all the relevant qualities of Y will be changed before the interaction takes place. An interaction with an unpredictable outcome would not be a determining interaction, because some of the qualities of Y after the interaction cannot be ascertained from a consideration of X. Now an interaction can be unpredictable in three ways: accidentally, partially, or arbitrarily. An accidentally unpredictable interaction is one that is not unpredictable in itself but is rendered unpredictable either due to the impossibility of ascertaining all relevant information needed to make a prediction or because the laws by which the interaction occurs are unknown. It is needless to say that free will cannot be established on the ground of accidental unpredictability. Nor can it be ascertained on the grounds of arbitrary unpredictability, meaning that there is no connection whatsoever between the initial conditions and the conditions after the interaction. Because I have defined free will as requiring an element of control, I cannot be satisfied by simply saying that the will arbitrarily acts, as this means there is actually randomization rather than responsible control. However, an event may be partially unpredictable if there are intelligible connections between the circumstances prior to the causal interaction and the effect of the interaction that are simply only observable after the event; that is, if when agent Q in situation R decides to perform action X, the relationship between R and X is intelligible but could not have been conclusively predicted before the action actually took place.
In order to bring these distinctions into what I hope might be a more sympathetic light, I would ask that one consider for a moment what a faculty like free will, if it could exist, would be. Clearly, it is considered a faculty of choice with regards to action: that is, my free will would have the power to determine myself with respect to some action or inaction. I would strongly disagree, however, with any attempt to separate my "self" from my "will" in an artificial way that makes the real "me" into some agent operating behind my will. Clearly, such a model would degenerate into an infinite regress of choosing. The real "me" would have to have some faculty of choice by which I choose what I would will, and so forth. My will is part of my self. Another important distinction I would like to make is to point out that my will does not operate in autonomous seclusion from the rest of me, including my intellect.
In any case, this faculty would clearly not be uncaused; whatever forces are responsible for my formation as a mature rational agent are also responsible for the formation and development of my free will as a faculty. But I would argue that because one of the requisite qualities for a free will is that it be able to determine myself with respect to some action or inaction, it is not necessary to believe that my free will is determined by its causes with respect to making that determination or choice. The free will would have to have this power in order to operate as it does. It would have to be the case that my will developed as precisely the sort of thing that has the power to determine itself with respect to actions and inaction.
There are three models I would put forward for discussing free will in the way I have outlined above. These are not three different models but rather three views of the same model, through which I hope to clearly describe the features of a free will which makes partially predictable choices. Underlying all of them is the concept that the will is incorporated into our rationality in such a way that it is not determined by its causal history, though it is limited by that history in certain ways which are actually necessary for free choice.
The first model is the causal gate model, in which the will is understood as being presented with a multiplicity of potential causal histories for its future actions. In short, as rational beings we are presented with the possibility of acting in response to different aspects of our own causal history and present circumstances.
The second model is the teleological model, in which the will is provided with the possibility of choosing among different proximate ends.
The third model is the indeterminism of thought model, in which the will's reciprocal control over our intellectual processes provides for the possibility of partially unpredictable, non-arbitrary judgment with regards to future action.
In the next part, I hope to discuss the first of these models: free will as the causal gate.