libertarianism

For a half - century, I have been developing a view of free will that is incompatible with determinism and, in the process, attempting to answer the Intelligibility Question about such a free will: Can one make sense of an incompatibilist or libertarian free will without reducing it to mere chance, or mystery, and can such a free will be reconciled with modern views of the cosmos and human beings? In this paper, I discuss recent refinements to my earlier writings on such a view, refinements developed in recent years in response to the large critical literature on my views in the past several decades. My view has usually been designated an event-causal (EC) view of libertarian free will and distinguished from non-causal (NC) and agent-causal (AC) libertarian views. But I was never happy with this designation of my view as “event-causal” and did not use it myself in earlier writings. In this paper, I explain why I now reject it altogether. I have come to believe that to avoid numerous misunderstandings in current debates about free will, we must distinguish four different kinds of libertarian theories, not merely three: in addition to non-causal (NC), agent-causal (AC), and event-causal (EC) theories, we need to add a fourth kind, which might be called an agent-causal/event-causal (AC/EC) theory. My view has always been of this fourth kind. It represents what I call in the title of this paper the “fourth way forward” for making sense of an incompatibilist free will.

This article articulates and defends an indeterministic weightings model of libertarian free will (LFW). It begins by defining the conception of free will at issue and then goes on to present versions of the luck objection which is often made against theories of LFW. It is argued that the sort of indeterministic weightings model of LFW which has been defended in the recent literature by Storrs McCall and E.J. Lowe (2005, 2008) and John Lemos (2018, Ch.5) has the resources to answer such luck objections while possessing virtues which some other libertarian views lack. According to the indeterministic weightings model of LFW, in making undetermined free-willed choices between two courses of action, A or B, the reasons for choosing each option don’t come with pre-established evaluative weights. During the process of deliberation, the agent assigns weight in an undetermined way to the reasons for each option and this typically leads to the choice of the option that has been assigned a greater value. In the paper, it is not only argued that this theory can resolve worries about luck but also that the view has virtues which certain other libertarian views lack, as such the view is contrasted with the modest libertarian models of LFW from Daniel Dennett (1978) and Alfred Mele (1995) as well as Robert Kane’s view (1996, 2002, 2007, 2011, 2019).

In recent work, I defend an indeterministic weightings model of libertarian free will. (Lemos, 2018, Ch. 5; 2021; 2023, Ch. 6). On this view, basic free-willed actions are understood as the result of causally indeterminate deliberative processes in which the agent assigns evaluative weight to the reasons for the different choice options under consideration. In basic free-willed actions, the assignment of weights is causally undetermined, and the choices are typically the causal consequence of these assignments of weights in which the choice option that is more highly valued is the choice option selected. In a recent article, Ishtiyaque Haji (2022) criticizes my view, arguing that it: (a) does not resolve worries about luck and (b) does not make coherent sense of the freedom of the weightings involved in free choices. I argue that his criticisms are based on misunderstandings of my position.

John Lemos defends an indeterministic weightings model of libertarian free will that is a variant of event-causal libertarian views. Many argue that these views are susceptible to the luck problem: an agent’s directly free choices are too luck infected for the agent to be morally responsible for them. The weightings model supposedly escapes this problem largely because in this model an agent’s reasons for choices do not come with pre-established values. Rather, an agent performs intentional acts of weighting that contribute to the value she assigns to her reasons. Decisions that are consequences of weightings are, thus, under the agent’s control and not subject to luck. In a recent paper, I argued that despite its weighting component, Lemos’s model succumbs to the luck problem. Lemos rejoins that my criticisms are based on misunderstandings and confusions. I deflect the charge of misperception and explain why the weightings model remains susceptible to the luck problem.