Paul M. Pietroski, Causing Actions.
The following assumptions are necessary to get the contemporary problem of mental causation off the ground:
(1) Non-Identity of the Mental and the Physical
(2) Causal Closure of the Physical
(3) Causal Exclusion
(4) Causal Relevance of the Mental
To many, especially those who flirt with Non-Reductive Physicalism (NRP), each of these is prima facie plausible, yet together they entail a contradiction. Or so it seems. Recently, it has been argued that another assumption is needed for a full-fledged contradiction, which we call
(5) Homogeneity of Mental and Physical Causation.
(5) says that the causal relevance of the mental, just as with the physical, involves a natural, non-projected causal sufficiency. We can summarize the central dilemma for NRP thus: If you are determined to accommodate mental causation while keeping (1) and (2), then either you reject (3) and accept systematic causal overdetermination (you double-count), or you reject (5) and claim there are two distinct metaphysical templates for causal relations (you double-think).
There are alternatives to this dilemma. Apart from the extreme move of rejecting (4), one may renounce (1) in favor of (multiple, local) type-identities, or deny (2) and adopt some robust version of (Substance- or Property-) Dualism. But serious doubts have been cast on the stability of positions that try to hold on to all of (1), (2), and (4).
A puzzling feature of Causing Actions is that its author methodically walks himself into a position that is less stable than NRP while accepting something approximating the union of the drawbacks of it and its competitors. Pietroski embraces an ontologically robust form of (1)--a Strawsonian personal dualism, with a token-dualism of Davidsonian event particulars. He assents to (4) by contending that distinctively mental trying-events bring about bodily behavior. However, he emphatically disavows Cartesian single-chain interactionism and is unwilling to abandon (2). This creates the first (nonstandard) tension in his book: trying to balance event dualism with causal closure, Pietroski claims that mental events globally supervene on physical events (GSV). Usually, supervenience is a byproduct of a fundamental conviction that the physical drives everything. (NRP notoriously combines a strict token-identity of the mental and the physical with only a very abstract kind of type-dualism.) But why should the unabashed dualist accept (2) and GSV, or claim that mental causes cannot make a nonredundant impact on the world (151)?
Pietroski recognizes that something is amiss when he concedes that he owes us an explanation as to why GSV holds. Adopting a Kripke/Stalnaker-account, he argues that possible worlds are individuated in terms of how their basic objects are arranged, together with a stipulation that only the physical is basic (6.3.2). Although we agree that GSV follows, that can hardly be considered an illuminating answer, rather than a mere restatement of the explanandum. And we should be especially suspicious here given the avowal that persons are ontologically primitive particulars (165).
Granting that (1), (2), and (4) can be made to fit together somehow, the second (standard-NRP) tension kicks in: double-count or double-think? Pietroski's predicament here differs from those of others with less ambitious ontological inclinations. Like Strawson, McDowell, and others, he recognizes two apparently independent frameworks: a normative realm of reasons, in which persons are considered as free and rational subjects of their actions, and a physical realm of causally closed mere happenings. Unlike these other authors, he wants to be more robust about this distinction: although the two realms don't intrude into one another, they are both causal. Moreover, he inclines (237f.) strongly toward an anti-projectivist, metaphysical oomph-theory of event causation (MOVE). Causal relations are purely extensional, independent of the way they are described (104ff.). So the question arises: how to avoid the odd consequence that our overt bodily behavior is constantly over-oomphed, contradicting (3)? Answer: deny that causation is a single, natural relation and argue that there are distinct, ontologically incommensurate levels of nomological explanation. We have the standard laws of the physical sciences and psychophysical, ceteris paribus laws yielding intentional explanations. Both kinds of explanation license singular claims about the causal sufficiency of an event. So although we don't double-count, each behavioral event is double-explained. Hence, Pietroski's explanatory compatibilism forces us to double-think after all--to accept, in Pietroski's terms (243), that causation is itself multiply realizable. MOVE for physical causes; supervenient causation, presumably cashed out in counterfactuals, for rationalizing causes. That is familiar from NRP: if we double-think, we don't have to double-count (244).
So, there can never be any nonredundant mental downward causation in the strong, impersonal MOVE-sense. Given Pietroski's strong commitments to (2), (3), and MOVE, however, isn't this tantamount to epiphenomenalism? Suppose the following are links in the supervenient (mental) causal chain: S [right arrow] M [right arrow] T [right arrow] B. The link between stimulus S and mental event in is oomphed (impersonal). But the link between M, trying-event T, and bodily behavior B is of a rationalizing kind only. Labels aside, are there reasons to claim that rational explainers play an equally natural causal role? To answer this, it seems, we need to know exactly what rationalizing causes are. Pietroski is less than ideally perspicacious here, and in any case does little to dispel a natural skepticism toward the idea of two distinct, incommensurate kinds of explanations, both of which are sufficiently strong to license singular causal claims, without any breach of (2) or (3). (If Pietroski were simply to deny that reasons are causes, then our contemporary problem would, of course, not arise, since we have gotten rid of (4). On the other hand, if Pietroski were to deny in Fodorian fashion that there are any interesting differences between intentional explanations and hedged physical laws (111), then he jeopardizes his project of personal dualism, including the freedom to choose one's own actions.)
Finally, let us briefly probe Pietroski's discussion of the causal role of states, and its potential results for MOVE (98-101). Endorsing a distinction between events and states, and opposing a reification of the latter as state tokens, he contemplates the idea of construing states in terms of properties, gesturing towards Shoemaker's notion of properties as sets of dispositional causal powers (99). A natural move from here would be to strengthen MOVE by distinguishing genuine from spurious events, based on the kind of properties involved. Based on a sparse property ontology encompassing only a few fundamental physical properties, MOVE would pick out and relate natural events whose occurrences make a causal difference. This more restrictive version of MOVE would explain why the basic physical sciences are able to yield a privileged description of the world and provide an objective basis for the projectibility of natural kind predicates in causal laws. It also reveals a clear rationale for rejecting (2) and GSV, if one is serious about augmenting one's ontology to include mental events. But Pietroski himself is ambivalent, wanting his theory to comport equally with Davidsonian Nominalism. This, of course, makes it difficult to find a causal role for properties, let alone to ground naturalness. And in fact, Pietroski denies that causal claims can be sensitive to type-descriptions (96f.). Lacking a robust conception of properties, his proposed sufficient condition for event causation in terms of explanation (basically ceteris paribus laws, notably involving those couched in intentional terms) must remain circular, as he himself admits (219). Barring projectivism, ceteris paribus laws are explanatory only if they relate events as cause and effect. Hence, relying on ceteris paribus laws to launch a defense of event dualism (3.2) presumably has little force against a physicalist, while it is dispensable for one committed at the outset to Strawsonian personal dualism (220-34).
Let us take stock. Pietroski's commitment with respect to (1) is considerably higher than your usual NRP-type dualism, but the benefit is the same: mental causation is possible only as ersatz causation. That is clearly a negative payoff. Why even put dualism in, if ersatzism is all you can get out? Conversely, if you are a dualist, why downplay the causal relevance of the mental, salvaging (2) while rejecting (5)? Despite many intriguing suggestions and lines of thought throughout Pietroski's dense and varied discussion, Causing Actions disappoints through its failure to effectively answer these challenges.
GEORG THEINER and TIMOTHY O'CONNOR
|Printer friendly Cite/link Email Feedback|
|Author:||Theiner, Georg; O'Connor, Timothy|
|Publication:||The Philosophical Review|
|Article Type:||Book Review|
|Date:||Apr 1, 2002|
|Previous Article:||Christopher Hookway, Truth, Rationality, and Self-Control: Themes from Peirce.|
|Next Article:||R.J. Hankinson, Cause and Explanation in Ancient Greek Thought.|