Naturalism and libertarian agency STEWART GOETZ
Naturalism and material objects 123
majority of sperm don’t manage to fertilize an egg; the lucky few that do can’t properly be accused of failure to function properly, on the grounds that they do things not done by their colleagues. Most baby turtles never reach adulthood; those that do are not on that account dysfunctional. Obviously you can function properly even if you don’t function the way most of your peers do.26
I take it that these examples do indeed show that statistical normality cannot be the whole story about proper system-function (and perhaps that is all they are intended to show). But none directly show that it is not the case that a thing X has F as its proper part-function just in case X belongs to a kind whose members statistically normally have F as their function. The reason is that none of Plantinga’s examples seem to provide us with an X that satisfies the right-hand side of the biconditional. Hands, no doubt, have functions, but it is not at all obvious that having a thumb is among them.
Sperm might also have various functions, but, again, we might well doubt that fertilizing eggs is among them. As I noted earlier, function ascriptions are system-relative; thus, to say that sperm have the function of fertilizing eggs is to say that there is some larger system of which the sperm are parts and in which they have the function of fertilizing eggs. But what system might that be? Certainly not the male organism, for sperm do not fertilize eggs in the male organism. But not the female organism either, for sperm are not parts of a female. Perhaps there is such a thing as Shakespeare’s beast with two backs, and perhaps that counts as a system of which sperm are parts and relative to which they have the function of fertilizing eggs. But such a system is not a biological organism, and it is not obvious that there are biological systems other than organisms; hence it is not clear that fertilizing eggs can count as a biological function of sperm, and many will doubt that sperm, as biological objects, can have any functions but biological ones. And, of course, similar reasoning will strongly suggest that cats and baby turtles lack functions as well.
Still, Plantinga’s examples do have indirect bearing on SN-accounts of proper part-function. They point out that there is no reason to think that what is statistically normal is metaphysically privileged in any way. Most male cats are neutered, but it does not follow from this that being neutered is a metaphysically privileged way of existing for a male cat, or that male cats are supposed to be neutered in any deep and interesting sense of “supposed to.” Similarly, then, it is hard to see why an object’s belonging to a kind whose members normally have F as their function should imply that it is, in some metaphysically important sense, supposed to have F as its function. The proper function of a human optic nerve is to transmit information to the visual cortex.
But this is not because it is statistically normal that it do so. If it became statistically abnormal for optic nerves to transmit information to the visual cortex, then malfunction of the optic nerve would be statistically normal. It would not be the case that optic nerves were no longer supposed to play the role in vision that they in fact
play. Furthermore, if it became statistically normal for optic nerves to produce sensations of pleasure that stimulate philosophical creativity, it would not follow from this that producing such sensations is what optic nerves are supposed to do.
If this is right, then SN-accounts of proper function will not do for our purposes.
But what of etiological analyses? We should note first that what initially suggests itself as the most intuitive etiological analysis is also most likely to be unacceptable to a naturalist. The analysis I have in mind is this: the (or a) proper function of X is to F just in case X was designed to F. Though this analysis is quite intuitive, it commits naturalists to thinking that natural objects, one and all, lack proper functions (again, assuming that the methods of science alone do not reveal the existence of a cosmic designer). Thus, insofar as the naturalist will want a unified account of proper function for both natural objects and artifacts, and insofar as she will want to say that at least some natural objects have proper functions, she is committed to saying that it is some sort of natural etiology that confers proper function – not an etiology that somehow involves a designer. As it turns out, the etiology of this sort that is universally invoked by naturalistic defenders of proper function is reproductive, or evolutionary etiology. Thus, whereas no one explicitly defends an SN-account of proper function, everyone who explicitly defends a naturalistically acceptable account of proper function defends a broadly evolutionary account. I will discuss only one such account
– one proposed and defended independently by Karen Neander and Ruth Millikan.27
This is not the only account of proper function in the literature, but it is the most prominent and the objections to it carry over to the others.
The full statement of Millikan’s account of proper function occupies two chapters in Language, Thought, and other Biological Categories. But in her “In Defense of Proper Functions,” we find the following summary of her view: Putting things very roughly, for an item A to have a function F as a “proper function”, it is necessary (and close to sufficient) that one of these two conditions should hold. (1) A originated as a “reproduction” (to give one example, as a copy, or a copy of a copy) of some prior item or items that, due in part to possession of the properties reproduced, have actually performed F in the past, and A exists because (causally historically because) of this or these performances. (2) A originated as the product of some prior device that, given its circumstances, had performance of F as a proper function and that, under those circumstances, normally causes F to be performed by means of producing an item like A. Items that fall under condition (2) have “derived proper functions”, functions derived from the devices that produce them.28
Though this statement does not include all of the subtleties of her more precise statement of the view, it will suffice for present purposes. Neander’s account, on the
Naturalism and material objects 125
other hand, is much more simply stated. She writes (under the assumption that the unit of selection in natural selection is a genotype): It is the proper function of an item X of an organism O to do that which items of X’s type did to contribute to the inclusive fitness of O’s ancestors, and which caused the genotype, of which X is the phenotypic expression, to be selected by natural selection.29
Before turning to objections, it is important to note that both Millikan and Neander deny that their theories are intended to analyze the ordinary concept of proper function.30 Millikan eschews analysis altogether, claiming that she is giving a
“theoretical definition” of a “technical term” that is of interest because of its usefulness in solving certain problems and not because of any relation it bears to ordinary notions of function or purpose.31 Neander, on the other hand, embraces analysis but (a) denies that giving an analysis involves giving necessary and sufficient conditions and (b) claims that the concept of proper function that she is analyzing is a technical concept in biology, not our ordinary concept.32 Despite initial appearances, then, it is doubtful that either Millikan or Neander means to be giving an account of proper function as it is understood in the present chapter. Thus, the objections I shall be raising should not at all be taken as objections to the projects that Millikan and Neander intended to engage in. Rather, they should be taken simply as objections to the idea that their accounts of proper function in their sense(s) of the term could serve as adequate accounts of proper function in my sense of the term.
That the Millikan/Neander account will not do as an account of proper function in the ordinary sense can be seen by considering the now familiar example of reproductive clay crystals. In The Blind Watchmaker, Richard Dawkins explains how natural selection could operate on a population of clay crystals.33 He invites us to imagine a kind of clay that improves its own chances of being deposited by damming up streams. The streams form shallow pools which then dry up. The clay dries and is blown away as dust, only to be deposited in other streams. The new crystals reproduce themselves and dam up their respective streams. And the process begins anew. Here, we have all of the elements of natural selection: the crystals reproduce, display random variation in the traits they possess, pass on traits to their descendants, and sometimes display traits that are either adaptive or maladaptive. But, as Mark Bedau also indicates in his own discussion of this example, no one would want to say that there is any metaphysically important sense in which the crystals are supposed to dam up streams.34
The reproductive history of an item makes no difference with respect to what it is supposed to do.
Granted, crystals are not biological organisms, so strictly speaking, though we do have a counter-example to Millikan’s account (construed as an account of proper
function in the ordinary sense), we do not have a counter-example to Neander’s. Still, the example casts doubt on Neander’s account. If, as we all want to say, the reproductive history of a population of clay crystals doesn’t tell us anything at all about how those crystals are supposed to behave, why should we think that the reproductive history of a population of biological entities tells us anything at all about how those biological entities are supposed to behave? There seems to be no relevant difference between the two cases.
We might put this point another way: Suppose we concede (as seems correct) that at least Neander’s view provides a materially sufficient condition for something’s having a proper function in the ordinary sense. That is, perhaps it is in fact the case that everything that satisfies her account also has a proper function. Still, that doesn’t guarantee that her account is adequate, for it doesn’t guarantee that there is any metaphysically or epistemically important connection between the properties by virtue of which something satisfies her account and the properties by virtue of which it counts as having a proper function. Furthermore, there is reason for thinking that in fact there is no important connection between the two sorts of properties. One reason is just the fact that clay crystals can exemplify many of the most important properties (namely, those involved in a natural-selective history) that Neander thinks guarantee proper function without themselves having a proper function. Another reason is that we can imagine cases where proper function and reproductive history might diverge. For example, we might imagine a world in which an incompetent deity creates biological organisms that evolve in ways contrary to the intentions of that deity.
In such a case, it seems clear that whatever we are inclined to say about what (if anything) counts as proper functioning for such organisms, we will not say that there is an objective, metaphysically privileged sense in which the organisms are supposed to function in the way that they evolved to function. Perhaps we will say that they are supposed to function in the way they were designed to function, or perhaps we will say that in this case there is no such thing as proper function for the organisms in question. But we will not say that the evolutionarily determined way of functioning is the objectively proper way.
It is trendy now to object to imaginary examples simply by appeal to the popular idea that conceptual analysis is a misguided project. But I find the objection unpersuasive. Even if conceptual analysis is a misguided project, I see no reason to doubt the evidential value of imaginary examples, especially when the examples are not highly contrived. But for those who insist on rejecting imaginary examples, we may also provide real examples where reproductive history and proper function seem to diverge, thus casting doubt on the idea that Neander has even provided a materially sufficient conditions for something’s having a proper function. Consider genetically engineered organisms. Will it be the objective, metaphysically privileged, proper function of a structure altered by engineering to display the reproductively established trait, the engineered trait, or neither? If one of the latter two, then there
Naturalism and material objects 127
are cases where proper function and reproductive history do diverge and we have immediate counter-examples to Neander’s view. If the former – if engineering always produces malfunction – then questions arise about what we should say about cases where an engineered trait becomes reproductively established. It seems clear that in such cases we should not say that a genuine malfunction has become a proper function. But if we should not say that, then, again, there are (or could be) cases where reproductive history and proper function diverge.
It should be clear that the problem here is quite general. Etiology might guarantee a kind of weak normativity. It might guarantee that there is a weak sense in which a thing is supposed to perform a certain function. But etiology alone cannot guarantee any sort of strong, metaphysically privileged sense in which a thing is supposed to perform a certain function. It simply doesn’t follow from the fact that a thing has a certain kind of causal history that it is in some metaphysically important way supposed to do whatever it is that its causal history has determined it to do.
If this is right, then we seem forced to the conclusion that the methods of science alone cannot reveal that anything has a proper function. Science can reveal facts about statistical normality; it can also reveal facts about reproductive history. But, as we have seen, there is no obvious epistemic or metaphysical connection between such facts and facts about proper function. Thus, there is no reason to think that in revealing such facts, science reveals facts about proper function. Furthermore, there are no naturalistically acceptable alternatives – which is just to say that there are no other scientifically accessible facts that naturalists might reasonably take as indicative of proper function. Hence, science cannot reveal proper function.
So far, I have been assuming that the naturalist will want a unified account of proper function – one that will cover both natural objects and artifacts. But, strictly speaking, I have said nothing to rule out the possibility that (a) there are no biological proper functions but (b) some artifacts have proper functions by virtue of various intentions and goals had by their human designers. If this is right, then since human intentions and goals are accessible to science, some proper functions are accessible to science after all.
But there is a problem here. Intentions and desires are accessible to science only insofar as they are physical states of biological objects (i.e. brains or parts of brains) or perhaps artifacts (such as computers). But, as earlier arguments in this section showed, we need biological proper function to render belief in mind-independent biological objects naturalistically acceptable. What this means, then, is that whatever a naturalist thinks about intentions and desires, she cannot think that they are physical states of persisting things that have their persistence conditions independently of human mental activity. In other words, she must give up the commonsense view that mental activity can only take place in a persisting, mind-independent subject. That by itself is not an objection to the view, but it is a reason for thinking that our best science cannot find anything in nature that satisfies the commonsense conception of an
intention or a desire. But it is intentions and desires as commonly conceived that are widely seen as being able to confer proper functions. Who knows whether intentions and desires under the revised conception will plausibly be able to do so? Indeed, who knows whether our best science will even leave room for intentions or desires? Thus, at least in the absence of further information from the field of cognitive science, it seems clear that if it is naturalistically unacceptable to believe in biological proper functions (and hence mind-independent biological objects), it will be naturalistically unacceptable to believe in any proper functions. And this, of course, entails that belief in mind-independent material objects generally is naturalistically unacceptable.
I conclude, then, that (3.1) is false. But, again, apart from the truth of (3.1) I see no other way to explain how we acquire justified beliefs about classification short of saying that facts about how objects are to be classified depend importantly upon our mental activity. Thus, I conclude that we cannot acquire naturalistically acceptable beliefs about classification.
In this chapter, I have argued that belief in mind-independent material objects depends for its justification upon our ability to acquire the justified belief that modal properties are exemplified in at least some of the regions of space-time that we take to be occupied by material objects, and I have argued that beliefs of this latter sort depend for their justification upon our ability to acquire justified classificatory beliefs.
Furthermore, I have shown that the most promising suggestion for how a naturalist might be able to be justified in classificatory judgments is a failure, and I have observed that if this suggestion fails, there is good reason to believe that any other will fail as well.
Thus, I conclude that belief in mind-independent material objects is incompatible with commitment to naturalism.
One might be tempted to respond by saying that non-naturalists are in a similar boat. It is very difficult for anyone to tell how we acquire justified classificatory beliefs; the problem is not peculiar to naturalism. However, there is a relevant difference between the position of a naturalist and the position of a non-naturalist with respect to questions about the justification of our classificatory beliefs. Non-naturalists are free to say that classificatory beliefs are justified a priori; and they can furthermore observe that the mystery of how such beliefs are justified is no different from the mystery of how any beliefs might be justified a priori. Naturalists, on the other hand, are not free to make this move. Naturalists are committed to thinking that classificatory beliefs admit of a particular kind of justification – namely, scientific. Thus, showing that there is good reason to doubt that such beliefs could be justified through the methods of science provides reason for naturalists to be skeptical where non-naturalists need not be skeptical.
As I indicated earlier, some naturalists will be content to abandon the ontology of mind-independent material objects. Some philosophers are attracted to stuff ontologies; and many of the same philosophers are attracted to either conventionalism or conceptualism. Other philosophers are happy with a
Naturalism and material objects 129
thoroughgoing agnosticism about ontology (which is, perhaps, the more appropriate stance in light of what I have argued about the status of our knowledge of the modal character of the world). Nevertheless, many have the strong conviction that it is much more reasonable than not to believe that there really are some things (e.g. people and plants) that last over time and that have their persistence conditions independently of human mental activity and semantic conventions. If the arguments in this chapter are sound, however, those who share this conviction ought to reject naturalism.
Notes
1 Work on this project was supported by a University of Delaware General University Research Grant and by a grant from the Pew Evangelical Scholars Program. I also have an embarrassingly long list of people to thank for advice. Fred Adams, Chris Boorse, Jeff Brower, Tom Crisp, Jeff Jordan, Alvin Plantinga, Bill Ramsey, David Silver, Jim Stone, and Dale Tuggy all read early drafts and provided many helpful comments. A version was also read at the Fourth Annual Mighty Midwestern Metaphysical Mayhem Conference at the University of Notre Dame. Conversations with the participants in that conference – particularly David Chalmers, Brie Gertler, John Hawthorne, Larry Lombard, Brian McLaughlin, and Laurie Paul (as well as others already mentioned) – were extremely helpful. Finally, I am especially grateful to Michael Bergmann, Trenton Merricks and Joel Pust, who were subjected to multiple drafts and offered many very useful suggestions. I would also like to thank Marty Strachan, my research assistant during the summer of 1998, for valuable information on the topic of conventionalism.
2 For example: D. M. Armstrong, “Naturalism, Materialism, and First Philosophy,” reprinted in Contemporary Materialism, ed. Paul Moser and J. D. Trout (London: Routledge, 1995), 35–
47.
3 For example: Michael Devitt, Coming to Our Senses (Cambridge, Mass.: Cambridge University Press, 1996); W. V. Quine, “Naturalism; Or, Living Within One’s Means,” Dialecitca 49 (1995): 251–61; W. V. Quine, “Epistemology Naturalized,” in Ontological Relativity and Other Essays (Cambridge, Mass.: Harvard University Press, 1969), 69–90; and Alex Rosenberg, “A Field Guide to Recent Species of Naturalism,” British Journal for the Philosophy of Science 47 (1996): 1–29.
4 For example: Jerrold Katz, Realistic Rationalism (Cambridge, Mass.: MIT Press, 1998); Jean Hampton, The Authority of Reason (New York: Cambridge University Press, 1998); and Barry Stroud, “The Charm of Naturalism,” Proceedings and Addresses of the American Philosophical Association 70 (1996): 43–56.
5 W. V Quine, “On What There Is,” in From a Logical Point of View (Cambridge, Mass.: Harvard University Press, 1961).
6 Quine, “Naturalism; Or, Living Within One’s Means”; Michael Ruse, Evolutionary Naturalism (London: Routledge, 1995); Mark Bedau, “Naturalism and Teleology,” 23–51 in S. Warner and R. Wagner, eds., Naturalism: A Critical Appraisal (Notre Dame: University of Notre Dame Press, 1993).
7 Armstrong, “Naturalism, Materialism, and First Philosophy”; Hilary Kornblith, “Naturalism: Both Metaphysical and Epistemological,” 39–52 in P. French, T. Uehling and H. Wettstein
(eds.), Midwest Studies in Philosophy, Volume 19, Philosophical Naturalism (Notre Dame: University of Notre Dame Press, 1994).
8 John Dewey, Sidney Hook, and Ernest Nagel, “Are Naturalists Materialists?” Journal of Philosophy 42 (1945): 515–30; William Dennes, “The Categories of Naturalism,” 270–94 in Y.
Krikorian (ed.), Naturalism and the Human Spirit (New York: Columbia University Press, 1944).
9 Wilfrid Sellars, “Empiricism and the Philosophy of Mind,” in Science, Perception, and Reality (London: Routledge & Kegan Paul, 1963), 173.
10 Michael Rea, World Without Design: The Ontological Consequences of Naturalism (manuscript in preparation). This view of naturalism is also accepted by Peter Forrest. See Peter Forrest, God Without the Supernatural (Ithaca, N.Y: Cornell University Press, 1996), 89.
11 Note that if facts about what changes a thing can and cannot survive depend upon God’s beliefs or attitudes, or upon the beliefs and attitudes of some other non-human being, that thing will still count as mind-independent as I am using the term. Hence the imprecision.
But the term is convenient, and the problems are negligible in the present context, so I will continue to use the term in the way just described.
12 See, for example, Thomas Blackson, “The Stuff of Conventionalism,” Philosohpical Studies 68
(1992): 65–81; Michael Jubien, Ontology, Modality, and the Fallacy of Reference, (Cambridge: Cambridge University Press, 1993); and Alan Sidelle, Necessity Essence and Individuation (Ithaca, N.Y: Cornell Unviersity Press, 1989). Arda Denkel has recently argued that it is impossible to have matter without objects (Arda Denkel, “Matter and Objecthood,” Dialogue 28 (1989): 3–16). But see also the replies by Laycock and Sidelle (Henry Laycock, “Matter and Objecthood Disentangled,” Dialogue 28 (1989): 17–21; and Alan Sidelle, “Formed Matter Without Objects: A Reply to Denkel,” Dialogue 30 (1991): 163–71.) 13 Rudolf C arnap and Allan Gibbard both seem to be conceptualists. See Rudolf C arnap, Meaning and Necessity (Chicago, Ill.: University of Chicago Press, 1947) and Allan Gibbard,
“Contingent Identity,” reprinted in Michael Rea (ed.), Material Constitution: A Reader (Lanham, Md.: Rowman & Littlefield, 1997). Alan Sidelle defends in detail a version of conventionalism independently described but not defended by Paul Coppock. See Paul Coppock, Review of Reference and Essence by N. Salmon, Journal of Philosophy 81 (1984): 261–
70 and Sidelle, Necessity, Essence, and Individuation.
14 But see William Dembski, “Naturalism and Design” in this volume.
15 By “modal realism” I just mean the denial of modal anti-realism as it is characterized above.
Modal realism is not to be understood as the thesis (endorsed by David Lewis and others) that possible worlds are concrete objects spatiotemporally disconnected from our world.
16 Michael Rea, “The Problem of Material Constitution,” Philosophical Review 104 (1995): 525–
52.
17 The example is from David Wiggins, “On Being in the Same Place at the Same Time,”
reprinted in Michael Rea (ed.), Material Constitution: A Reader.
18 Note that the kind of classification under discussion here is metaphysical rather than, say, biological or physical. Few people nowadays think that biological classifications are metaphysically informative. Note furthermore that, on this view, it is possible to be nominally a K without being classifiable as a K, but not vice versa. For a defense of my views about kind-membership, see my “Constitution and Kind Membership,” forthcoming in Philosophical Studies. Nothing of substance turns on the assumption that there are two ways of belonging to a kind; but it would be inordinately complicated at best to discuss the points that will follow
Naturalism and material objects 131
in completely theory-neutral terms.
19 Peter van Inwagen, Material Beings (Ithaca, N.Y.: Cornell University Press, 1990).
20 See, for example, John Dupré, The Disorder of Things (Cambridge, Mass.: Harvard University Press, 1993) and W. V Quine, “Whither Physical Objects?” pp. 497–504 in R. S. Cohen, P. K.
Feyerabend, and M. W. Wartofsky (eds.), Essays in Memory of Imre Lakatos (Dordrecht: D.
Reidel, 1976).
21 Crawford Elder defends views that are in some ways suggestive of this one (though not nearly enough to justify attributing the view to him). Elder maintains that the essential properties of a thing are learned by learning facts about its distinctive nature; he also maintains that the proper functions of things are essential to them (see Crawford Elder, “An Epistemological Defense of Realism about Necessity,” Philosophical Quarterly 42 (1992): 317–36, and Crawford Elder, “On the Reality of Medium-Sized Objects,” Philosophical Studies 83 (1996): 191–211).
If he is right, then perhaps one implication of his views is that we acquire justified classificatory beliefs by way of acquiring justified beliefs about proper function. However, I hesitate to attribute this sort of view to Elder since (a) he doesn’t explicitly defend it, (b) both of his papers presuppose that we can make classificatory judgments in arguing for their respective conclusions, and (c) the second of the two papers also presupposes that modal intuitions (such as intuitions about the essentiality of origins) can help to establish facts about the essential properties of material objects.
22 I assume that systems are wholes; hence the objects “in” the system are just parts of it. Note that if function ascriptions are indeed system relative, then to say that some xs are functionally organized entails that they are parts of a whole. Thus, we have a quicker route to a conclusion – that being functionally organized is sufficient for being parts of a whole – that I defended on purely intuitive grounds on p. 354 of Michael Rea, “In Defense of Mereological Universalism,” Philosophy and Phenomenological Research 58 (1998): 347–60.
23 Here I assume that if A-facts supervene on B-facts, and if science can reveal A-facts, then it can also reveal B-facts. I do not assume the converse – that if science can reveal B-facts then it can also reveal A-facts. I also do not assume the general principle that if way of knowing W can reveal A-facts, and if A-facts supervene on B-facts, then way of knowing W can reveal B-facts.
The falsity of this general principle can readily be seen by substituting “introspection” for
“W”, “facts about mental phenomena” for “A-facts,” and “facts about brain-chemistry” for “B-facts” (I thank Trenton Merricks for this point).
24 This is in fact a rough summary of Larry Wright’s account of what it is for something to have a function (see Larry Wright, “Functions,” The Philosophical Review 82 (1973): 139–68. For counter-examples, see Christopher Boorse, “Wright on Functions,” The Philosophical Review 85 (1976): 70–86, and Elizabeth Prior, “What is Wrong with Etiological Accounts of Biological Function?” Pacific Philosophical Quarterly 66 (1985): 310–28). Nevertheless, many of those following Wright have adopted roughly this sort of view to explain what it is for something to have a proper function. In fact, most of those who write on proper functions tend to conflate the notions of function and proper function (see, for example, Ruth Millikan, “In Defense of Proper Functions,” Philosophy of Science 56 (1989): 288–302, especially p. 293; Karen Neander, “Functions as Selected Effects: The Conceptual Analyst’s Defense,”
Philosophy of Science 58 (1991): 168–84, esp. p. 173; and Michael Levin, “Plantinga on Functions and the Theory of Evolution,” Australasian Journal of Philosophy 75 (1997): 83–98).
25 Alvin Plantinga, Warrant and Proper Function (New York: Oxford University Press, 1993), 199.
Plantinga also claims that the analysis is suggested, though not explicitly stated, in John
Pollock’s “How to Build a Person,” in James Tomberlin, (ed.), Philosophical Perspectives, volume 1: Metaphysics (Atascadero, Calif.: Ridgeview, 1987), 109–54.
26 Plantinga, Warrant and Proper Function, 200–1.
27 Karen Neander, Abnormal Psychobiology (Ph.D. diss., La Trobe University, 1983); Neander,
“Functions as Selected Effects”; Ruth Millikan, Language, Thought, and Other Biological Categories (Cambridge, Mass.: MIT Press, 1984); Millikan, “In Defense of Proper Functions”.
Among those who extend their view to include artifacts are Paul Griffiths, “Functional Analysis and Proper Functions,” British Journal for the Philosophy of Science 44 (1993): 409–22; Crawford Elder, “On the Reality of Medium Sized Objects”; and Beth Preston, “Why is a Wing Like a Spoon? A Pluralist Theory of Functions,” Journal of Philosophy 95 (1998): 215–54.
28 Millikan, “In Defense of Proper Functions,” 288–9.
29 Neander, “Functions as Selected Effects,” 174.
30 Even noting this, however, does not entirely remove the possibility of being misunderstood.
For example, John Post accuses Alvin Plantinga of misreading Millikan; but, ironically, the accusation itself rests on a misreading of Plantinga (see John Post, “Critical Notice of Ruth Millikan’s White Queen Psychology and Other Essays for Alice,” Philosophy and Phenomenological Research 58 (1998): 233–7). Post claims that “Plantinga…badly misreads [Millikan] as attempting an analysis, then tries to counter-example accordingly” (233). In fact, however, it is Post, not Plantinga, who is guilty of poor exegesis. Plantinga never suggests that Millikan is attempting an analysis, and he explicitly acknowledges that “her project is not as it stands directly relevant to the question at issue: the question whether there is available a naturalistic understanding or analysis of proper function.” (Plantinga, Warrant and Proper Function, 201).
He goes on to say that “Millikan’s account…is subtle and challenging (and formidably difficult), and it is worth looking to see whether in fact it provides the materials for an adequate or accurate analysis of the notion of proper function” (202). But, of course, in light of his earlier remarks it is abundantly clear that Plantinga has not made the mistake of thinking that Millikan was actually attempting to provide “the materials for an adequate or accurate analysis of the notion of proper function.”
31 Millikan, Language, Thought, and Other Biological Categories, 17; “In Defense of Proper Functions,” 289–93.
32 Neander, “Functions as Selected Effects,” especially pp. 168–73.
33 Richard Dawkins, The Blind Watchmaker (New York: W. W. Norton & Company, 1996), 150–4.
34 Bedau, “Naturalism and Teleology,” 36–42.
Charles Taliaferro
J. L. Mackie developed an argument against moral realism he called “the argument from queerness.”1 He contended that, given a materialist understanding of the cosmos, the existence of objective values would be queer or, to use less colorful language, objective values would be unique and unlikely. “If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe.”2 Mackie did not conclude that the strange and unlikely has happened, but instead that there are no objective values. The argument has such currency that it is even an entry in the Penguin Dictionary of Philosophy under the heading “queerness.” The argument runs along the same lines as Mackie’s reasoning in The Miracle of Theism. Following Hume, Mackie argues that theism is so unlikely it seems like a miracle that anyone believes it; the existence of God would be “queer” given everything else we know about the cosmos, and thus it is best denied. A yet further, similar argument has been advanced by many philosophers about consciousness.
Some philosophers contend that the existence of consciousness is itself an entity, quality or relation of a very strange sort, a reality that seems utterly different from anything else in the universe. While the universe is fundamentally physical in its origin and constitution, consciousness seems so radically different that we are disposed to think of it as something deeply mysterious, perhaps even something non-physical.
Given a general physicalist outlook, contemporary philosophers are reluctant to posit consciousness as some elusive, immaterial object. An overriding physicalist framework leads philosophers either to eliminate consciousness and the mental, lest these entities become untamed emergent items, or it leads philosophers to acknowledge consciousness and related mental states, but to treat these as physical.
Paul Churchland describes the prevailing physicalist orientation of much philosophy of mind:
Most scientists and philosophers would cite the presumed fact that humans have their origins in 4.5 billion years of purely chemical and biological evolution as a weighty consideration in favor of expecting mental phenomena to be nothing
but a particularly exquisite articulation of the basic properties of matter and energy.3
This observation fits with Jaegwon Kim’s estimation of the current field. Kim points out that the goal for many philosophers is to accommodate physicalist, background assumptions about the nature of the world while not explaining away the mental altogether and eclipsing vital, human values:
The shared project of the majority of those who have worked on the mind–body problem over the past few decades has been to find a way of accommodating the mental within a principled physicalist scheme, while at the same time preserving it as something distinctive – that is, without losing what we value, or find special, in our nature as creatures with minds.4
The goal is either to accommodate consciousness, minds, values et al. as denizens of a physical world or to eliminate them. Either way, materialism is the order of the day. As William Lycan writes in Consciousness and Experience, “Few theorists question the eventual truth of materialism.”5
Naturalism admits of a range of definitions and scales. There are, for example, strong versus weak naturalists, radical and agnostic naturalists, and so on. Sometimes naturalism is simply another name for a thoroughgoing form of physicalism, while at other times naturalism is simply any view of the world that is incompatible with supernaturalism. In a volume dedicated to raising difficulties for naturalism, I shall point to the problem of incorporating the mental in a physicalist schema, and I shall then endeavor to clear the ground for a view of human nature that is resisted by all physicalists and virtually all naturalists. If this non-materialist, non-naturalist alternative may resolve the problems facing materialism and naturalism, and if this alternative may be seen to have some overriding plausibility, then this essay will have done its part in the overall case against naturalism. The non-materialist, non-naturalist philosophy of mind I shall defend is a form of mind–body dualism. Daniel Dennett’s comment on dualism captures the prevalent, naturalist disdain toward dualism. In Kinds of Minds he describes dualism as the “view that minds are composed of some nonphysical and ultimately mysterious stuff . . . relegated to the trash heap of history, along with alchemy and astrology.”6 My case for dualism will not lean on alchemy and astrology, but it will be in concert with philosophical theism, as sketched in a final section of this paper.
Not long ago, philosophers did not believe that a person is “a particularly exquisite articulation of the basic properties of matter and energy.” It was commonplace, rather, to believe that we live in a theistic cosmos and that human beings are an exquisite articulation of the Creator of the cosmos. A range of philosophers today link the case
for dualism and theism.7 I shall take this link seriously. Essentially, my goal is to develop a position similar to George Mavrodes’ in his reply to Mackie’s argument from queerness. In “Religion and the Queerness of Morality”8 Mavrodes argues that objective values look less “queer” if seen against a theistic back-drop than they do if seen in light of Mackie’s naturalism. Following Mavrodes’ lead, I contend that the case for dualism is far more promising once debate in philosophy of mind is carried out in a broader metaphysical context.
Problems with eliminative materialism
Eliminative materialism is not as popular as it was some decades ago. A major problem has been the task of developing a version of eliminative materialism that is not self-refuting or self-contradictory. Some eliminativists appear to be in the unenviable position of claiming to believe that there are no beliefs. Another difficulty is the problem of being able to accommodate human reasoning. A further worry still is that eliminativism is flatly refuted by experience.9
The next section of this paper will consider the problem of treating consciousness and experience within a form of reductive or identity materialism. Insofar as experience does have the characteristics I shall argue that it does, this counts as a problem for eliminativism as well. But here I wish to consider a problem with the way eliminativism is articulated in the literature. Eliminativism has been advanced in a fashion that appears to leave little unsettled in our everyday discourse and our understanding of social relations and the like. I believe this is deeply problematic.
Consider, for example, Paul Churchland’s The Engine of Reason, the Seat of the Soul.
At the outset, The Engine of Reason, the Seat of the Soul appears non-eliminative.
Churchland asks: “How does the brain work? How does it sustain a thinking, feeling, dreaming self?”10 This seems to involve a straightforward acknowledgment of thinking, feeling, dreaming, and of a unitary subject that engages in each. Indeed, some of Churchland’s depiction of the findings of the brain sciences seem compatible with a philosophy of mind and body that is either a form of non-eliminative materialism or even dualism:
[W]e are now in a position to explain how our vivid sensory experience arises in the sensory cortex of our brains: how the smell of baking bread, the sound of an oboe, the taste of a peach, and the color of a sunrise are all embodied in a vast chorus of neural activity.11
Reference to how a psychological phenomena “arises” suggests a distinction between the psychological and mental, and talk of “embodiment” seems as natural for noneliminative materialists as it is for dualists who uphold the metaphysical
distinction between the mental and physical and yet insist upon their causal interaction. But this open-ended description of the mind and body is soon put aside.
Churchland has elsewhere construed dualism as “mankind’s official view of itself ”
and so it is often Churchland’s foil in the course of his arguments for a thorough materialism.12 But Churchland challenges not only dualism, but non-reductive and identity materialism as well. One of the themes of Churchland’s book is that our customary theory or working assumption about ourselves needs to be radically revised in light of the natural sciences. Our “official” and unofficial, ordinary conceptions of ourselves is mistaken:
Is our basic conception of human cognition and agency yet another myth, moderately useful in the past perhaps, yet false at edge or core? Will a proper theory of brain function present a significantly different or incompatible portrait of human nature? . . . I am inclined toward positive answers to all of these questions.13
The challenge to folk psychology comes in stages. At certain points, a unitary subject distinct from the brain is the target and the properties that would customarily be attributed to this subject are attributed to parts of the brain. Churchland comments on vision:
But who . . . can be watching this pixilated show? The answer is straightforward: no one. There is no distinct “self” in there, beyond the brain as a whole. On the other hand, almost every part of the brain is being “watched” by some other part of the brain, often by several other parts at once.14
But then the claim that we folk customarily watch things – when this is interpreted in terms of sensations, concepts, and beliefs – is itself critiqued.
Churchland assumes that the average reader, or newcomer to the field, is not an eliminativist when it comes to sensations, beliefs, perceptions, desires and preferences:
You came to this book assuming that the basic units of human cognition are states such as thoughts, beliefs, perceptions, desires, and preferences. That assumption is natural enough: it is built into the vocabulary of every natural language. . . .
These assumptions are central elements in our standard conception of human cognitive activity, a conception often called “folk psychology” to acknowledge it as the common property of folks generally. Their universality notwithstanding, these bedrock assumptions are probably mistaken.15
This is a radical claim; it takes us some distance beyond the customary attack on dualism. Churchland claims that the disposal of such folk “bedrock assumptions” and the building up of an alternative, scientifically-informed understanding of human nature will lead us to new insights into moral life. New scientific discoveries can “set us free, and allow us to achieve a still higher level of moral insight and mutual care.”16 He endorses a form of moral realism and indicates his preference for an Aristotelian as opposed to Kantian ethic. Brain and physiological testing provide Churchland with evidence that moral judgments on specific cases comes prior to our grasping and applying moral rules.
Churchland proposes that a successful, eliminative account of persons will not leave us with some “cold” or soulless view of ourselves: One’s first impulse, perhaps, is to see the vocabulary and framework of a general theory of the brain as something alien and cold. But it will not be alien if it depicts all of us, at last, as we truly are… Whatever the distractions, we must continue to exercise our reason. And whatever the temptations we must continue to nurture our souls. That is why understanding the brain is so supremely important. It is the engine of reason. It is the seat of the soul.17
Churchland does not see that eliminativism will threaten “the soul” and, perhaps to put on display his settled view on the compatibility of eliminativism and his philosophy of personal relations, he includes several personal allusions in The Engine of Reason, the Seat of the Soul. There is, for example, a photograph of his daughter and her “soul-mate,” which is then analyzed in terms of retina-visual cortex interaction.18 And there is figure 7.1, an MRI image of Patricia Churchland’s brain.19 Paul Churchland comments: “This particular brain is in fact well known to me via more conventional informational pathways. It is the brain of my wife and colleague, Patricia Churchland, and it is very dear to me.”20
But regrettably, it is not very clear how Churchland can succeed both in advancing eliminativism and redeeming the promise that it will assist us in an effort to “nurture our souls” and care for one another. Obviously the brain sciences can assist us insofar as they enable us to treat physiological illness and the organic basis for psychological dysfunction. They can provide the non-eliminative philosopher with insights as to the material underpinnings of our mental life. But it is not clear how we can make use of notions life “assist,” “care” and so forth, if we must shed bedrock assumptions about thoughts, beliefs, perceptions, desires and preferences. In most, if not all cases, “care,”
“compassion,” and “love” appear to make sense only within a conceptual context of beliefs and desires. At a minimum, love between persons seems to involve beliefs about oneself, beliefs about the beloved, and some desire for one another’s well-being. In The Engine of Reason, the Seat of the Soul we are not given a clear guide as to how what, in
folk language, we refer to as “love,” “care,” “compassion,” and so on, can be given any purchase in an eliminative world.
Notwithstanding the pride of place that Churchland gives to the moral benefits of the position he defends, I think the overall, eliminative project gives rise to similar worries about whether the brain (or “enbrained body,” to use an awkward phrase) as Churchland describes it can indeed function as an engine for reason. In my view, it is difficult to conceive of how any reasoning can survive in an eliminativist framework.
In a world that eliminates the intentional world of folk psychology, how can one make sense of reasoning? In a standard case of reasoning, one accepts a conclusion in virtue of believing the premises of an argument and acting on one’s grasp of the nature of entailment. Jane Doe accepts conclusion B, because she believes that ‘If A then B’ and she believes there is compelling evidence that A. I am skeptical about whether reasoning can be given any intelligible place in our view of what we do when we undertake philosophical reflection (or any other norm-governed reflection in which conclusions are derived on the basis of beliefs) if one does away with beliefs – or, if one acknowledges the existence of beliefs and yet gives them no role to play in a causal explanation of why, say, Jane Doe reasons as she does. Thus, even when Churchland permits (if only as by way of marking a starting point for further analysis) attributions of “beliefs” in describing our activities, I contend that these beliefs are not given a sufficient role to describe and explain (what at least appears to be) our reflection – our coming to hold certain beliefs on the basis of others. Here, I think Churchland’s privileging of physics and chemistry causes problems. He is careful to delimit causal accounts to physics and chemistry:
I do not mean to suggest that social properties are anything more, ultimately, than just intricate aspects of the purely physical world. Nor do I wish to suggest that they are independent causal properties over and above what is captured by physics and chemistry.21
But unless physics and chemistry include the stuff of reason and beliefs (which, at least in conventional versions of physics and chemistry, they do not), I do not see how Churchland can do justice to our capacities to reason and argue rationally.22
I cited Churchland’s claim earlier that a general theory of the brain could not be
“alien and cold” so long as “it depicts all of us, at last, as we truly are.” But the question that needs to be raised is the prior one of whether eliminativism can depict us. For what account may we have of personal identity without taking seriously our beliefs, desires, hopes, loves, and the like? If Churchland is right that a true, general theory of the brain cannot be alien and cold, and the theory of the brain that is at the heart of Churchland’s eliminativism seems to leave us in alien and cold waters, then we have reason to think eliminativism is mistaken.23
Assuming there is some reason to recognize the existence of beliefs, desires, and other “folk” items, what are the prospects of identifying them as no more than physical processes and states?
Problems with identity materialism
Condensing the problem with materialism into the simplest terms, the chief obstacle to identifying a person’s mental life with their body or bodily parts and processes, is that they do not appear to be identical. Identifying the two seems to cut directly against what we (at least seem to) know of our experience and our bodies. On the face of it, conceiving of some mental state like being in pain or being puzzled philosophically is radically different from conceiving of being in whatever physical states that are posited or described by materialists. It appears that one may conceive of the mental without the physical, and that conceiving of one does not involve or entail the other. The problem is not that the mental and physical appear to be different given our commonplace theory about human nature, but that their apparent difference constitutes a reason for adopting a theory of human nature that does not collapse the two into one thing or kind of thing. As Colin McGinn writes “Consciousness certainly seems quite different from a mere brain process. My hearing of a loud bang, say, presents itself as a different kind of thing from electrical activity in a certain part of my brain. And thinking about a trip to the beach does not feel like the spiking of innumerable neurons in my cortex.”24
The rationale for differentiating the mental and physical relies upon a widespread method for differentiating all cases of objects and properties. At the very foothills of cognitive development one learns to differentiate objects and states of ourselves by noting how objects may be conceptually identified independently of one another.
One does not differentiate oneself from one’s care-giver, for example, simply by undergoing an array of sensory experiences, but by conceptually grasping that conceiving of him/her is not the same as conceiving of oneself. It is this conceptual power which, when wedded to visual and other sensory experiences, allows us to perceive the relevant differences.25 And, a bit later on in our cognitive life, it is because I can conceive of the property of Being a philosophy professor without thereby conceiving of the property of Being the President of the United States that I recognize these as distinct properties. If someday I discover that there is someone who is both a philosophy professor and the President, I do not thereby discover that the properties are identical but I discover instead that the same person is lucky enough to have both properties.26
A related, but I think unsatisfactory, way of putting the above point about differentiating properties (or things in general) is that properties are differentiated when they are describable in terms that are not synonymous. It is because “Being a philosophy professor” and “Being a President of the United States” mean different things that we differentiate professors and presidents. But I believe the descriptions
mean different things because of our power to conceive of the one property without the other; we do not conceive of them differently because of the difference in the meanings of the terms describing them. Besides, the power to differentiate properties and things based on our conceptual powers may operate without descriptive terms. I believe that X (some thing to my right that I have no term for) is different from Fred (where imagine “Fred” names someone but the proper name “Fred” has no descriptive content) because I can conceive of the one without the other.
The appeal to our conceptual powers in differentiating properties is not advanced here as a power that operates independent of our other powers to sense, perceive, reflect on, and remember things. Our ability to differentiate things works in concert with these other powers and is integral to them. And it is on the basis of extensive conceptual investigation that one discovers the relation between various properties, whether these are discoverable a priori (as I believe that some are) or a posteriori, as discovered in experience.
Dualists have advanced a host of thought experiments to capture the differences between the mental and physical. Examine someone’s brain states, their body as a whole and any of their parts, while the person is in certain mental states (e.g. he is in pain and also puzzled about philosophy). Does one thereby see any mental states? It appears that one does not. Examining pain as a feeling seems to be one thing and examining the relevant physical states seem to be two different things. The most famous, Cartesian strategy is to argue that the mental can be conceived of without conceiving of anything spatially extended and thus material, and vice versa. A popular, contemporary way of differentiating the mental and physical is by imagining that someone knows all that can be known about the physical world, but that the person lacks the ability to see color. Arguably, if she were to gain visual experience of color she would then grasp something true about the world (e.g. she would know what the color red looks like) that she would not know before.27 Arguably, this exhibits an important difference between the physical basis or cause of color and other sensory, mental states and those mental states themselves.
In opposing a form of identity materialism, I will not offer a refined analysis of the mental, systematically delimiting the extent and interrelationship of different mental properties and states. I will refer to the mental as a general realm that includes sensing, perceiving, and having intentional attitudes (hoping, loving, hating, remembering, et al. ). Some philosophers elect to refer to the mental in terms of consciousness (McGinn), experience (Strawson), “something it is like” (Nagel), phenomenally conscious states (Block), and qualia (Stubenberg). My principle goal is simply to make trouble for naturalism, and point the way to what I think is a more constructive account of human nature, not to carry out a substantial investigation of the mental and establish a uniform, orthodox system of categories. It will suffice to point out how difficult it is to accommodate some mental properties and states, and leave to another occasion the project of refining our treatment of the mental (e.g. addressing the
question of whether one should posit intending as a basic mental category or whether intending can be analyzed in terms of beliefs and desires). I shall adopt what Ned Block in a similar context has referred to as “a liberal terminological policy” .28
It may will be useful to consider the dualist appeal to experience in the context of weighing four materialist objections. I take these objections from work by Georges Rey, William Lycan and Paul Churchland. While an eliminativist, Churchland has also offered powerful anti-dualist arguments that would clear the way for identity materialism. Like the philosopher John Foster who is an idealist as well as a defender dualism as his next-best philosophy, Churchland defends the identity form of materialism even if his final preference is for eliminativism.
Referential opacity
Some materialists claim that the dualist appeal to experience runs aground because of its failure to recognize referential opacity. The same object may be referred to in different ways and in terms that are not synonymous, and yet there is only the one object being referred to. Consider Georges Rey’s analysis of a dualist argument. Rey considers the proposal that the mind or the mental seems to be so radically distinct from the physical that it has the property of being possibly unextended, that is, it might exist without thereby having spatial extension. But my body and bodily states do not seem to have this property. It seems absurd to think one’s body (or any physical thing whatever) might have the property of existing without spatial extension. He casts the argument as follows:
(1) My mind has the property of being conceivably unextended.
(2) My body does not have the property of being conceivably unextended.
Therefore (3) My body is not my mind.
Rey analyses this argument as follows:
Now this argument is patently invalid. The same form of argument could defeat almost any identity claim, consider, for example, the identity of Mt. Everest with the highest mountain on earth. By parallel reasoning, we could argue: (1*) The tallest mountain has the property of being conceivably different from Mt.
Everest.
(2*) Mt. Everest does not have the property of being conceivably different from Mt.
Everest.
Therefore: (3*) The tallest mountain is not Mt. Everest.
. . . By the same reasoning, the morning star couldn’t be identical with the evening star, Mark Twain wouldn’t be identical with Sam Clemens, and the square of 14
might not be 196! Clearly something has gone wrong.29
But his purported analogous arguments do not run to parallel the original dualist one.
If my mind has a property that my body lacks, then the mind is not the same thing as the body. Following the principle of the indiscernibility of identicals (if a=b, then any property of a is a property of b and vice versa) if Mark Twain does have some property different from Sam Clemens, then we have two people. But in fact we do not have two people and moreover if the terms “Mark Twain” and “Sam Clemens” do pick out the same individual (call this person X) then whatever is true of Mark Twain is true of Sam Clemens. What seems to be in confusion here is a different point, namely Being called Mark Twain is a different property from Being called Sam Clemens. X has both properties. Similarly the property of Being called the Evening Star is different from Being called the Morning Star, and Being the tallest Mountain is a distinct property from Being called Mt. Everest despite the fact that a planet and a mountain each have one pair of properties. Being the square of 14 is also a distinct property from Being 196 and the infinitely many other properties had by that number, e.g. Being the successor of 195, being an even number, and so on. The appeal to referential opacity does not belie the appearance that the mental and physical are distinct. If one has reason to believe that conceiving of one’s mental state does not require conceiving of the physical state it is supposedly identical with, then one has reason to think the two states are indeed two.30
Bridging the gap with more science
Some materialists claim that greater scientific awareness will help us see the mental in (or even as) the physical. C onsider C hurchland’s response to a dualist thought experiment referred to earlier in which one imagines looking at the brain but failing to see or thereby identify the relevant mental states. Churchland offers this reply to Leibniz’ dualist scenario in which one may ( contra Leibniz) recognize some mental state such as the taste of peaches in the appropriate physical configuration: It remains possible, even granting Leibniz’ story, that the taste sensation of a peach is identical with a four-element activation vector in the gustatory pathways.
And it remains possible that, should you and I happen to know what vectors constitute what sensations, and should we happen to know where and how to look for those activation vectors, then we might recognize those sensations, from our mitelike perspective, as they go by.31
Churchland claims that this spatialization of taste helps shed light on our ability to compare tastes. The reason why tasting pears and apricots is similar is because of the proximity and resemblance between the relevant activation pattern across the four types of tongue receptors as re-represented downstream in one’s taste cortex. The peach pattern differs from the apricot pattern by only a few percentage points in each of the four dimensions.32
But in knowing what peaches taste like and in knowing the location and character of a four-element activation vector, are we attending to two properties or activities or just one? I believe that a natural (and plausible) view is that we do not know, nor can we observe, the constitution of one by the other, but, rather, we can take note of the link, the interaction or embodiment of the one by and in the other. We do not observe the taste in the pattern, nor do we observe the brain pattern in reflecting on the taste, but, so it appears, we may well come to conclude that the two are intimately, causally connected such that the differences in taste are causally explained (in part) by the configuration and activity of brain activity. One can conceive of subjective properties like being such that one tastes peaches without conceiving of being gustatory pathway XYZ (however this is described in the language of current physics and chemistry) and this is an important ground for concluding that the properties are distinct. If this criterion is plausible, then one has grounds for regarding the relevant mental and non-mental, physical properties as distinct. For scientific and other practical purposes, one may well elect to treat the two as a single thing, but it remains a philosophically substantial issue as to whether one should collapse the two metaphysically. This is the point that many dualists have made: correlation is not equivalent to identity (e.g. the older generation of twentieth-century dualists: H.D. Lewis, A.C. Ewing, C.A. Campbell, C.J. Ducasse, et al, and “the next generation,” e.g. Richard Swinburne, W.D. Hart, Stewart Goetz, et al. ). Correlation is a necessary but not sufficient condition for identity. Even correlation in all possible worlds does not amount to identity (presumably the properties being the successor of 5 and being the smallest perfect number are distinct but necessarily co-extensive).
Confusing metaphysics and epistemology
Some materialists charge that the dualist appeal to experience fails to take seriously the difference between metaphysics and epistemology. Churchland diagnoses the appeal of dualism as “a conflation between different ways of knowing on the one hand, and different things known on the other.”33 To bring to light his point, Churchland notes how the same physical state, the location of one’s limbs for example, can be known in different ways:
Here, however, the object of knowledge is exactly the same from both perspectives, the subjective and the objective, and it is something paradigmatically physical: the configuration of your body and limbs.34
Other analogies involve access to one’s bladder, bowls, stomach, micro-muscles in one’s skin, lungs, flushed skin, and so on.
Such examples can be tripled, quadrupled, and more, but these eight will serve to make the point. The existence of a proprietary, first-person epistemological access to some phenomenon does not mean that the accessed phenomenon is nonphysical in nature. It means only that someone possesses an information-carrying causal connection to that phenomenon, a connection that others lack.35
Has Churchland thereby defeated the dualist appeal to subjectivity? I do not think so.
The different ways of knowing the object(s) Churchland cites – by feeling it oneself rather than not feeling it and knowing it in some other way, e.g. observing it scientifically – constitute different things that are knowable. To know the feel of one’s flushed cheeks is different – as a feeling – from knowing what it looks like from a few yards away. The point of dualist arguments is that the perspectives of the mental and physical are different sorts of things (where a “thing” is understood broadly to cover properties, activities, processes, and not limited to concrete individual substances).
Insofar as Churchland allows that there truly is a difference between the subjective appraisal of one’s body, desires, tastes, and knowing about these in other ways, by external observation, say, then he is paying homage to the kind of subjectivity that dualists seek to bring to light. Churchland’s reference to “information carrying processes” to cover both first-person observation and the kind that is captured in the language of physics and chemistry appears to override the radical differences in the mental as gleaned, on the one side, in our first-person feeling and subjective appraisal of our own states and attitudes, and, on the other, that which is depicted in physics and chemistry. The distinct reality that is manifested in subjectivity is not exhausted by citing its information content; its content is non-propositional. Even when one is thinking propositionally, I suggest there is a felt awareness of the process (there is qualia or what its like to thinking). Apprehending information involves subjective awareness.36
William Lycan advances a similar attempt to defuse the case for dualism by claiming that dualists fall prey to an epistemic confusion. He thinks dualists commit what he calls the “stereoptic fallacy”:
The Steroptic [or steroscopic] fallacy of supposing that, because to have a vivid perceptual experience of oneself is nothing like observing the brain of someone else who is having it, having such an experience must be entirely different in nature from any goings-on in the brain that underlie one’s having it.37
And Lycan’s contempt for this is clear: “Even to call this inference a ‘fallacy’ is to flatter it.” He goes on to reprimand those who think otherwise:
No materialist theory of the mind ever entailed that watching the gray cheesy brain of someone who is having an intense cyan (or whatever) sensation is qualitatively or in any other way like having that sensation oneself. Watching the brain produces gray, cheesy visual sensations in the watcher, subserved by whatever neurophysiology underlies gray, cheesy visual sensations, and gray, cheesy visual sensations are, of course, neurophysiologically and functionally quite unlike the intense deep cyan sensations being held by the subject whose neurophysiology is being watched.38
But the problem for materialism is not overlooking the fact that looking at a gray brain will mean looking at something gray. The problem is that if a person claims to have sensations, intense deep cyan sensations, and these are held to be physical, we should be able to locate them as physical states or activities. But which physical property is the very same thing as the intense sensation? If there is the feeling of being irritated by a fallacy, which property in the physical sciences is this? Lycan describes materialism, his view, as follows: “If materialism is true, then human beings are large collections of small physical objects and nothing more, ontologically. It follows that any human being could be described, and described completely, in purely scientific terms.”39 But how might one find the scientific term for being irritated by a fallacy and other psychological properties?
Arguments from analogy
I note finally two attempts to dismiss the dualist argumentation on the basis that the dualist position is no better than certain inadequate analogies.
Churchland objects that dualists who resist identifying the mental and physical are no different from a stubborn vitalist who assumes that life must be different from what one observes in the natural sciences. Notice first Churchland’s depiction of a vitalist claim:
However closely you might watch these molecular structures folding, unfolding, hooking together, unhooking, and drifting aimlessly around in the soup, it is obvious that you would never observe the impulse of life that urges its growth; you would never observe the telos of life that knows and guides its species-specific development.40
Qualia are as useless in providing reasons for dualism, just as this predicament is without any force in favor of vitalism.
The problem is that if vitalism is understood as asserting that there is such a thing as knowing, felt urges, guidance on the basis of some grasp of (or maybe belief in)
purpose and so on, then the difficulty of identifying these with physical processes observed in the natural sciences is very much in place. It is only because the concept of “life” as it is employed by most contemporary biologists seems to be a straightforward non-mental category (defined in terms of cell replication and so on) that there is no philosophical puzzle about the claim to observe life or identify it with physical processes. The question remains, however, whether the dualist position about our own (and some non-human animal) mental lives should be jettisoned as quickly as we have given up thinking of “life” itself in quasi-mental terms. It is one thing to give up the notion that “life” should be understood in mental categories and another to give up the notion that our thinking, believing, feeling, sensing and the like should be understood as bona fide mental categories.
Churchland’s disregard of qualia comes up in another analogy in which he charges that dualists commit a confusion between types of awareness as one finds in a commonplace manual. Churchland complains about the confused metaphysics in a microwave instruction book. The manual includes this claim: The microwaves agitate and vibrate the moisture molecules at such a great rate that friction is created; the friction, in turn, creates heat and the heat causes the food to cook.41
Churchland comments:
This raises a problem: how to connect heat with the rest of what is going on. Here the authors fall back on their prescientific folk understanding of one of the many things that can cause heat: friction! The result is massively misleading to the innocent reader, who is left with the impression that rubbing two molecules together causes heat in the same way that rubbing your two hands together causes heat.42
There is a confusion in the manual, but I take issue with Churchland’s diagnosis.
The puzzle created by the manual results because it can be read as switching from talk of “heat” as molecules in motion to talk of “heat” as a feeling of warmth, and then the manual appears to revert back talk about molecules in motion. In outlining how a microwave cooks (though not how microwaved food tastes) there is no more need to invoke feelings of warmth than there is a need to invoke vitalistic talk of “urges” when describing the growth of a tree. Instead of playing into Churchland’s hands, I believe his pointing to an awkwardness in the manual brings to the fore an awkwardness of entirely dispensing with folk, subjective notions of warmth. I suggest we do wind up with a truncated and unsatisfactory theory of what it feels like to be warm if we were to construct a manual on rubbing hands together using Churchland’s data that only
referred to molecular friction. In a sense, then, Churchland’s use of the example can be turned around. He seems to acknowledge there is a difference between “heat”
when this is engendered by rubbing hands and “heat” when this designates molecules in motion when no sentient being is in the vicinity. Advocates of the dualist appeal to experience would pay him the complement of taking this distinction more seriously, though they would carry it to a conclusion Churchland would reject.
Nonreductive materialism and the limits of the natural world If one grants a distinction between the mental and the physical as the physical is conceived of by Lycan, Churchland, Rey, Dennett et al. , why not simply extend the concept of the physical to cover this stray data? A range of philosophers, including Noam Chomsky, John Searle, and Galen Strawson, seem to challenge the limits set on the physical by standard forms of materialism. Churchland is bent on not allowing entities to escape his materialist net. On proprioception he writes: “There is nothing supraphysical, nothing beyond the bounds of physical science here.”43 And later:
“Does the undoubted existence of this unique way of your knowing about your own internal states mean that there is something nonphysical about those states, something that must transcend representation within physical science?”44 But why not allow that there is something beyond a limited notion of what is physical?
One needs some notion of what might be beyond the natural or physical world in order to see what might be at stake in loosening or tightening our concept of what counts as physical. Leopold Stubenberg makes the excellent point that if there is no alternative to materialism, we might just as well treat qualia, subjectivity and experience, as something that will eventually be incorporated into materialism even though we currently have little idea on how this might be accomplished: Materialistic science stands unrivaled. The belief that consciousness will force this giant onto its knees may even seem slightly mad.45
Colin McGinn also holds that a materialist naturalism is the only real alternative.
It is either eliminativism or miracles or hidden structure. Absolute noumenalism is preferable to denying the undeniable or wallowing in the supernatural.46
McGinn argues for a form of naturalism that posits a hidden structure linking the mental and physical.
I think this approach is sensible. In the absence of an alternative to materialism we should stick it out with materialism, opting either for McGinn’s hidden structures or
working to display these structures in the clear light of day. But in staying with materialism we will be resisting not just some of the points made above against identity materialism, but also resisting our (at least apparent) ability to see that the mental and physical are only contingently related. Thought experiments abound in the philosophical literature that cause difficulties for identity and non-reductive materialists who believe that the mental and physical are inextricably bound together (or, using popular, current terminology, that the mental supervenes on the physical).
Plausible thought experiments describe cases in which a person has inverted senses (seeing, smelling or tasting in ways that are systematically different from others notwithstanding identical physical conditions), absent qualia (a person lacks a qualia that should be present, given current materialist claims), zombies or superspartans (in which consciousness is missing when, if materialism is right, consciousness should be present), nations instantiate the materialist analysis of mental states and yet do not have the requisite mental states, persons switch bodies without any bodily parts switching, or persons become altogether disembodied.47 None of these ostensible possibilities seem to be bona fide possibilities given most forms of materialism.
One materialist strategy at domesticating these thought experiments is to propose that they represent pictures of worlds in which dualism is true but, as it happens, in this world, dualism is false and thus these are not real possibilities. But this seems to avoid getting at the deeply rooted sense of contingency captured in the thought experiments. Just as my imagining I could be in Rome seems to be grasping a genuine possibility about me in my present circumstances and not just grasping something about me in some different universe, the above thought experiments can be structured so that they reveal what appear to be genuinely contingent relations in this world.48 Theism seems to provide a framework in which we may understand the relevant contingency and incorporate this contingency in a broader, explanatory framework.49
According to theism, the origin and continued existence of the cosmos is in virtue of the intentional activity of an omnipotent, omniscient, good being, God. Theists do not see intentionality as an emergent entity from an utterly mindless background.
Behind the emergence of animal and human consciousness is the antecedent consciousness or mind of God. Consciousness is not, then, something queer or anomalous, a ‘parlor trick’. So, McGinn puts the alternatives as follows: One wants to insist, consciousness cannot really be miraculous, some kind of divine parlor trick. It must fit into the natural order of things somehow. Its relation to matter must be intelligible, principled, law-governed. Naturalism about consciousness is not merely an option. It is a condition of understanding.
It is a condition of existing.50
But in a theistic view of consciousness, there is no parlor trick or discrete miraculous act of God behind the emergence of consciousness. Consciousness emerges from the physical cosmos through an abiding comprehensive will of God that there be a world of physical and non-physical objects, properties, and relations. The relation between matter, energy, consciousness, the laws of space-time, tout court, all stem from an overarching, divine activity.
In a theistic metaphysics, the materialist reductive proposal of Dennett and others are reversed. As mentioned at the outset, contemporary philosophy of mind is almost entirely materialist. For many philosophers, the key philosophical project is therefore to locate the mental in the physical world, and this means explaining the mental in terms that are not mental. So, Daniel Dennett privileges explanations of nature that are, at base, non-mental and make no use of mental terms like intelligence and purpose. In his view, to achieve an ultimately satisfactory explanation of human intelligence requires that one not invoke ever increasingly intelligent forces in the universe, but, rather, one accounts for intelligence in terms that ultimately can account for it without invoking intelligence. Dennett contends that accounts that still leave intelligence unexplained are question begging:
The account of intelligence required of psychology must not of course be question-begging. It must not explain intelligence in terms of intelligence, for instance by assigning responsibility for the existence of intelligence in creatures to the munificence of an intelligent Creator . . .51
Dennett’s position will thereby feed into the project of securing the unity of science.
By explaining intelligence in categories that do not involve it, one can aim to unify the kinds of accounts used in science. Dualists by comparison seem left with intelligence plus all the rest; the mental appears to dangle. “Only a theory that explained conscious events in terms of unconscious events could explain consciousness at all.”52 Georges Rey adopts a similar strategy:
Any ultimate explanation of mental phenomena will have to be in non-mental terms, else it won’t be an explanation of it. There might be explanations of some mental phenomena in terms of others – perhaps hope in terms of belief and desire –
but if we are to provide an explanation of all mental phenomena, we would in turn have to explain such mentaliastic explainers until finally we reached entirely non-mental terms.53
This seems to be an intelligible, materialist goal. Theism reverses the order of explanation.
Theism treats all the laws of nature and the entire cosmos as the result of intentional activity. As such, the account of physical realities and non-conscious, non-mental events, must, in the end, appeal to an intentional reality. One might well reverse all of Dennett’s claims within a theistic framework and claim that any explanation of the physical, non-intentional world that did not break out into a deeper, intentional account would be question-begging and not genuinely explanatory. From the vantage point of a fundamentally materialist cosmology, the emergence of consciousness seems strange; it is likened to claiming “then a miracle happens.”54 But from the vantage point of theism, the emergence of consciousness may be seen as something deeply rooted in the very nature of reality. The creation of animal and human consciousness is not some isolated miracle, but a reflection of the underlying structure of reality.
The appeal to a combined case for theism and dualism may be further appreciated in the course of addressing three areas: the nature of divine properties, the nature of mental causation, and the way in which both theism and dualism may be developed along integrative lines. I shall address each briefly.
The nature of divine attributes
Some critics of theism underestimate the resources of theism. Consider, for example, Brian O’Shaughnessy’s comment on the problem of matching theism with the precise nature of the cosmos as revealed in modern science:
Well, four centuries of triumphant advance by the rock-bottom physical sciences of physics cannot but leave some mark on philosophy. When you can predict the wave length of a spectrum line to eight decimal places it is rather more difficult to believe that the underlying reality of everything is spiritual, e.g., an immaterial Deity. After all, should a Deity be so fastidious?55
O’Shaughnessy seems to suggest the peculiar position that if there is a God, we should expect the universe to be unspecified, vague or blurred without having any precise features that are measurable to eight decimal places of some length. Alternatively, perhaps if there is a God, God would only make objects that are easily delimited by measurements in whole numbers without any use for fractions. Clearly, the God of classical theism is radically different from O’Shaughnessy’s portrait. Traditionally, theists understand God to be unsurpassably great in power, knowledge, and goodness.
Naturalists often fail to adequately appreciate classical Jewish, Christian, and Islamic theistic claims of God’s essential aseity or ultimacy. McGinn objects to theists who account for the emergence of consciousness in nature in terms of a broader divine consciousness:
The theory takes for granted the existence of the conscious agent who is held to explain the existence of all other conscious agents. But if there is a problem about how the conscious beings we see round us come to exist, then there is equally a problem about how the conscious being who creates each of those conscious beings came to exist.56
To court theism is to entertain the thesis that there is a being whose properties of omniscience, omnipotence, goodness and aseity are not derived from some other agency or explainable in terms of the laws of nature. The theist asks you to consider the possibility that there is such a being and then look at the world to see whether the features of the world make greater or lesser evident sense. Theism is a comprehensive philosophy that offers an explanation of the very existence and continuance of the world and its constituents. I believe that a devalued view of divine attributes leads McGinn and other naturalists underestimate the resources of theism for explaining the very existence of our contingent cosmos. “There is nothing about the intrinsic structure of lumps of mere matter that suggests a divine origin. When we look at rocks through a microscope, no trace of the divine greets the eye.”57 But our eyes are greeted by a cosmos that is contingent (on this point, many naturalists would agree) and it is this feature of the universe along with the emergence of consciousness, that theism can address with great power.58
The nature of mental causation
Critics object that dualism leaves us with a complete mystery in terms of mind–body interaction. How can things so remote interact? Invoking theism seems to heighten rather than reduce the problem, for doesn’t theism require us also to believe that there can be causal interplay between radically distinct levels of reality? To appeal to theism here would be to explain the obscure in terms of the more obscure.
As R.C . Richardson, B. Mijuskovic and R. Swinburne have forcefully argued, dualist interaction between the mental and physical only looks peculiar if one assumes that the only causal interaction that is intelligible is physical to physical interaction.59
There is no compelling reason to conclude at the outset that causal interaction cannot be between something physical and something non-physical. There is little consensus today on the nature of physical-to-physical causation, and so little common ground for using an uncontroversial theory of causation in an attack on dualism. Moreover, the assumption that contemporary science somehow rules out such interaction or shows there is none, seems motivated more by philosophy than science. Consider, for example, Churchland’s summary of the scientific data against dualism.
It will be evident from the rest of this book that this familiar hypothesis [dualism]
is difficult to square with the emerging theory of cognitive processes and with the experimental results from several neurosciences. The doctrine of an immaterial soul looks, to put it frankly, like just another myth, false not just at the edges, but to the core.60
But none of the solid scientific findings he cites shows that one cannot square dualism with the neurosciences, and it is easy to wonder whether the reason why dualism (or dualism plus theism) looks mythic is because of prior materialist commitments. Given an overall materialist metaphysic, there is not going to be much motivation for recognizing some relatively local exceptions in human and non-human consciousness to the general order of nature.
By way of underlining the way in which C hurchland blends his science and metaphysics, I take note of the ease with which a dualist can accommodate many of his scientific claims. Dualists can readily understand Churchland to have located key physical causes in shaping our consciousness. Addressing some particulars, Chruchland writes: “This is where past learning shows itself, where character and insight come in, and where intelligence is ultimately grounded.”61 This may readily be seen as identifying the bodily backing or foundation of our mental lives, but it is quite a different, and decidedly metaphysical, matter to charge that intelligence and so on is the brain and body. The blending of science and philosophy emerges at various points. I suggest the following claim seems more metaphysics than science:
“Without a neural network in place, there can be no self, neither an emotional self, nor a perceiving self, nor a deliberating self, nor any other kind of self.”62 Imagine science shows the exact correlation in our own case of selves and neural networks and does not give us evidence of there being any immaterial being (God, say) that is a person or person-like. Why would this be taken to demonstrate a necessary connection between selves and neural networks? I think Churchland’s claim is informed by the sciences, but it reflects considerable speculation beyond what is dictated by the sciences alone.
Integrative dualism and theism
Finally, I note that both dualism and theism have been criticized for foisting on us a splintered view of ourselves and the cosmos. Anthony Kenny, P.M.S. Hacker and others propose that dualism offers a huge bifurcation between the person and body such that dualists leave us with a dark skepticism about the mental life of others.
But this also seems like a caricature, especially as dualism is seen in light of theism.
In classical theism the world is understood as fashioned by an all-good God who insures the proper functioning of our cognitive, sensory, and volitional powers. The goodness of God is the guarantor of the fit. Given the goodness of God, we might employ Mackie’s terms and conclude that a darker, skeptical hypothesis about the
adequacy of our cognitive faculties would be ‘queer’ (or unlikely).
Some dualists do not emphasize sufficiently the way in which dualism may offer a fully integrated understanding of persons and bodies. The fact, if it is one, that the person and body may be contingently related is no reason to think that in virtually all practical contexts the person and body are functionally identical. Richard Swinburne, for example, depicts a person’s relation with their body as follows: “An agent has a body if there is a chunk of matter through which alone he or she can make a difference to the world, and through which alone he or she can learn about the world.”63 As a dualist, I do not deny the claim, but I think it is important to emphasize that in a fully functioning, embodied life, the person and this chunk of matter are functionally indistinguishable. To touch someone’s body is not to touch something they are tied to, but, in a perfectly common sense, ordinary sense of the terms, to touch them.64
In this last section of this chapter, I have made a plea to understand the philosophy of mind in a broader setting. Since World War II we have seen philosophy of mind expand in important respects; some of this has been motivated by scientific advance and speculation, and some by incorporating other domains in a fuller intellectual environment. For example, much work in philosophy of language has been brought into philosophy of mind, and with great profit, owing to J.L. Austin’s seminal contributions. I propose that philosophy of mind may be more fruitfully debated and explored if it too might be incorporated and debated within a broader metaphysical context, one that pares materialism and naturalism over against theism. In this broader setting, we may well see a theistic-dualist philosophy, not as a case of
“wallowing” in the supernatural (to use McGinn’s term), but as offering a credible, comprehensive philosophy of nature that is profoundly unified, coherent, and able to account for things that otherwise seem queer and anomalous.65
Notes
1 J.L. Mackie, Ethics: Inventing Right and Wrong (Harmondwsorth, England: Pelican, 1977), 38–
42.
2 Ibid., 38.
3 Paul M. Churchland, The Engine of Reason, the Seat of the Soul (Cambridge, Mass.: MIT Press, 1995), 211.
4 Jaegwon Kim, Mind in a Physical World (Cambridge, Mass.: MIT Press, 1998), 2.
5 William Lycan, Consciousness and Experience (Cambridge, Mass.: MIT Press, 1996), 1.
6 Daniel Dennett, Kinds of Minds (San Francisco: Basic Books, 1996), 24. See also Dennett,
“Living on the Edge,” Inquiry 36 (1993): 135–59.
7 Nagel suggests a direct link between anti-theism and the growth of “scientism and reductionism of our time” (Thomas Nagel, The Last Word [Oxford: Oxford University Press, 1997], 131). See also Georges Rey, Contemporary Philosophy of Mind (Oxford: Blackwell, 1997), 23. Churchland refers to dualism in various places as “the religious hypothesis.”
8 George Mavrodes, “Religion and the Queerness of Morality,” in Rationality, Religious Belief and Moral Commitment, ed. Robert Audi and William J. Wainwright (Ithaca, N.Y.: Cornell
University Press, 1986), 213–26.
9 For versions of these arguments, see Galen Strawson, Mental Reality (Cambridge, Mass.: MIT
Press, 1994); Lynne Rudder Baker, Explaining Attitudes: A Practical Approach to the Mind (Cambridge: Cambridge University Press, 1995); William Hasker, The Emergent Self (Ithaca, N.Y.: Cornell University Press, 1999); and Victor Reppert, “Eliminativism, Cognitive Suicide, and Begging the Question,” Metaphilosophy 23 (1992): 378–92.
10 Churchland, The Engine of Reason, xi.
11 Ibid., 3.
12 Paul M.Churchland, Metaphysics and the Mind–Body Problem (Oxford: Clarendon Press, 1979), 64.
13 Churchland, The Engine of Reason, 19.
14 Ibid., 8.
15 Ibid., 322.
16 Ibid., 17, 18.
17 Ibid., 324.
18 Ibid., 58 ff.
19 Ibid., 155.
20 Ibid.
21 Ibid., 131.
22 The threat of epiphenomenalism to materialism is lampooned by Dennett in Brainchildren (Cambridge, Mass.: MIT Press, 1998), but I think he should be far less sanguine. Kim’s Mind in a Physical World provides a compelling, sober counter-point.
23 Some materialists note the way in which ordinary terms might be refined under further analysis, scientific and philosophical. Note, for example, William Lycan’s comments on
“belief”:
The word “belief” . . . points dimly toward a natural kind that we have not fully grasped and that only mature psychology will reveal. I expect that “belief” will turn out to refer to some kind of information-bearing state of a sentient being . . . but the kind of state it refers to may have only a few of the properties usually attributed to beliefs by common sense (William Lycan, Judgement and Justification [Cambridge: Cambridge University Press, 1988]).
24 Colin McGinn, The Mysterious Mind (New York: Basic Books, 1999), 24.
25 Charles Taliaferro, Consciousness and the Mind of God (Cambridge: Cambridge University Press, 1994); and W.D. Hart, The Engines of the Soul (Cambridge: Cambridge University Press, 1988).
26 See Taliaferro, Consciousness and the Mind of God, chapters two and three for an elaboration and defense of these points.
27 See Frank Jackson, “What Mary Didn’t Know,” Journal of Philosophy 83 (1986): 291–5.
28 Ned Block, “Consciousness,” in A Companion to the Philosophy of Mind, ed. Samuel Guttenplan (Oxford: Blackwell Publishers, 1994), 213.
29 Rey, Contemporary Philosophy of Mind, 57.
30 I defend the reliability of thought experiments and this criterion of identity in “Modal Reliabilism,” paper in preparation.
31 Churchland, The Engine of Reason, 193.
32 Ibid., 23.
33 Ibid., 201.
34 Ibid., 197.
35 Ibid., 198.
36 See Strawson, Mental Reality; and Taliaferro, Consciousness and the Mind of God.
37 Lycan, Consciousness and Experience, 6–7.
38 Ibid., 48.
39 Ibid., 45.
40 Churchland, The Engine of Reason, 192.
41 Ibid., 207.
42 Ibid.
43 Ibid., 197.
44 Ibid.
45 Leopold Studenberg, Consciousness and Qualia (Amsterdam: John Benjamins Publishing Co., 1998), 32.
46 Colin McGinn, The Problem of Consciousness (Oxford: Blackwell, 1991), xii.
47 See Taliaferro, Consciousness and the Mind of God, chapters two and three for a review and analysis.
48 See Hart, The Engines of the Soul; J. Harrison, Essays on Metaphysics and the Theory of Knowledge (Aldershot, England: Avebury, 1995); Richard Swinburne, The Evolution of the Soul (Oxford: Clarendon, 1997); Charles Taliaferro, Consciousness and the Mind of God; Taliaferro. “Animals, Brains, and Spirits,” Faith and Philosophy 12 (1995) ; Taliaferro, “The Perils of Subjectivity,”
Inquiry 40 (1997); Taliaferro, “Possibilities in the Philosophy of Mind,” Philosophy and Phenomenological Research 57 (1997): 127–37.
49 I defend this in Consciousness and the Mind of God as well as in “Mysterious Flames in Philosophy of Mind; Reflections on McGinn’s Naturalism,” forthcoming in Philosophia Christi.
50 Colin McGinn, The Problem of Consciousness, (Oxford: Blackwell, 1991), 47.
51 Daniel Dennett, Brainstorms (Cambridge: MIT Press, 1978), 83.
52 Daniel Dennett, Consciousness Explained (Cambridge: MIT Press, 1991), 454.
53 Rey, Contemporary Philosophy of Mind, 21.
54 Dennett, “Living on the Edge,” 141.
55 Brian O’Shaughnessy, The Will, vol. 1 (Cambridge: Cambridge University Press, 1980), xvii.
56 McGinn, The Mysterious Mind, 86.
57 Ibid., 79.
58 See Taliaferro, Consciousness and the Mind of God; and Taliaferro, Contemporary Philosophy of Religion (Oxford: Blackwell, 1998), for a detailed defense of this claim.
59 See R. C. Richardson, “The ‘Scandal’ of Cartesian Interactionism,” Mind 92 (1982): 20–37; B. Mijuskovic, “Brentano’s Theory of Consciousness,” Philosophy and Phenomenological Research 38 (1978): 315–24; and Richard Swinburne, The Evolution of the Soul.
60 Churchland, The Engine of Reason, 17.
61 Ibid., 11.
62 Ibid., 308.
63 Richard Swinburne, Providence and the Problem of Evil (Oxford: Oxford University Press, 1998), 98.
64 See Taliaferro, Consciousness and the Mind of God.
65 My thanks for comments on different versions of this paper to the Philosophy Departments at Bethel College, St. Cloud State University, the University of Chicago, New York University, and to Professor Dore Ashton’s seminar at Copper Union.
agency
Stewart Goetz
Introduction
David Papineau introduces his book, Philosophical Naturalism, with the observation that though the term “philosophical naturalism” (“naturalism”, for short) is a familiar one nowadays, there is no universal consensus about its meaning.1 Papineau believes that naturalism is a commitment to the completeness of physics, where physics is complete in the sense that a purely physical specification of the world, plus physical laws, will always suffice to explain what happens. The concepts of physics, however, change over time. What categories, therefore, will qualify as ‘physical’ in the ultimate or final physics? We cannot, says Papineau, presently answer this question with any certitude. At best, we can specify one category which will not qualify, and that is the category of the psychological which involves propositional attitudes (e.g. beliefs, desires) that represent things as being a certain way (intentionality).2
When I say that a complete physics excludes psychology, and that psychological antecedents are therefore never needed to explain physical effects, the emphasis here is on “needed”. I am quite happy to allow that psychological categories can be used to explain physical effects, as when I tell you that my arm rose because I wanted to lift it. My claim is only that in all such cases an alternative specification of a sufficient antecedent, which does not mention psychological categories, will also be available.3
David Armstrong is another philosopher who addresses the issue of naturalism.4
According to him, naturalism is “the doctrine that reality consists of nothing but a single all-embracing spatio-temporal system.”5 Contemporary materialism is a form of naturalism and maintains that the single all-embracing spatio-temporal system contains nothing but the entities recognized by physics. Irreducible purpose or teleology has no place in this (or any other) spatio-temporal system as an explanatory
Naturalism and libertarian agency 157
principle because it entails the characteristic of irreducible intentionality, and irreducible intentionality implies the falsity of naturalism. Thus, Armstrong says, “I suppose that if the principles involved [in analyzing the single, all-embracing spatiotemporal system which is reality] were completely different from the current principles of physics, in particular if they involved appeal to mental entities, such as purposes, we might then count the analysis as a falsification of Naturalism.”6
If we take our lead from Papineau and Armstrong, naturalists are those who maintain that the ultimate explanation of things will not include psychological (intentional, mental) categories and, therefore, will exclude teleology. What, then, is the ontological status of the psychological according to naturalism? Is is not even real?
‘Hard’ naturalists not only deny its relevance for ultimate explanation, but also its reality.7 ‘Soft’ naturalists, while denying the psychological’s ultimate explantory relevance, try to preserve its reality. Such naturalists often invoke the concept of supervenience and maintain that the psychological supervenes on the physical. Thus, David Chalmers argues for naturalism and maintains that while “experience is superfluous in the explanation of behavior,”8 the fundamental theory of the world will include psychophysical supervenience laws which tell us how experience arises from physical processes.9 John Heil, who is also a naturalist, says the following about naturalism: “Naturalism in the philosophy of mind is the view that mental characteristics are determined by or supervene on features of agents comprehended by the natural sciences.”10
A common thread running through all conceptions of naturalism, then, is that the fundamental explanation of any event is non-psychological in nature. Naturalism, therefore, ultimately implies the falsity of teleological explanation. Because naturalism excludes teleology, it also excludes libertarian freedom or agency.
According to libertarianism, a choice is an undetermined mental action which is explained teleologically in terms of a purpose or goal of its agent. A teleological explanation of a choice, because it makes reference to the concept of a goal, end, or purpose derived from the representational content provided by a propositional attitude, is believed by the naturalist to be either eliminable for, reducible to, or dependent upon a fundamental non-teleological explanation of that event.
In addition to rejecting the reality of libertarian freedom, naturalists also deny substance dualism (dualism, for short), which is the view that there is a non-physical soul or mind which has irreducible psychological properties.11 Given the naturalist’s rejection of both libertarianism and dualism, an interesting and important question is whether the truth of the former implies the truth of the latter. If it does, it would not be surprising to find that naturalists reject libertarianism because they reject dualism.
The course of my argument in this paper is as follows. In Section I, I set forth a non-causal account of libertarian agency in which a choice is an essentially uncaused mental action explained by a telos or purpose of its agent, and I make clear why such
an account is an intuitively plausible conception of free mental action. Surprisingly, but for no good reason, a non-causal account of human freedom is normally rejected by libertarians themselves. Therefore, in Section II I defend my non-causal agency theory against the criticisms of other libertarians as a way of highlighting the centrality of teleological explanation for libertarianism. Section III consists of an explanation of why naturalists believe that teleological explanation of choice implies the truth of dualism and a discussion of the alleged problem of causal interaction which they assert is a decisive reason for rejecting dualism. I claim that dualism is no worse off than any soft naturalist’s view of the mind when it comes to explaining the relationship between the psychological and the physical. Thus, if naturalism implies the falsity of dualism and libertarianism entails dualism, if we have libertarian freedom, which one prominent soft naturalist concedes we certainly seem to have,12 naturalism is false.13
Finally, in Section IV I briefly examine an emergentist alternative to the view that libertarianism implies dualism, and explain why it is inadequate.
I
A desideratum which any libertarian theory must satisfy is that of being able to explain how it is that mental actions such as choices differ ontologically from mental non-actions or mere happenings. One way to account for this is in terms of the extrinsic relationship of causal ancestry: mental actions differ from mental non-actions in terms of their different causal histories. This is the approach taken by libertarians who espouse agent-causation. A more intuitively satisfying view is to say that mental actions differ intrinsically from mental non-actions. Because a choice is a mental action, how can it be intrinsically active? In the following way: most generally, a mental power is an ontologically irreducible property which is exhibited by an entity. Corresponding to a mental power is the exercising of that power which is an event. Exercising a mental power is acting, and mental action is essentially uncaused. One kind of mental power an agent possesses is the mental power to choose. When an agent exercises it, he chooses. Thus, (i) a choice is the exercising by an agent of his mental power to choose, where (ii) the exercising of the power to choose is essentially an uncaused event.
Consider (i). Carl Ginet has recently defended an account of mental action which maintains that any mental act differs intrinsically from passive mental events.
According to him, a mental act’s intrinsic active nature consists in its having an ‘actish phenomenal quality’. This actish phenomenal quality lacks the complex structure of a causal relation and, by itself, is enough to make a mental event a mental action.14 I believe it is a mistake, however, to characterize the intrinsic active nature of a mental action as a phenomenal quality.15 Such a characterization suggests that a mental act has a distinctive quale or feel about it which makes it intrinsically active and distinguishes it from a passive mental event. The exercising of the power to choose,
Naturalism and libertarian agency 159
however, like any mental act, has no intrinsic feel or quale about it. Thus, while I may feel tired after exercising my power to think, which I do when I think for a long time about the issue of free will, this active thinking itself has no intrinsic quale any more than choosing does. The active nature of a mental event consists solely in its being the exercising of a mental power.16
If a mental action is an event which is the exercising of a power, what is a mental non-action or mere happening? It is what I will call the actualization of a mental liability.
Like a mental power, a mental liability is an ontologically irreducible property which is exhibited by a subject. When a subject’s mental liability is actualized, a caused event occurs with respect to which he is a patient. One mental liability a subject has is the liability to believe. When his liability to believe is actualized, he believes a proposition and he is a patient with respect to believing that proposition.
In summary, there are two types of mental properties, namely, powers and liabilities. These two kinds of properties are inherently different from each other and each is an ultimate category in our ontology. Corresponding to these two kinds of mental properties are two kinds of events, namely, the exercising of a mental power (an action) and the actualization of a mental liability (a passion). Like the properties themselves, these two kinds of events are inherently different from each other such that any exercising of a mental power is intrinsically distinguished from any actualization of a mental liability.
Support for (ii) is conceptual in nature. An event which is efficiently caused is produced by that cause and as such is an occurrence with respect to which its subject is essentially passive. An event is being made to occur to the subject and it (the subject) is not active with respect to that event. Because an exercise of mental power is active in nature, it is not produced and, thus, cannot be caused.
The intrinsically active and uncaused nature of choosing (ontology) is confirmed by the epistemology of action. In commenting on causal theories of action, Harry Frankfurt makes the following point about an agent’s knowledge of his action: They [causal theories] are therefore committed to supposing that a person who knows he is in the midst of performing an action cannot have derived this knowledge from any awareness of what is currently happening but that he must have derived it instead from his understanding of how what is happening was caused to happen by certain earlier conditions. . . . This is what makes causal theories implausible. They direct attention exclusively away from the events whose natures are at issue, and away from the times at which they occur.17
Frankfurt’s remarks suggest the following point about mental action such as choice: It is an epistemological feature of an agent who knows that he is making a choice that he knows this while he is choosing. Given this fact, it is natural to think that the agent
knows that he is choosing by being aware of the choice which he is making. However, if he knows in this way that he is choosing, it seems to follow that choosing is intrinsically different from a mere happening or passive event and that he is aware of this difference. On a causal theory of action, however, an agent who knows that he is choosing cannot possess this knowledge in virtue of his awareness of the choice itself.
This is because on a causal theory of action a choice is not intrinsically different from a mere happening. In themselves, the two are indistinguishable. Therefore, an agent can know that he is choosing only by being aware of causal differences which distinguish the two events.
According to Frankfurt, a causal account of action implies that the agent’s attention must be directed away from the mental action which he is performing in order for him to know that he is acting. Thus, a causal theory is unable to account for a significant epistemological feature of mental action. However, while Frankfurt has directed our attention to an important epistemological feature of mental action, he unnecessarily obscures his point by suggesting that the problem with a causal theory of action stems from what it implies about the temporal distance separating a mental action from its cause. Thus, Frankfurt not only maintains that a causal theory entails that an agent who knows that he is choosing must derive his knowledge from an awareness of what causes his choice, but also he maintains that the cause is a certain earlier event. Because it is, the agent who knows that he is choosing must be directing his attention away from the time at which he is choosing.
If the problem with a causal theory of action pointed out by Frankfurt were essentially linked with this temporal issue, the causal theorist would have a rather obvious response. He could simply stipulate that the cause which distinguishes his choice from a passive event happens simultaneously with the choice.18 Because the cause occurs simultaneously with the choice, there is no epistemological problem of the kind noted by Frankfurt. It is true that the agent who knows that he is choosing must look to a causal antecedent, but the causal antecedent is, in virtue of its simultaneity with its effect, known immediately by the agent.
The epistemological problem with a causal theory of action pointed out by Frankfurt is not essentially linked with temporal considerations. Rather, it arises in light of the fact that agents who know that they are performing mental actions possess this knowledge by being aware of the actions themselves, without reference to any cause of them. In the case of a choice, the exercising of the power to choose by the agent is essentially uncaused and intrinsically active and the agent knows that he is choosing by simply being aware of the choice (the exercising of the power to choose) itself.
If agents are directly aware of choosing and choices are essentially uncaused events, it seems to follow that agents are directly aware of the lack of causation of their choices. But is this plausible? Some critics of libertarianism maintain that it is thoroughly implausible “because undetermined choice (as opposed to our lack of
Naturalism and libertarian agency 161
recognition of causes for our choices) is not introspectible at all.”19 We cannot introspect indeterminacy because we cannot “observe the lack of causes.”20
If an agent had to conduct a search for the lack of causes, it would be implausible to maintain that he is aware that his choice is uncaused because he is aware of the lack of any causes which produce his choice. It would always be the case that he might have overlooked a cause which is actually there. It does not follow from this point, however, that an agent can at best only fail to observe the presence of any causes. It is a conceptual truth that a choice is uncaused because a choice is the exercising of the power to choose and exercisings of mental powers are essentially uncaused. Thus, because an agent can be directly aware of exercising his power to choose, he knows that his choice is uncaused because this is entailed by the nature of what it is to be a choice.
If a choice is uncaused, what explains its occurrence? Because a choice is a mental action, the most plausible and intuitively satisfying answer is that a choice, like any action, is explained in terms of a reason, where a reason is a goal, end, or purpose for which an agent chooses. The concept of acting for a reason is correctly captured or expressed with the “in order to” locution. Thus, we say that Kathryn chose to be an exchange student in Germany in order to learn to speak the German language. To explain a choice by a reason is to explain it teleologically, not causally.
An agent’s reason for choosing is grounded in the representative, semantic contents of certain of his psychological attitudes. These psychological attitudes are his desires and beliefs. For example, when Kathryn chose to be an exchange student in Germany, she had a desire to learn the German language and believed that spending several months living with a family in Germany and attending Hochschule would achieve the goal of learning German. The desire that she learn German provided her with her purpose or goal expressed by “in order to learn German” which explained her choice to be an exchange student. Thus, the semantic content of Kathryn’s desire was not identical with her reason for choosing, but provided the basis for the reason for her choice. Similarly, the event of her coming to have or the event or state of her continuing to have the relevant desire did not cause her to make her choice, as is typically maintained by causal theorists.21
II
I have developed an account of libertarian freedoom (free will) where indeterministic choices are uncaused events which are adequately explained in terms of teleology alone. Not all libertarians have endorsed such a non-causal account. Indeed, most have been reluctant to do so.22 In this section, I will set forth and respond to five libertarians who have claimed that teleological explanation of choice must be supplemented by causal explanation.
To begin, consider Richard Taylor’s criticism of non-causal libertarianism.
According to Taylor, “[o]nly the slightest consideration will show that this simple denial of determinism has not the slightest plausibility.”23 Why not? Because, while non-causal libertarianism avoids picturing an agent as a puppet, it does so by substituting something even less like a human being; for the conception that now emerges is not that of a free person, but of an erratic and jerking phantom, without rhyme or reason at all. . . . There will never be any point in asking why
[choices] occur, or in seeking any explanation of them, for under the conditions assumed there is no explanation. They just happen, from no causes at all.24
From the fact that a choice is uncaused, however, it does not follow that it has no explanation. To conclude this is to identify causation with explanation. Contrary to what Taylor’s argument assumes, causation is only one kind of explanation. Causation stands to explanation as species to genus. Teleology is another form of explanation.
Moreover, Taylor’s rejection of non-causal libertarianism is puzzling because he recognizes that “[i]n the case of an action that is both free and rational, it must be such that the agent who performed it did so for some reason, but this reason cannot have been the cause of it.”25
Next, consider Robert Kane’s view of free will. He asserts that in formulating their accounts of free will, libertarians should not appeal to special kinds of entities or causes (substances, properties, relations, events, states, etc.) which are not needed by non-libertarian accounts of free will.
The only difference allowed between libertarian and nonlibertarian accounts is the difference one might expect – that some of the events or processes involved in libertarian free agency will be indeterminate or undetermined events or processes. But these undetermined events or processes will not otherwise be of categories or ontological kinds that do not also play roles in nonlibertarian accounts of free agency . . . – the difference being that in nonlibertarian theories, these events or processes need not be undetermined. Such differences as there are between libertarian and nonlibertarian theories should flow from this difference alone, and the task will be to make sense of a libertarian freedom satisfying the plurality conditions, given this difference.26
The plurality conditions to which Kane refers in the last-quoted sentence consist of an agent’s having reasons for alternative courses of action; there being no compulsion or coercion of the choice which is made by him for the relevant reasons; and his having control over that choice in the sense that he makes it for those reasons.27
It is clear that the non-causal account of free will I have developed in which a choice
Naturalism and libertarian agency 163
is an uncaused exercising of the power to choose does not introduce or appeal to categories or kinds of entities beyond those mentioned by Kane. An uncaused choice is an exercising (an event) of a power (a property) which is causally undetermined.
Moreover, my account of free will satisfies the plurality conditions set forth by Kane.
Nevertheless, he criticizes the libertarian position which maintains that only reasons-explanations can be given of choices because
it stands in the way of understanding the place of free will in the natural order in which causal explanations of deterministic or probabilistic kinds hold sway.
Causal explanations cannot be the whole story of free will, but they must be part of the story if free will is to have a place in the natural order. And since deterministic causal explanations are ruled out for undertermined free willings, it is important to ask how explanations of them in terms of reasons are related to probabilistic or nondeterministic causal explanations.28
Kane asserts that the choice which is made by an agent must be probabilistically and, thereby, indeterministically caused by the reasons for making that choice,29 where both the choice30 and the reasons31 for which it is made have corresponding neural connections or brain-wave patterns. But what do indeterministic, probabilistic causes explain about the making of a choice that is not teleologically accounted for by the reasons for making it?32 Kane’s view seems to be that causation involving an agent’s reasons is required as a link to explain why the choice is “the agent’s doing rather than a mere happening”.33 Three responses are in order.
First, if Kane is claiming that causation is required to make the choice an action, he is simply mistaken. As I explained in setting forth my account of the nature of choice in the previous section, the view that a choice is intrinsically active is most plausible in light of both libertarian and epistemological considerations about action.
Second, if Kane is claiming that causation is necessary to account for why a choice is mine, as oppposed to some other agent’s, he is again mistaken. As I also asserted in Section I, the view that a choice is an exercising of the mental power to choose, and that a mental power is an ultimate and irreducible property of an agent, is plausible in its own right. In addition, it is supported by the view that the essences of physical particles are constituted, at least in part, by (1) their causal powers to attract and repel other particles with similar powers and (2) their liability to be moved. Moreover, one would want to know from Kane what makes the reasons which, on his view, probabilistically cause my choice, my reasons as opposed to some other agent’s. If Kane claims that these reasons just are my reasons (a view which seems as eminently plausible as that about choice which he rejects), then it is incumbent upon him to explain why it is that the reasons which probabilistically cause a choice can just be an agent’s, but the choice made for those reasons cannot. Some property of an agent
must essentially characterize it, and there is no better candidate for such a property than the power to choose (though the power to choose need not be the only such property).
Finally, I believe it is plausible to maintain that probabilistic causation is not a form of explanation at all. It is no more than a description of how often an event of a particular kind occurs in light of its explanation. Kane, himself, seems to realize this. Thus, he states that “while the probabilistic explanation tells us which outcome is more likely to occur, it does not tell us why one occurs rather than the other in the particular case; and this is true whether the outcome that occurs in the particular case is more probable or less probable.”34 Kane also acknowledges that there is a teleological explanation of a particular choice.35 Thus, what is more or less probable is whether a particular choice will be made by an agent for a relevant reason in light of his past record of having or not having made choices of the same kind in similar situations.
Randolph Clarke is a third libertarian who maintains that a non-causal libertarianism cannot adequately account for free will (choice). According to him, free will is a certain kind of agent-control, where agent-control is fundamentally a matter of what does and does not causally produce a free act of will.36 Clarke notes that it is possible to maintain that agent-control is a matter of the complete absence of causation of an action. Like most other libertarians, however, he thinks a non-causal libertarianism is inadequate37 and causation in the form of agent-causation must be invoked to account for free will. As best as I can ascertain, Clarke’s reason for believing that a non-causal libertarianism is inadequate is an argument put forth by Donald Davidson that reasons-explanation must be understood as a form of causal explanation. Clarke maintains that no non-causal theory has met the challenge of this objection.38 Therefore, in what follows, I will summarize and respond to Davidson’s argument.
According to Davidson, if reasons-explanation is not a form of causal explanation, then there is no way to account for the distinction between having a reason to choose and choosing with it, and having a reason to choose and choosing because of it. One may justify a choice by citing a reason one has even if in choosing one did not choose because of it, but one cannot explain a choice by citing a reason unless one chooses because of it.39
To illustrate this objection, consider an executioner named Smith. Smith is ordered to put to death a certain individual, Charles. Charles was caught in the act of, confessed to, and was convicted for murdering a woman who happened to be Smith’s wife. Smith has two reasons for putting Charles to death. On the one hand, he believes that it is his job and moral duty to put to death persons guilty of the heinous crime of murder. On the other hand, Smith desires to get revenge against Charles by killing him. Smith has more than one reason to kill Charles. Both reasons justify the act. After putting Charles to death, Smith tells a reporter that he chose to put Charles to death
Naturalism and libertarian agency 165
in order to carry out what he believed was his job and moral duty, and he did not choose to kill him in order to get revenge. According to Davidson, this can only mean that Smith’s having the one reason caused him to make the choice and his having the other reason did not.
On a teleological account of how reasons explain an action, Smith chose to put Charles to death in order to carry out what he believed was his job and moral duty and did not put Charles to death in order to get revenge. In saying that Smith chose in order to do the former but did not choose in order to do the latter, it is the case that the former was an end for which he chose and the latter was not. The former’s being an end for which Smith chose is no less objective than its being the case that one event was the cause of an effect and another was not.40
A fourth libertarian who maintains that teleological explanation alone is not adequate to explain libertarian choice is Timothy O’Connor.41 He maintains that there must be causation to link the reason for making a choice with the choice, because he, too, thinks that teleological explanation alone cannot account for the distinction between having a reason and choosing with it and having a reason and choosing because of it. According to O’Connor, the requisite causation is agent-causal in form. He asserts that an agent has the causal capacity to cause the coming to be of an intention to act, where the coming to be of an intention is an event-part of a choice which is made for a reason. He claims that the causation by the agent of the coming to be of an intention has no cause and is “dependent upon the reason he has . . . for acting
[choosing] in that way. . . . For the agent’s free exercise of his causal capacity provides a necessary link between reason and action, without which the reason could not in any significant way explain the action.”42
Contrary to what O’Connor believes, if there is a problem with explaining how a reason explains an uncaused choice, there is a problem with explaining how a reason can explain an agent’s uncaused exercising of his power to cause the coming to be of an intention, where the exercising of this power depends upon the reason for the choice of which the coming to be of the intention is a part. Because there is no problem for the non-causal libertarian view which maintains that a reason alone can explain an uncaused choice, agent-causation is superfluous to an adequate account of freedom.
The last libertarian who contests a non-causal libertarianism of the kind I advocate is Ginet.43 Like Clarke and O’Connor, Ginet believes that there must be some way to account for the distinction between having a reason and acting with it and having a reason and acting because of it. Ginet maintains that with a typical bodily action done for a reason, such as opening a window in order to let in fresh air, an agent S intends of that action that by it she will let in fresh air. The intention explains the action in virtue of its content referring directly to the action. In this example, the content of the intention is ‘by this opening of the window I will let in fresh air’. Moreover, it is not
required of the intention that it play a causal role in the opening of the window. It is possible for S not only to have a reason to open the window but also to have a reason for keeping the window closed (e.g. in order to stay warm). In this case, the agent acts because of the first reason and only with the second in virtue of the specified intention.
The intention allegedly forges a link between the reason for which the action is done and the action itself. To illustrate Ginet’s position, consider an example of his where S urgently needs her glasses which she has left in a person, R’s, room where R is now sleeping.
S has some desire to wake R, because she would then have R’s company, but also some desire not to wake R, because she knows that R needs the sleep. S decides to enter R’s room in order to get her glasses, knowing as she does so that her action will satisfy her desire to wake R. Could it nevertheless be true that S did not intend of her action that it wake R? . . . It seems right to say that S did not intend to wake R if S was so disposed that, had it turned out that her entering the room did not wake R, S would not have felt that her plan had failed to be completely realized, and she must then either wake R in some other way or decide to abandon part of her plan. And S’s being thus uncommitted to waking R is quite compatible with S’s expecting and desiring to wake R. 44
Before evaluating Ginet’s position, it is helpful to clarify how his libertarian account of freedom differs from my own. On my view, freedom pertains to an agent’s choice, and while the concept of intention provides an explanatory connection between the choice and the chosen action, it provides no such connection between the reason and the choice itself because the concept of choosing for a reason is primitive in nature.