I can see the broken window, but can I also see why the window is broken?  In this ambitious and interesting paper, Church argues for an affirmative answer to this question.  Just as we can have perceptual knowledge of a state of affairs, so too can we have perceptual knowledge of the reason for that state of affairs.  Her defense of this claim proceeds in roughly two stages.  First, she develops an account of perceptual knowledge, i.e., an account of what makes a given piece of knowledge perceptual rather than non-perceptual.  Second, she shows how this account of perception can be extended to the case of reasons.

The first stage of her argument depends on three key claims:

(1) Perceptual knowledge can be distinguished from non-perceptual knowledge in terms of justificatory immediacy: “perceptual knowledge is immediate knowledge in the sense that it does not depend on any other knowledge for its justification.” (640) 

But unless we develop an explanatory account of perceptual immediacy, Church worries that this claim may be construed in a deflationary way.  She thus argues that:

(2) Perception requires the experience of objectivity: “An experience is a perceptual experience precisely when the independent reality of its object is evident from within that experience.”  (644)

This in turn requires her to explain how we can achieve this experience of objectivity (what I will call experiential objectivity).  How can we experience a state of affairs as existing independently of our experience of it?  Her answer, which draws on an intriguing claim by P.F. Strawson that one’s perception of an object is “infused with … other past or possible perceptions of the same object,” is to invoke imagination:

(3)  Our experience of something as objective depends on our imagining alternative perspectives of it.  Via the imagination, we can occupy perspectives and modalities different from the ones we are presently occupying, and it is these imaginings that serve to ground experiential objectivity.

It is important to recognize what a strong claim (3) is.  Church is not merely invoking our capacity to imagine alternative perspectives in an effort to explain experiential objectivity; rather, she explicitly claims  that “we actively imagine alternative perspectives whenever we experience something as an objective state of affairs.”  (649)  Her argument is largely a transcendental one:  “it is only by imagining alternative perspectives and imagining no-actual possibilities that we could perceive states of affairs, because only then could the objectivity of what is seen be evident from within perception.” (659)  The transcendental considerations, however, are supplemented by empirical ones, for she also suggests that we have good reason to believe that we engage in such imaginings based on the phenomenology of perception.  As she notes, we are not typically aware of such imaginings (one might ask: are we ever aware of them?), but introspective exercises suggest that “we can transform our ordinary experiences of objects into experiences of patches of light and color … by ceasing to imagine alternative perspectives.” (658)  Here, though I am sympathetic to her claim, I would have liked to have heard a bit more.  She rightly notes, in my view, that there is a phenomenological (and not merely cognitive) difference between seeing something as two-dimensional and seeing it as three-dimensional (649), but more could be said to show that active imagining is involved in the explanation of this phenomenological difference.

In the second stage of her argument, Church applies the account of perception that she has developed to the case of perceiving reasons.  She first focuses on explanatory reasons.  When we see why the window is broken in addition to seeing that the window is broken, we must see what explains the broken window, and we must also see it as explaining the broken window.  Explanations might be causal, constitutive, or some combination of the two.  With respect to causal explanations, Church discusses three different analyses of causality: (1) Regularity; (2) Energy transference; (3) Counterfactual, but she focuses largely on the third of these.  How could we perceive causes if to do that we have to perceive counterfactual possibilities?  To see that C caused E—that E would not have occurred if C had not occurred—would seem absurdly to require that we see the non-actual case in which C fails to occur.  The problem is easily solved, however, on Church’s analysis of perception in terms of active imagining: “the non-actual alternatives are alternatives that, when actively imagined, can inform and infuse our current experience in such a way as to make causality perceivable.” (656)

Turning to constitutive explanations, Church offers a similar story.  We explain the shape of an object in terms of its atomic structure, which in turn involves constitutive dependencies: the object would not have the shape that it does if its constituent parts were differently arranged.  What enables us to see these constitutive dependencies is our active imaginings of various alternatives.  As Church notes, not all of these constitutive dependencies are easily seen, and this can be explained by the fact that they cannot be easily imagined.  For example, when we see why a particular diamond is hard, it is not enough merely to have various beliefs about atomic alignments and so on, nor can we simply imagine geometrical structures.  Rather, “our view of the diamond must be informed and infused with our imaginings of various changes in atomic structure and their various effects.” (660)

After extending her account of perceiving explanatory reasons to the case of perceiving justificatory reasons, Church turns in the final section of her paper to a discussion of three advantages of seeing reasons, as opposed to gaining knowledge of them some other (indirect or reflective) way.  First, if our knowledge of reasons is perceptual, and hence immediate, it relies much less heavily on memory than inferential knowledge does, and this gives it greater security.  Second, via perception we are well placed to generate further knowledge from the knowledge that we already have.  Perception makes certain facts evident in a way that facilitates new discoveries.  Finally, we are typically more motivated by beliefs gained through perception compared to beliefs gained through inferential reasoning.  The importance of this fact becomes clear when we consider both moral contexts and other contexts where quick action is advantageous.  As Church notes, “Insofar as increased responsiveness to the world and to others is desirable, then seeing reasons will be preferable to understanding that remains non-perceptual.” (667)

Reviewed by Amy Kind

Claremont McKenna College

Main authors discussed: Gregory Currie, Ian Ravenscroft, David Velleman, Shaun Nichols, Stephen Stich

Some philosophers (e.g., Currie and Ravenscroft, Velleman) have recently argued that pretense provides a counterexample to the standard Humean Theory of Motivation (HTM). Pretense behavior is a novel kind of mental state, namely, an imaginative analogue of desire, that motivates action. For example, when a child pretending to be a Jedi Knight swings a tree branch in his hand, his action is motivated by his imaginative desire to destroy his opponent with his “lightsaber.”

Funkhouser and Spaulding (F&S) reject this anti-Humean explanation of pretense action.  Although they are willing to grant the existence of imaginative desires (302), they deny that such states provide the motivation for pretense. On their view, we need not reject the spirit of Humeanism about psychological motivation to explain the pretender’s motivation; rather, we need only reject the traditional and overly narrow HTM. Their own account of pretense is given by their Guider-Motivator Thesis: “For every intentional action, there is a guider-motivator pair that both causes and rationalize that intentional action” (307).

Before making their positive case for this thesis, F&S respond to objections against the standard Humean account of pretense (most notably associated with Stich and Nichols). One class of such objections suggest that, by explaining all of a pretending child’s actions in terms of a desire to pretend, such an account over-intellectualizes the pretense behavior, leaves the pretending child outside the pretense, and denies the child’s creativity. In response, F&S argue that a child who is motivated by the desire to pretend need not have the concept of pretense. Children need only have a behavioral, not a mentalistic, understanding of pretense in order to engage in pretend behavior, i.e., “their desire to pretend that p is simply a desire to behave, loosely as if p were the case” (300). Relatedly, a child motivated by the desire to pretend may engage in pretend behavior without consciously consulting or being cognizant of that desire (301). Though Velleman suggests that a Humean account of pretense makes a child who pretends seem depressingly adult-like, F&S astutely counter that “It is not depressing, but rather comforting, that pretending children retain a relatively firm grip on reality” (301-2); otherwise, their pretend actions would be hard to distinguish from the delusional actions of a schizophrenic. With respect to creativity, F&S note that the Humean’s denial that imaginative desires motivate pretense is comaptible with the claim that children who pretend “have a genuine desire to act out their imaginings,” and this means that their creativity is limited only by their imagination (302).

In addition to defending Humeanism against these objections, F&S offer criticisms of the anti-Humean account of pretense. For reasons of space, I will here consider just one of these criticisms, namely, their worry that imaginative desires are not intrinsically motivational and hence cannot be motivational at all. As they argue, “Motivation is typically an intrinsic and essential feature of motivating states. For example, the motivational component of a desire is, at least typically, intrinsic to it” (303). But the imaginative desires that we experience in contexts other than pretense, such as when we engage with fiction or daydream, do not seem to have motivational power at all. Thus, the anti-Humean must claim “that imagination has distinctive motivational powers in the context of pretense” (303), and F&S find this claim implausible.

Here I worry a bit about their argument. In a footnote, F&S consider examples of various kinds of desires that seem not to be motivational. What should we think about desires that the agent believes are impossible to satisfy, such as the desire to square the circle or to travel to a distant galaxy? F&S note that they are inclined to deny that such states are really desires; rather, they are mere wishes. Dealing with such counterexamples by terminological fiat seems to me unsatisfying. Moreover, I doubt that the anti-Humean must deny that imaginative desires are intrinsically motivating. Our desires only move us to action in the context of the appropriate background beliefs, and one might think the anti-Humean could suggest something similar about imaginative desires.

After making their case against the anti-Humean view, F&S attempt to spell out their own view of the role of imagination in pretense in their defense of the Guider-Motivator Thesis. Rather than seeing intentional action as caused or rationalized by belief-desire pairs, F&S see it as being caused by a guider-motivator pair, where “motivators provide the thrust or pick the destination, and guiders direct or tell us how to get there” (307).

F&S thus part from traditional Humeans by broadening the class of mental states that can cause or rationalize action; on their view, imaginings as well as beliefs can serve as guiders. Importantly, this claim is also compatible with the views of their anti-Humean opponents who tend to claim not only that some imaginings can serve as guiders, but also that some imaginings can serve as motivators. It is this latter claim that F&S argue against in their paper. According to F&S, imagination can serve only a guiding role when it comes to pretense.

In the final part of their paper, F&S introduce the notion of a script for the class of guiders that provide a model for action: “A script, unlike the typical belief used in means-end reasoning, is something that can be imitated or enacted. Significantly, these scripts can sometimes guide action without being mediated by a belief or knowledge state” (309). Imagination can be one such script, but others can also guide action, and some of these might be external to the agent, such as when an actor’s behavior is guided by her reliance on a physical script, or when children’s pretense behavior is guided by the behavior of other children or various social norms. In accepting the existence of external scripts as well as internal ones, F&S commit themselves to a version of the extended mind thesis advanced by Chalmers and Clark.

On the whole, I find the arguments that F&S make against the anti-Humeans compelling, and I wholeheartedly agree with their claim that Humeanism is best defended by adopting a pluralistic rather than unitary account of the motivation for pretense. Though I find less satisfying their own positive view of imagination which considers it on analogy with a script, I nonetheless found their discussion of scripts to provide some important insights into the complexities of the interplay between imagination and pretense, and I recommend this paper as a must-read to anyone thinking about pretense, the imagination, or the Humean theory of motivation.

Reviewed by Amy Kind
Claremont McKenna College