Motivation as an epistemic ground

Abstract

In several papers, Mark Wrathall argued that French phenomenologist, Maurice Merleau-Ponty, identifies a sui generis type of grounding, one not reducible to reason or natural causality. Following the Phenomenological tradition, Merleau-Ponty called this form of grounding “motivation,” and described it as the way in which one phenomenon spontaneously gives rise to another through its sense. While Wrathall’s suggestion has been taken up in the practical domain, its epistemic import has still not been fully explored. I would like to take up the epistemic dimension of Wrathall’s thought in this paper. Following Wrathall, I explain how motivation can help us understand the manner in which perceptions ground singular, experiential judgments. But I extend this work in two ways. First, I suggest some additional considerations that support Wrathall’s view. Second, I argue that motivation can also help account for the way perceptions ground general judgments. My aim here will not be so much to rule out other answers to these longstanding epistemological questions, as to show that motivation carves out an attractive epistemological space.

In several papers, Mark Wrathall (2005a, 2005b) argued that French phenomenologist, Maurice Merleau-Ponty, identifies a sui generis type of grounding, one not reducible to reason or natural causality. Following the Phenomenological tradition, Merleau-Ponty called this form of grounding “motivation,” and described it as the way in which one phenomenon spontaneously “triggers” another, “not through some objective causality, … but rather through the sense it offers.”¹ While Wrathall’s suggestion has been taken up in the practical domain—as a ground for action—its epistemic import has still not been fully explored.² I would like to take up the epistemic dimension of Wrathall’s thought in this paper. Following Wrathall, I explain how motivation can help us understand the manner in which perceptions ground singular, experiential judgments. But I extend this work in two ways. First, I suggest some additional considerations that support Wrathall’s view. Second, I argue that motivation can also help account for the way perceptions ground general judgments. My aim here will not be so much to rule out other answers to these longstanding epistemological questions, as to show that motivation carves out an attractive epistemological space.³

1 What is Motivation?

First of all, let’s be clear about the two forms of grounding to which I propose motivation as an alternative: reason and natural causality. The distinction I have in mind here can be found, for example, in McDowell’s distinction between the logical space of reason and the logical space of nature. As Sellars puts it, “In characterizing an episode or a state as that of knowing, we are not giving an empirical description of that episode or state; we are placing it in the logical space of reasons, of justifying and being able to justify what one says” (1997, 76). In other words, the space of reason is a normative space, in which we are fundamentally concerned with grounds for knowledge, i.e., with justification. In contrast, McDowell defines the logical space of nature as the arena of the natural sciences, in which we are concerned not to justify, but rather to explain. Socrates sitting in an Athenian prison, for example, is not justified by bones and sinews—these explain his position—but by a conception of the good and how to realize it. To explain an event, then, is not to justify it with a reason, but to identify its natural causes. Natural causes operate, then, in a manner that is lawful but not normative.

Merleau-Ponty introduces the concept of motivation, in the Phenomenology of Perception, to describe the grounding processes characteristic of perception, which according to him (though not in these terms) do not fit squarely in the logical spaces of reason or nature. According to Merleau-Ponty, one element of the perceptual field grounds another, not in virtue of causing it or serving as a reason for it, but rather in virtue of motivating it. As a first pass, consider the following series of dots:

. .      . .      . .      . .      . .

These dots are almost invariably perceived as grouped into nearest pairs. But there is no evident reason for grouping the near dots rather than the far ones: nothing in the sensory properties of the dots per se supports the one grouping rather than the other. Obviously one could attempt to rationalize the grouping—to reconstruct reasons that would support it, e.g., “one should group the near dots”—but not only do we have no reflective access to such reasons, neither do they carry any real weight for us. On the other hand, it’s hard to see how anything about the sensory properties of the dots per se would cause this grouping. Doubtless, there is a causal story to be told about goings on in the brain that explain the perception, and surely the relative proximity of the dots will play a role in this story. But the claim we are considering is not about grounding processes going on in the brain. It’s about grounding processes going on in the perceptual field: What is the grounding relation between the perceptions of the dots and the grouping, such that from the first we see them as grouped in just this manner? Nothing about the sensory properties of the dots force upon us the one grouping rather than the other.

Or take the following case, invoked by Wrathall. “It took centuries of painting before the reflections upon the eye were seen, without which the painting remains lifeless and blind …. The reflection is not seen for itself, since it was able to go unnoticed for so long, and yet it has a function in perception, since its mere absence is enough to remove the life and the expression from objects and from faces. The reflection is only seen out of the corner of the eye. It is not presented as the aim of our perception, it is the auxiliary or the mediator of our perception. It is not itself seen, but makes the rest be seen” (2012, 322–3). Now, the reflection in the eye seems not to be a reason to perceive liveliness, according to Merleau-Ponty, because the former can ground the latter without one’s being able explicitly to attend to it or identify it as one’s ground. On the other hand, it’s unlikely that the reflection causes the perception. It seems plausible that the reflection grounds the perception in virtue of its familiar perceptual meaning: nothing about a white splotch on a black splotch per se gives rise to a perception of liveliness, but the perceptual meaning under which it is intended (namely, “reflection in the eye”) does. However, it is generally agreed that natural causation does not rely on meanings—there is no meaning operative in the way one pool ball imparts kinetic force to another (e.g., there is no role that “striped,” “blue,” “the 10 ball,” etc. can play in this interaction).

While these examples are far from decisive, they should suffice to give Merleau-Ponty’s notion of “motivation” some initial plausibility. And if they are indeed examples of motivation, they allow us (provisionally) to note some distinctive features of this form of grounding.

First, motivation is spontaneous. I mean this in the sense Merleau-Ponty does when he writes that the perception of the near dots as grouped is a “spontaneous valuation” (ibid., 465). In other words, Merleau-Ponty distinguishes the kind of grounding characteristic of the grouping, as spontaneous, from the active form of grounding characteristic of reasoning or the passive form of grounding characteristic of natural causality. When we reason, we actively give and take reasons and revise our judgments and decisions in response to those reasons. When we are caused in some way, we simply undergo an effect. But motivation is not quite like either of these options: I don’t simply undergo the perception (again, nothing in the sensory properties of the dots requires the grouping), but neither do I actively decide on it (as evidenced by the fact that I cannot simply see a different grouping should I wish to). Rather, I spontaneously arrive at the perception.

Second, like reasoning, but unlike natural causality, motivation is a relationship between meanings.⁴ This is not to say that the sorts of meanings that populate perception are just like those that populate reasons (presumably, the sorts of meanings that propositions are composed of); the perceptual sense of “reflection in the eye” need not be exactly like its corresponding conceptual sense. Whatever the proper characterization of this perceptual sense, it seems to be in virtue of just this sense or meaning that the perception of the reflection in the eye grounds the perception of liveliness.

Third, at least according to Merleau-Ponty, and like natural causality but unlike reason, motivation can be implicit: I needn’t attend to a motive for it to give rise to its motivatum. Now, obviously, there is little agreement that reason, in contrast, does need to be explicit.⁵ I tend to think it does, but I will leave this criterion aside.⁶

Fourth, and again like reason but unlike natural causality, motivation can be normative. When I am motivated, in an appropriate way, to have a certain perceptual experience (e.g., when the reflection in the eye motivates a perception of liveliness), I not only have that perceptual experience, but I find it right so to perceive. If someone tells me they experience the world otherwise, I am not indifferent to this fact; I think that one of our experiences must be mistaken. This conviction is conspicuously absent with regard, e.g., to our imaginings—presumably because imagination lacks normative import. One way to articulate this natural conviction would be to say—to paraphrase Pryor’s description of norms for belief—“When you are in conditions C, you should have a perceptual experience of P.”⁷ If perception is grounded in processes of motivation, then motivation would seem to be responsible for this sense of normativity.⁸

So, in brief, we can say that motivation is a form of grounding that is spontaneous, normative, and is a relation between meanings (though not explicit ones, necessarily). In contrast, natural causation is passive, not a relation between meanings, and not normative. On the other hand, like motives, reasons are normative and relations between meanings, but unlike motives, are active (and by some accounts must be explicitly held in order to function as reasons). Let this suffice for an initial characterization of motives.

2 Singular Judgments

Now let’s turn to the epistemic import of motivation, starting with the most basic epistemic role played by perception: the relation between a perception and a singular judgment about that perception. By “singular judgment,” I just mean a judgment about a particular individual, in this case, the individual experienced in a perceptual experience. I will only be interested in singular judgments about individuals that are experienced and that are based on that experience, and so—while singular judgments need not be about experienced individuals or based on experience—I will use “singular judgment” to refer only to this kind of judgment. I contrast singular judgments with general judgments, which are not about any particular individual.

There are familiar accounts of the relation between perception and singular judgments in terms of both reason and natural causality. McDowell, for example, holds that the relation between perceptions and singular judgments must lie within the space of reasons if our beliefs are to be normatively constrained by the world: otherwise perception provides only exculpations for beliefs, not justifications (1996, 8). In contrast, in a well known argument, Davidson claimed that only propositional attitudes can count as reasons: “nothing can count as a reason for holding a belief except another belief” (1986, 310). Of sensation, he writes, “The relation between a sensation and a belief cannot be logical, since sensations are not beliefs, or other propositional attitudes” (ibid., 311).⁹ Instead, the former must simply cause the latter.

Now, we might object to assumptions made by either of these arguments. On the one hand, many think that perception is a propositional attitude. On the other, one might doubt that perception really does normatively constrain our beliefs. Let us assume, however, for the moment, that both of these premises are good ones—as, indeed, I think they are. Whereas these premises lead Davidson and McDowell to opposite conclusions, I would like to suggest, following Wrathall, that there is a way to make them compatible. But it will require us to think of the relation between perception and experiential judgments in terms of motivation. I have suggested that motivation is both normative and a relation between meanings that may or may not be propositional. If this is true, we can allow both that.

A) perceptions, despite being non-propositional, ground singular judgments, and

B) perceptions ground singular judgments in a normative manner.

In contrast, if we describe this relation in terms of natural causality, we can accept A) at the price of rejecting B), since natural causality is not intrinsically normative; if we describe it in terms of reason, we can accept B), but have to reject A).¹⁰

Now let me say a bit more about why we might make these two assumptions. In my view, both assumptions articulate an intuitive account of the relation between perception and judgment.

On the one hand, there are a number of reasons to think that perception does not have propositional content.¹¹ One such reason can be couched in terms of Husserl’s distinction between meaning-intention and meaning-fulfillment. I can think a proposition—a proposition can have meaning for me—without my being actually presented (for example, perceptually) with the state of affairs that proposition is about. In this sense, a thought of a proposition is intrinsically “empty,” but can be "fulfilled" by perceptions; acts of thinking propositions are “significative,” whereas perceptions are “intuitive.”¹² As Hopp puts it, propositions are “detachable” from perceptions, and so perceptual contents cannot be defined simply in terms of propositions.¹³

One might also object to Davidson’s claim that there is no way for mental states with only non-propositional content to serve as reasons for beliefs. Perhaps this is too high a bar for what counts as a reason. Were Davidson wrong, then reason could account for feature A) of the relation between perception and empirical judgments. But Davidson’s position is highly plausible. If perception does not have the propositional content “That’s an owl,” one way to describe the content of perception would be in terms of a simple conceptual content, e.g., “owl.” McDowell seems to think something like this, in holding that the content of perception is conceptual but not exactly propositional.¹⁴ But the sole content “owl” cannot serve as a reason for judging "That's an owl." Presumably, "owl" doesn't have a truth value in the manner that "That's an owl" does, and even if it did, "owl"'s justificatory import would be underdetermined: "owl" could serve as a reason for "That's an owl" or "That's not an owl."¹⁵ I’m not convinced one should accept this account of perception, though, and to my mind Davidson’s view becomes especially compelling when couched in terms of the distinction between significative and intuitive content just suggested. Reasons, I take it, are relations between "significative" mental states: I must be able to affirm a reason without being perceptually or intuitively presented with it. But perception—if it’s not a propositional attitude—is an intuitive mental state (which is not to say it doesn't have a kind of meaning or sense of its own). It seems to follow that if perception is not propositional, then strictly speaking it cannot serve as a reason for propositions.¹⁶

On the other hand, it seems true that the relation between perception and judgment must be normative. Not only is such normativity a strong theoretical desideratum—as McDowell puts it, without this kind of normativity our beliefs seem to be a “frictionless spinning in a void”—but it seems phenomenologically true that this relation is normative (1996, 11). There is a phenomenal difference between my holding the belief, "That's an owl," i) emptily, ii) in conjunction with perceiving an owl, iii) in conjunction with imagining an owl.¹⁷ The difference between i) and iii) on the one hand, and ii) on the other, is that we, so to speak, are right or wrong to believe "That's an owl" in case ii). Assuming we have no outside sources of evidence (e.g., testimony), there is no such normative demand in i) or iii). The reason why perception does this, in my view, to use the language above, is precisely because it fulfills, or provides intuitive content, to a proposition. My perception of an owl grounds (I would say, motivates) the judgment “That’s an owl,” precisely because it fulfills that judgment. Fulfillment is normative: it is prima facie right to affirm a fulfilled judgment, and not prima facie right to affirm an empty one.¹⁸

Again, there is room to question any of these premises. Assuming these views have an initial intuitive appeal, though, we have a strong motive to seek an account that can accommodate both these features, as motivation does. At a minimum, describing the relation between perception and judgment in terms of motivation opens up an interesting theoretical space.

This, as I understand it, is the thrust of Wrathall’s argument. But, quite independently of allowing us to reconcile compelling aspects of both McDowell’s and Davidson’s accounts, motivation makes sense of a number of other features of the relation between perception and experiential judgments. Specifically, it uniquely allows us to accommodate the following features of this relation:

C) The relation between perception and empirical judgment is not active.

D) Perception and empirical judgment relate to each other in virtue of their meanings.

Let me explain these claims.

Suppose I see an owl outside a window at a friend’s house, and on the basis of this perception, judge “There’s an owl.” First, note that perception does not simply bring it about that we judge one way or another. I can perfectly well perceive the owl without stopping and forming the belief “There’s an owl.” Indeed, if I have good reason to doubt the perception—say, my friend tells me that, in fact, it’s only a fake, meant to scare away squirrels—then I can easily judge, “That’s no owl.” Still, lacking such a countervailing reason, perception ordinarily does incline or dispose us to judge in a certain way. So, let’s say that the relation between perception and judgment is not simply passive.

But neither does it seem to be active. Start with a less controversial point: whereas I am free to actively affirm or deny a judgment, I am not so free to affirm or deny a perception.¹⁹ When my friend tells me that that is not a real owl, I can, as we said, judge “That’s no owl”—but I may or may not succeed in revising my perception. I mean this in the sense that when I see an optical illusion, I may know the illusion to be false but continue to perceive the illusion. In the owl case, it may happen that I suddenly notice the crudity of the imitation, that false shine of its feathers, etc., and succeed in seeing it for what it is. But I may also fail to achieve this perception: it may stubbornly appear to be an owl, despite my judgment to the contrary. So perception is not revisable in the way judgment is: I cannot decide to see a fake—this perception either pops into view or it does not. In other words, whereas I have active control over my judgments, my perceptions are spontaneous.²⁰ And this quality of perception spills over to the relation between perception and judgment. If I fail to see “the owl” as a fake, then my perception will continue to dispose me to form the judgment “That’s an owl,” even though I actively bracket this disposition in forming my conclusion. Once I learn about the fake, my perception does not simply become indifferent, but stubbornly persists in supporting its judgment. A good indicator of this is the subtle unease we feel in situations where we know the world is not as we perceive it: while I will no doubt judge “That’s no owl,” part of me will likely feel unsure until I succeed in seeing “the owl” as a fake. It’s hard to see why I would feel this unease if perception didn’t continue to dispose me to judge “That’s an owl.” Just like my perception, then, the relation between perception and judgment is spontaneous.

Someone might object that if I am able to bracket my disposition to judge “That’s an owl,” then I do actually have active control over this disposition. But this isn’t quite right. I do have the ability actively to bracket the relation between perception and judgment, but I don’t have the ability to revise it—at least not the way I can revise my judgments. In other words, all I have active control over is whether or not to accept the disposition to judge that perception gives me. I don’t have the ability to revise what perception disposes me to judge. Thus, the relation between perception and judgment—the disposition—is not under my active control. To say otherwise would be a bit like saying that, because I have the ability to decide whether or not to accept someone’s love for me, I have the ability to decide whether or not they love me.²¹

Thus, we should say:

C) The relation between perception and empirical judgment is not active.

But were the relationship reason-constituting, it would have to be active. Consider McDowell’s articulation of what it takes for a relation to stand within the space of reasons. According to McDowell, for a relation to count as “reason-constituting,” it “must be able to come under the self-scrutiny of active thinking,” in the sense of being “liable to revision, if that were to be what the self-scrutiny of active thinking recommends” (1996, 52–3). As we have seen, the relation between perception and judgment does not meet this criterion.

Of course, one can actively judge about one’s perception and employ this as a reason. For example, I could judge “I perceive an owl” and infer from this that “That’s an owl.” And one may have active control over the former judgment. Upon learning the figure is a fake, I can judge “I do not in fact perceive an owl.” But this response really just defers the problem, because now the question is about the relation between my perception and my judgment that “I perceive a ….” The same considerations would apply to this relation as to the simple relation between my perception and my judgment that “That’s a ….”

On the other hand, if one took a more epistemically externalist approach to warrant than McDowell, C would not appear very compelling. Take Burge’s (2003) distinction between justification, which requires access to one’s warrant, and entitlement, which does not. On such an account, we might be entitled to a singular belief on the ground of perception simply by something like reliable veridicality. Obviously, active revisability is not a condition for this kind of warrant. My point here, though, is not to dismiss every alternative to my account, only to draw the contours of what motivation can do philosophically. Part of what it can do is to accommodate A and B in an internalist mode. It is this philosophical space that I wish to explore, and this space seems also to require C.

Let us now turn to feature D, namely that perception seems to ground judgment in virtue of its meaning. My perception of an owl does not ground the corresponding judgment simply as the sensation of various colors and shapes, but precisely because it is a perception of an owl as an owl. A good indicator of this is that if I don’t know what an owl looks like, then perceiving the owl could not fulfill the judgment “That’s an owl.” I might be able to believe that that over there is an owl, but the basis of this belief couldn’t be my perception.²² What I mean is that only in virtue of its perceptual meaning “owl”—again, however exactly we should describe perceptual meanings—does my perception ground my judgment. Thus,

D) Perception and judgment relate to each other in virtue of their meanings.

This feature could be accounted for by placing the relation within the space of reasons. The only mental states that can serve as reasons, I presume, are those composed of meanings.²³ For example, pure sensations can’t serve as reasons. A bare sensation of pain cannot serve as a reason for thinking I have been stabbed, though presumably the belief “I feel pain” can. Similarly, the bare sensation of shapes and colors does not serve as a reason to judge “That’s an owl” (though a belief, e.g., “I see such and such shapes and colors” could serve as a reason for judging “That’s an owl”). As Davidson puts it, “Not all causes are reasons: the activation of our retinas does not constitute our evidence that we see a dog, nor do the vibrations of the little hairs in the inner ear provide reasons to think the dog is barking” (2005, 135).

But while reason can account for feature D of the relation between perception and judgment, natural causation cannot. As we already saw, natural causes seem not operate in virtue of their meanings, and so if perceptions are supposed to cause judgments, it’s unclear how it would do so in virtue of its meaning.

Thus, I would suggest, we have good reason to seek a motivation view of the relation between perception and judgment. Only motivation can account for features A-D of this relation. While reason can account for B and D, and causation can account for A and C, neither can account for all. Needless to say, we might have good reasons to doubt any of these features. But, at least initially, they map an intuitive account of this relation, and so we ought to take seriously a theoretical framework that allows us to accommodate all four features, as motivation indeed does.

3 General Judgments

Let us now consider how motivation can help us think about general judgments. My claim will be that, just as motivation can offer an interesting alternative to existing accounts in the case of singular judgment, so it can offer an interesting alternative in the case of general judgment, including those necessary, universal judgments traditionally termed “a priori.”

A central motivation for the Rationalist position—namely that some knowledge must derive its justification independently of experience—has always been that we seem to know certain propositions that perception is unable to justify. Take one of Bonjour’s examples of putatively a priori knowledge, that “There are no round squares.” Certainly, I don’t perceive any round squares, but more, in principle I cannot perceive a modal truth of the type “squares cannot be round.” This, I take it, is also the objection to Empiricism that Leibniz voiced when he wrote:

Although the senses are necessary for all our actual knowledge, they are not sufficient to provide it all, since they never give us anything but instances, that is particular or singular truths. But however many instances confirm a general truth, they do not suffice to establish its universal necessity; for it does not follow that what has happened will always happen in the same way. … From this it appears that necessary truths, such as we find in pure mathematics and particularly in arithmetic and geometry, must have principles whose proof does not depend on instances, nor consequently on the testimony of the senses, even though without the senses it would never occur to us to think of them… . (1996, 49).

Bonjour proposes a different argument. He notes that there is a broad class of knowledge that perception cannot directly justify, e.g., generalizations or laws of nature. If perception is to justify these beliefs indirectly, then we will need some set of principles for moving from particular perceptions to these general laws. But such principles will themselves be general, and so cannot be directly justified by perception. In order to avoid a regress, such principles will have to be known a priori.²⁴

Thus, though they may reach this conclusion by different paths, Rationalists share the view that:

E) Not all knowledge is justified by perception.

If this is true, then such knowledge must rely on some kind of direct insight—not unlike the way perception provides us with intuitive content, but at an intellectual rather than perceptual level. Bonjour calls this kind of intuition “rational insight.”

Responses to Bonjour’s argument typically come in two parts: a) the suggestion of an alternate, experience-based account of the justification for seemingly a priori knowledge; b) considerations in favor of this alternate account. Consider Devitt’s (2005) arguments against Bonjour. First, he suggests an account of justification on which justification is not a one–one relation between a perception and a belief, but a matter of a system of beliefs as a whole facing the tribunal of experience. Second, he argues that this account is preferable to the Rationalist’s, since the latter account is deeply obscure.²⁵ While we have at least a general and intuitive idea of how empirical justification is supposed to work—namely, the worldly fact that p makes the belief that p true, and experiences produced by this fact are involved in justifying this belief—no parallel account can be supplied for rational insight. While this objection is surely not decisive, it does mitigate against the Rationalist view.

There is a core insight to this kind of response that makes it recognizably Empiricist, namely, that the touchstone of knowledge is that it passes the “tribunal of experience.” There are many ways of spelling out this insight: one might accept it while denying Devitt’s holism, for example. However we articulate this insight, the core idea is that:

F) Experience norms all our knowledge, including knowledge of general judgments.

One cost of not accepting F is having to accept either a) skepticism about a large portion of our knowledge, or b) an obscure faculty of direct intellectual insight. Now, of course, one might be willing to accept a) or b), and so deny F. On the other hand, one might think something stronger than F, namely that E is false and perception can justify knowledge, as Devitt thinks (and, I suppose, most Empiricists would). My intention is not to argue any more exhaustively for E and F than I did for A or B, but just to point out as I did with those claims, that E and F are not actually exclusive. There is a way to accommodate both claims, and motivation is the key to doing so. Motivation can reconcile E and F because it is a normative relation, but one not contained within the space of reasons, such that it allows a justificatory gap between experience and at least some general knowledge, but not a normative gap. In other words, if we allow that perceptual experiences motivate general judgments, then perceptual experience could very well norm our general judgments without fully justifying all of them.

Again, what does it mean to say that motivation is normative in this context? As I put it earlier, the idea is that perceptual experience inclines or disposes us to form certain general judgments. The manner in which it does this is not arbitrary, nor a simple matter of fact. Rather, in doing so, perception makes it (prima facie) right for us to form just those judgments. Again, compare the way in which a perception disposes me to form a general judgment and an imagination does. There is a phenomenal difference between the way my perception of a bird inclines me to judge that “Birds have wings,” and the way my imagination of a Jaberwock inclines me to judge that “Jaberwocks have wings.” As I suggested in section II, this difference lies in the way that perception fulfills the former judgment, whereas imagination does not fulfill the latter (at least in the same way).²⁶

But the relation between perception and general judgments—despite its normativity—also does not belong within the space of reasons. The same concerns that applied in section II apply here as well—namely that this relation is not fully active. But, in the present case, a number of additional considerations mitigate against taking this relation as rational, namely, the sorts of Rationalist considerations I explained above.

This is not to deny that, at a higher level, empirical and other judgments can be combined in logical ways to ground “a priori” judgments. Coherentist accounts of how this occurs may well be in the right, for example. It is just to say that, at a base level, the relation between perception and general laws need not be justificatory, but could yet be normative.

Note that the considerations pointed to by the Rationalist are not an obstacle for a motivation view. The motivation view could perfectly well allow for Bonjour’s argument. It may be that for perception to justify general knowledge, we would require prior knowledge of some third term, a principle for moving from particular perceptions to general propositions. But no principle is required for perception to motivate general judgments, since motivation is not a logical relation. No prior item of knowledge is required here: perception can independently motivate a general proposition.

On the other hand, even if Leibniz were right that perceptions of particular and contingent facts cannot justify universal and necessary claims, the former might still be able to motivate the latter. Take a general proposition, such as that when fire is applied to a cotton ball, under appropriate conditions, the cotton will burn. Of course, this general fact is not the content of any particular perceptual experience. Nevertheless, a particular perception can motivate the general proposition. Even if I see a cotton ball catch fire once, this perception may suffice to dispose or incline me to make the general judgment, and the more I have such perceptual experiences, the stronger this disposition will become.²⁷ Nor is this disposition arbitrary (though it could, of course, be mistaken): it is not as if it were an imagining of a cotton ball catching fire that was disposing me to make this judgment. I mean, the disposition is normative—even if it’s normativity is not as strong as justification—and it is normative because it is a response to the fact that my perception fulfills the general proposition.

An obvious objection at this point is that, if general propositions cannot be contents of perceptions, then how can a perception fulfill a general propositions? The phenomenological answer to this question is, perhaps unsurprisingly, fairly complicated, and is tied together with an account of how we form general ideas at all and what the contents of perception actually are—an account that would require an extended exposition. For now, my aim is simply to suggest that plausibly, perceptions can fulfill general propositions. Again, there is a phenomenal difference between merely thinking that “Kiwis have long beaks,” and forming this proposition in conjunction with perceiving (or remembering perceiving) some kiwis. It would be natural to chalk up this difference to the way the perception fulfills the proposition. I think one could give different answers as to how, exactly, this fulfillment occurs (one might, for example, attempt a more holistic account than the one provided here). One possibility, however, is suggested by Husserl. Briefly, his claim is that a perception can fulfill even necessary, universal judgments—such as “Nothing can be red and green all over”—only when its object is grasped as an exemplar, i.e., when grasped not in its particularity, but simply as, say, a colored object. Then, by varying the properties of the object in imagination, we can note essential limitations on its properties, e.g., wherever I extend the red surface on a ball, it ceases to be green, and vice versa.²⁸ If I grasp the ball as an exemplar of colored object, then I do not merely think that “no object can be red and green all over,” I actually have a particular kind of fulfillment for this thought, one based on a peculiar grasp of a perceived object, albeit mediated by imaginative activity. While more, obviously, could be said about how this is supposed to work, let the foregoing suffice as a sketch.²⁹

4 Conclusion

In this paper, I have examined the epistemic significance of introducing motivation as a sui generis kind of ground, and provided some considerations that support taking this ground to have an epistemic function. As I have argued, we should appeal to motivation as an epistemic ground because it allows us to accommodate a variety of highly plausible features of the relation between perceptions and judgments, whether singular or general. In contrast, these features can be only partially accounted for in terms of reason or natural causality. It’s worth emphasizing in conclusion, however, that this account doesn’t rule out justificatory relations between beliefs, including between judgments that perception has motivated. Clearly, I can adduce the judgment “I see an owl” as a reason for “That’s an owl,” for example. One may even wish to say that only at this higher level do we acquire knowledge. I would prefer not to say this—it seems to me the normative constraint provided by motivation suffices to count it as a genuinely epistemic ground. But, again, this relation between judgments occurs at a higher level than the relation I am trying to analyze: namely, the relation between judgments and perception itself. It will mark a significant step forward in our understanding of this relation, I have argued, if we account for it in terms of motivation.