RESPONSIBLE BELIEVING

S. Joel Garver

CHAPTER FOUR

Seeing Isn't Believing:
(And Other Thoughts on Doxastic Practices)

Section II: Part B

Eskimos famously have more ways of talking about what English-speakers refer to simply as "snow" and, it is thought, a far more specialized and fine-grained set of concepts regarding frozen precipitation. Similarly, the Hopi have one word and, it seems, one concept to cover what English-speakers refer to by means of the words "airplane," "pilot," and "dragonfly." Furthermore, in anthropological field work, ethnologists, linguists, and the like often have great difficulty in "mapping" the linguistically expressed concepts of other cultures back into their original language for the purpose of expressing these concepts to readers of reports, ethnologies, and so on. There is no simple fit between the two or more conceptual schemes or frameworks. Moreover, the ontological commitments of these various frameworks commit their participants to different objects, parts, kinds, relations, and so on. (For anthropological materials on these points see, e.g., Whorf 1956; Kearney 1972:43ff; also cf. Barfield 198 8 and Durkheim 1915.)

The upshot of these points is that two individuals from two different cultures, each with a distinct set of concepts, might well be having indistinguishable experiences and acquire very different sets of beliefs. These kinds of situations have various implications. I shall not draw out all possible implications nor shall I argue how we are to make sense of all these implications. Much of that work is beyond the scope of my present interests and beyond what is to be seen as more strictly epistemological. Nevertheless, allow me to suggest a general picture as a background to my more substantive comments.

It seems to me that however we are interpret these types of situations, we will do well to allow ample space for a "given"--that is, aspects of reality that are not reducible to or merely forms of human beliefs, concepts, practices, social struc tures, and the like. Now there are two kinds of "giveness." There is a metaphysical given, that is, the world of all things in heaven and earth, seen and unseen apart from what anyone might believe or say about it. There is also a phenomenological given, that is, the range of human (sensory and event memory) experiences including types and tokens of experiences, some of which are phenomenally indistinguishable, some of which are not, and all which may be variously grouped or classified. More will have be said on both these types of "giveness."

On the other hand, we must also account for the variableness of human descriptions and beliefs concerning the metaphysical given that are, in various ways, based on the phenomenological given. As a first approximation, let us consider cases in which [1] we have two normally functioning subjects (call them "Francis" and "Dominic") who are having experiences that possess indistinguishable phenomenology or identical phenomenal type, [2] the same metaphysical given presents itself in these experiences so that they count as perception, and [3] Francis and Dominic come to hold different beliefs concerning what it is they are experiencing.

It is clear to me that not every case of this sort involves some sort of error on the part of at least one of the subjects. In fact, it is clear to me that some of these cases are cases of normal, properly functioning perception in which Francis and Dominic come to hold very different, prima facie incompatible sets of beliefs that describe the way the world is, and, furthermore, the sets of beliefs that they hold are both true and count as knowledge.

For instance, suppose that Francis is a member of the Hopi nation while Dominic is an Englishman. We present them with flashcards which picture the following: an ocean, a stream, a bottle of water, a fountain, a cup of water, a ladle-full of water, a waterfall, a bucket of water, and a pond. We ask both of them (in their native tongues) to explain, in the most general terms they can, what they saw.

Dominic, the Englishman, says "Of course. You showed me pictures of water."

How many kinds of water, we ask.

"Just one kind, really--ordinary liquid water. It's not as if you showed me ice or something. Just water in different containers and environments."

We ask Francis, the Hopi, what he saw.

He answers (translating what we can), "Of course. You showed me pahe and keyi."

Was there anything more in common than that, we ask.

"Well...those are two kinds of wet things. But you didn't include snow or milk or any other wet things."

It seems then that the Hopis lack a concept of water as a natural kind. They only have the concepts referred to by the words "pahe" and "keyi" (very roughly speaking pahe is naturally occurring water or flowing water--pond, waterf all, fountain, ocean--while keyi is water that is isolated or contained for drinking, washing, and other social purposes). As a result the Hopi and the Englishman possess different beliefs regarding what they saw and the number of kinds of things they saw (see Whorf 1956). Both came to their beliefs normally, with properly functioning perception. Both sets of beliefs appear to be true and to count as knowledge (given the contexts)--but their beliefs sets are prima facie incompatible.

It is considerations of the kinds I have just outlined that, I believe, lead many epistemologists to hold that there must be something more to the justification of perceptual beliefs than the fact that these beliefs are reliably based on experience. Is it not obvious that conceptual schemes or frameworks are playing some role in the justification of these beliefs? I am in agreement with that much, but the question is whether such frameworks are best theorized as playing a role within the basis of perceptual beliefs and as playing that role qua beliefs or belief systems. Or is there another way to theorize about such situations?

Towards answering this question I will draw on some ideas and arguments put forward by Alan Millar (Reasons and Experience 1991). Millar suggests that many philosophers believe that sensory experiences have propositional structures. He writes,

Millar also believes such a view to be mistaken. He argues against it by trying to show that it is not the case "that the content of an experience is intrinsic to the type of experience, in the sense that any experience of the same type would have to have the same content" (1991:1). On the contrary, Millar argues that

By an "F-type experience" Millar means, roughly, an experience as if of an F or, better, an experience of the type that an F would yield under certain conditions (these obviously need refinement, but as a rough and ready definitions they will suffice for the time being).

The type of situation that Millar has described is precisely the kind of consideration that I have suggested underlies the philosophical intuition against immediate justification. In fact, Millar argues against both immediate justification and foundationalism. While I think his arguments ultimately fail, there is much that is instructive in his presentations. Particularly helpful are his arguments as to the nature of experience, belief-formation from experience, and forms of inferential competence and what he calls "quasi-inferential" competence. Let us examine these arguments in more detail.

It is to be noted that in Alston's discussion of the doxastic practice of sense perception it is precisely perception which concerns him and not so much phenomenal experience in general. That is to say, in Alston's account the focus is on cases of an object X appearing to or presenting itself to a certain subject S and presenting itself as something, e.g., as bumpy, as loud, as a Boston terrier, and so on. How one conceives of the nature of perception is not of great import, though Alston prefers what he calls the Theory of Appearing in which perception is unanalyzable beyond its being cases of Xs appearing to an S as so-and-so.

It is, however, in the language of "putative" perceptions that Alston's theory is couched, though for such putative perceptions to render certain classes of belief as likely to be true, such putative perceptions must usually be genuine. So much for Alston's views (drawn from 1991:54ff., 68ff.). Millar, in some contrast, focuses primarily on experience-types which may include mostly experiences that count as perceptions of an X. These experiences-types, however, are also neutral to the presence of an X in such a way as to include experiences, phenomenally indistinguishable from perceptions of an X, that are merely as if one were perceiving an X (1991:12-19). In the end, however, Alston and Millar's approaches amount to much the same thing since for both writers putative perceptions serve as reliable inputs only if they are regularly genuine.

In light of the preceding paragraph, the phrase "experience of an F" is not to be interpreted in terms of genuine perception, but rather in terms of experiences as if they were perceptions of an F. Let us move then from mere experience of an F to a case of having an experience such that it seems to you that an F is there. On Millar's account

Among countervailing considerations Millar includes "anything which casts doubt on the reliability of the relevant sense-modality, or which suggests that the intervening medium is not suitable for accurate perception, or which, independently of these considerations, casts doubt on whether an F is present" (1991:20). In general, these are parallel to the sorts of considerations that Alston includes among overriders of justification, though Millar is not so interested in these as abstract conditions that may or may not obtain, but as considerations that in some way may or may not be present to the subject.

The question that arises concerns the move from an experience of an F to having an experience such that it seems to you that an F is there. To put it another way, we are concerned with the move from experience as kind of phenomenal awareness to experience as having propositional content that gives rise to (or is identical with) a propositional attitude ("It seems that an F is there"). A certain move of this type (e.g., a move from your experience of an F to it seeming to you that an F is there) might be universally legitimate and entailment-like. If so, then experiences may have propositional content intrinsically. And Millar suggests that if experiences have propositional content intrinsically, then certai n principles must be true of experience--namely, [1] the Concept Principle and [2] the Intrinsicality Principle.

Allow me to briefly summarize his line of argument before we examine the details. Millar's Concept Principle states simply that "if a concept is an ingredient in the concept of a mental state then the subject of the state must grasp the concept " (1991:20). The Intrinsicality Principle states that "if token mental states in a given category (belief, desire, or whatever) have different [propositional] contents then they are of different state-types within that category" (1991:21). Millar 's basic strategy at this point is to argue that experiences as types of phenomenal awareness do not entail any particular propositional contents. That is to say that "token experiences may differ in content yet not be different qua experiences (phenomenally, qualitatively)" (1991:25). And so the Intrinsicality Principle is false in respect to experiences. Furthermore, he will argue that there is "nothing in the notion of an F-type experience which requires subjects of such experiences to possess the concept of an F" (1991:32). And so the Concept Principle is false in respect to experiences.

The upshot of these points is that there is a "distinction between an F-type experience and an experience such that it seems an F is there" and this distinction has two important consequences (1991:32). First, it disallows us from seeing the move from an experience of an F (an F-type experience) to an experience such that it seems to you an F is there as a move of logical or causal necessity. Rather, it is a move that involves the conceptual apparatus of the believing subjects. That is to say, since experiences do not have intrinsic propositional content, the concepts used in belief-formation are not logically necessitated by the type of experience (and so the fact that the Hopi and the Eskimos are different than us in respect to the content of their beliefs need not prove a logical puzzle).

Second, Millar uses his arguments to allow us to study the move from experiences to belief. He theorizes it as an exercise of a type of acquired habit, intellectual practice, or conceptual competency. So much for the outline of Millar's argument. Let us now turn to the details of the arguments themselves.

Let us return to the Intrisicality Principle. Recall that it demands by implication that "experiences which differ in [propositional] content are different experiences, in the sense of being of different experience types" (1991:22). The question then must be raised as to what it is for experiences to be of the same or different types. Millar begins answering this question by appealing to what he calls "the principle of phenomenal character" which states "experiences are of the same (different) types if and only if they have the same (different) phenomenal character(s)" (1991:23). This is not particularly helpful, but consider the following example by way of elucidation:

If propositional content is intrinsic to experience, then one might think the experiences of the two sisters are two different types of experience--the content of one experience would be specified as concerning Sally and the content of the other experience as concerning Sally's twin sister. This difference in propositional content would obtain despite the phenomenal indiscernability of the experiences and the consequent violation of the principle of phenomenal character.

It is conceivable that with appropriate background beliefs and the like, you might be led to believe on the basis of one experience that it is Sally before you and on the basis of the other experience that it is Sally's twin. Nevertheless, as Millar notes, it is "odd to suppose that such a difference amounts to a difference in the intrinsic features of the experience, a difference in the experiences qua experiences" (1991:24; remember also we are discussing experiences and not perception per se, since the notion of a perception of Sally or her sister does intrinsically involve the object of that perception). The natural response to this example is to "reject the view that experiences must satisfy the Intrinsicality Principle and retain the principle of phenomenal character" (1991:25). This does justice to what I earlier suggested to be the "phenomenal given."

Millar defends this line of reasoning by constructing a typology of experience. His basic thrust is this:

Various examples of experience types in various modalities might include (red thing)-type, (apple on a tree)-type, (rough and hairy thing)-type, or (clap of thunder)-type.

Millar adds a number of clarifications not all of which are of interest to us. Some, however, are of interest. First, not all F-type experiences are (necessarily) F-indicating (thus allowing artificial apples to produce apple-type experiences). Second, "numerically distinct token experiences which are both F-type may differ qua experiences (phenomenally, qualitatively) in many ways" (1991:28). For example, not all F-type experiences need also be (F against background B)-type experiences. Third, in the case of "a (grouse in grass)-type experience the intended range would exclude experiences obtained when looking straight at a grouse which is completely camouflaged by its colouring and background" (1991:30).

Now we are in a position to consider the point that there is "nothing in the notion of an F-type experience which requires subjects of such experiences to possess the concept of an F" (1991:32). And so we undermine the Concept Principle's application to experience and we disprove the idea that propositional contents are intrinsic to experience. Millar gives the following example:

Thus when an experience is such that it seems that an F is there, we have an interaction between experience and conceptual capacities. Millar writes "the idea is that for a person who has a mastery of the concept of an F, F-type experiences are apt to be experiences such that it seems to the subject that an F is there, where that is explained by the capacities constitutive of mastery" (1991:32).

Having considered the relation between experiences and perceptual beliefs, let us return to the issue of immediate justification. On Millar's account it is apparent that perceptual beliefs can be justified, at least in part, on the basis of experience. It is also apparent, however, that since experiences do not have intrinsic contents in the way, for instance, propositions do, beliefs cannot be based on experiences in a strictly inference-like way (as in some cases of beliefs based on beliefs). Furthermore, on Millar's view the content of beliefs based on experience also necessarily involve some interaction between the experience and a conceptual capacity. There is some question then as to what is entailed by the notion of a conceptual capacity, how it functions in belief-formation, whether its functioning necessarily involves other beliefs not drawn from experience (e.g., presuppositions), and whether those beliefs (if any) form part of the basis of perceptual beliefs. Only if such beliefs are necessarily involved in the basis of perceptual belief-formation will immediate justification be impossible. It is important to note that at this point the discussion has come full circle and we are reexamining some of what Alston discussed in connection with adequacy assumptions, contextual beliefs, and the like, though we are going somewhat further.

Millar believes that the involvement of conceptual capacities (as he conceives them) in perceptual belief-formation renders immediate justification impossible. Let us examine his case. I will not go into all the details of Millar's views on justification, but only as they bear on the immediate justification of perceptual beliefs. An important notion throughout Millar's discussion is that of "quasi-inference." Quasi-inference involves "correct or incorrect applications of concepts in response to experience...there being inference-like links from experience types to belief contents" (1991:116). Millar calls these links "quasi-inference" since they are both like and unlike ordinary inference patterns. The links are unlike ordinary inference since they do not conceptually link two entities that have propositional contents. The links are like ordinary inference in that they are concept-governed and concept-individuated. That is to say, the links essentially involve use of conceptual capacities and different concepts regulate different kinds of links. (See also Alston's discussion of "objectification" in 1983:107-109 and McLeod on "conceptual-reading practices" in 1993:35-36, 49-50.)

Let us consider a quasi-inference pattern that shall be called "Q". Roughly, Q says that if [1] S is having a F-type visual experience and [2] There are no countervailing facts with respect to there being a thing having the look of an F before S, then S is right in quasi-inferring that [3] A thing having the look of an F is before oneself. Millar explains Q in the following way:

The experience type and the proposition regarding countervailing facts together form what Millar calls an "unconditional yet inconclusive reason" for the conclusion.

Unconditional reasons are basically forms of inference involving a proposition or experience, the fact that the proposition or experience is strong fallible evidence for a further proposition, and there being no countervailing facts in respect to the further proposition (1991:64-75). In more Alstonian terms, one might think of unconditional reasons in terms of reliability or adequacy of grounds in the absence of overriders, or in terms of truth- conducivity, or in terms of being in a strong position in respect to ascertaining a certain true proposition. The details need not concern us here.

Millar argues that it is not enough for the justification of a perceptual belief simply that a subject S is having an experience of a certain type P, comes to believe a particular proposition that p, and this pattern mirrors an acceptable pattern of quasi-inference Q. Further conditions must be met. First, S's belief that p must be based on the experience of type P. Second, S must believe in line with the inference pattern Q that there are no countervailing facts to it being the case that p. Third, S's believing that p on the basis of the experience of type P and the belief concerning no countervailing facts must be "the outcome of the exercise of competence on his part in employing the concept" of a P. Fourth, S's belief that there are no countervailing conditions must be one he is in a position to take for granted (these points are drawn from 1991:117-8; emphasis mine).

There are obvious parallels here with Alston's notion of a justified belief being based on an adequate ground and there being no overriders. What is different is that Millar requires (to put it in Alston's way of speaking) that the subject believe there to be no overriders and to be in a position to take that for granted. Therein lies part of the difference between Alston and Millar and why Alston can allow immediate justification and Millar cannot. On Millar's view a justified belief that p always includes another belief as part of its basis, to wit, the belief that the belief that p has no countervailing facts against it. This alone is enough for Millar to disallow immediate justification. But this is not the only reason why Millar disallows immediate justification.

Note that in our discussion of quasi-inference, the pattern Q allowed one to move from an F-type experience (and the grantable belief that there are no countervailing facts against the conclusion) to the conclusion that there is a thing having the look of an F before one. This conclusion quite intentionally falls short of the conclusion that there really is an F before one. Consider an example. According to Millar an (apple on a tree)-type experience can, under certain conditions, justifiably lead one to believe that a thing having the look of an apple on a tree is before one. This, however, is not the same as believing that an actual apple on a tree is before one.

According to Millar, what is further required to reach the full-blooded conclusion that there really is an apple there, is the exercise and application of one's mastery of the concept of an apple. He argues that we have "a mode of applying the concept [of an apple] which presupposes that a thing's having the look of an apple is strong fallible evidence that it is an apple" (1991:123). This is the kind of belief that Alston earlier called an adequacy assumption. The further inference authorized by the adequacy assumption is contingent on our presuppositions concerning the way the world is. As Millar writes, "in a world with as many fool's apples as apples the evidential link would not obtain" (1991:123).

At this point we need to employ a distinction that Millar draws between mastery of a concept in a broad sense and in a narrow sense. It is best to quote him on this. To master a concept in the narrow sense "is simply to possess it, which, to a first approximation, means to be governed by the inference patterns by which it is individuated" (1991:87). To master a concept in the broad sense entails mastering it in the narrow sense "plus the mastery of any modes of application which though not entailed by possession nevertheless prevail among those for whom the concept has use" (1991:87). This distinction allows us to make sense of those situations in which, for instance, you can talk with some knowledge concerning seismographs (they measure movement of the earth's crust, their measurements are correlated to the Richter scale, etc.), but you wouldn't recognize one if it bit you (so to speak). How this distinction plays into the formation of perceptual beliefs will be clear shortly.

In the case of what Millar calls a strongly experiential concept only a narrow mastery of the concept is necessary for producing justified perceptual beliefs. For example, in the case of the strongly experiential concept of something being red the notion that red things have a distinctive visual appearance is built right into the concept. Thus one can justifiably move by quasi-inference from a (red thing)-type experience to the belief that something red is before you (that is, without an intervening belief to the effect that the thing before you has the look of a red thing). And one can make this move with only narrow mastery of the concept of a red thing. Furthermore, it is not merely contingent, Millar suggests, that having a (red thing)-type experience is strong fallible evidence of the presence of a red thing. The same is true of any strongly experiential concept.

On the other hand, in the case of what Millar calls a weakly experiential concept a broad mastery of the concept is necessary for producing justified perceptual beliefs. For example, in the case of the weakly experiential concept of something being an apple we need a broad mastery of the concept of an apple--incorporating our actual modes of applying the concept. Only then can one justifiably move to the belief that an apple is before you. Furthermore, this move has two parts: first there is a quasi-inference from an apple-type experience to the belief that a thing having the look of an apple is before one; and then the belief that the thing before one is an apple can be justifiably believed on the basis of the quasi-inferred belief.

Here, then, is another reason why many perceptual beliefs that one might take to be immediately justified cannot be so on Millar's account. According to him, justified beliefs of the form "That is an X" or "An X is there" where X is a weakly experiential concept are necessarily based on other beliefs of the kind Alston called adequacy assumptions.

I am in basic agreement with Millar on much. I agree that beliefs that are justifiably held on the basis of experience must be formed in accord with forms of inference and quasi-inference that are expressive of the concepts utilized. I agree that for such belief-formation to be justified it must actually be regulated by or involve an exercise of these (quasi-)inferential moves. I also agree that experience itself forms the basis of the belief. My two basic criticisms of Millar question whether beli efs play the sorts of roles he gives them in the basis of all perceptual beliefs. First, I believe Millar is mistaken to see beliefs of the form "There are no countervailing facts in respect to there being a thing having the look of an F before me" as having an essential role in patterns of quasi-inference. Second, I believe he is mistaken in holding that patterns of belief-formation involving weakly experiential concepts (and thus broad mastery) necessarily have two parts.

Regarding the first criticism, I take my cue from Alston's treatment of overriders. It need not be the case that a subject actually believe there to be no overriders (or, in Millar's terminology, countervailing facts); it is enough that there are, in fact, no overriders (countervailing facts) present. I fail to see why anything more is required. Certainly the exercise of a form of quasi-inference presupposes the absence of countervailing facts, but presupposition need not amount to belief. One more note on this matter: I would hold that if there are countervailing facts present, they ought to be sorts of facts that are accessible--though, as I have noted, I doubt that they must be accessible to the subject rather than simply the community that shar es that form of quasi-inference. If I am correct in this first criticism, then forms of quasi-inference do not necessarily involve beliefs as part of the basis for the beliefs they produce. Thus experience can be held to lead to immediately justified perceptual beliefs. More will be said on the nature of quasi-inference as we continue.

Regarding the second criticism, it is to be noted that Millar's view does not, on its own, threaten all sorts of immediate justification. Rather, it only threatens the immediate justification of beliefs that involve weakly experiential concepts. Such beliefs, however, include beliefs such as "That is an apple," "A pencil is on the table," "This rhubarb-pie is sweet," and other relatively simple perceptual beliefs. I find it incredible to think that such beliefs cannot be immediately justified given the sorts of quasi-inference Millar has already allowed. Certainly, by means of quasi-inference, an apple-type experience can justifiably yield the belief that something having the look of an apple is before one. But it also seems to me that there is no reason to think that the belief that something having the look of an apple is before me is necessarily involved in the move from an apple-type experience to the belief that there is an apple before me. Introspection at the very least affords no hint of such a two-step process of belief-formation. Furthermore, as Alston noted (in criticism of the idea that perceptual beliefs are justified by experiential beliefs), very seldom do we actually even form beliefs about our experiences (1991:82-83).

Rather, it is the case, I think, that broad mastery of the weakly experiential concept of an apple--though mutable and very much contingent on the scarcity of fool's apples--does in fact enable us to seamlessly move from an apple-type experience to the belief that there is an apple before me. The pattern of quasi-inference involved in this move is at the same time the pattern that allows us to move to the belief that something having the look of an apple is before me (if and when such a belief is form ed). While something's being an apple and something's having the look of an apple are not conceptually built into one another (as are something's being red and looking red), the broad mastery of the concept of an apple that is involved in the quasi-inference accounts for both beliefs. And so immediate justification of the belief that an apple is before one is possible.

In this section, then, we have, I hope, vindicated the possibility of immediate justification for perceptual beliefs. We have also begun to construct an account of what happens between the "meager inputs" and "torrential outputs" in belief-formation. What happens is that inferential and quasi-inferential functions are exercised and these functions are basically kinds of conceptual competence. More needs to be and will be said concerning these functions. For now let us simply keep in mind a couple of points.

First, as forms of conceptual competence these inferential patterns need not be engaged in consciously, intentionally, deliberately, or otherwise. Though deliberate engagement is perhaps possible in some cases the capacities mostly function on "cruise control," though that does not necessarily exclude extensive and pervasive forms of indirect influence and therefore responsibility.

Second, we have provided a first step towards accounting for social, cultural, linguistic, religious, and other sorts of variations between discursive formations or forms of life or conceptual frameworks. If Eskimos or Hopi or some other community utilize concepts that differ significantly from our concepts (e.g., cutting the world at different joints or being more or less specialized or fine-grained), then their forms of quasi-inference will also differ alongside. A number of differences can be accounted for in this way. Furthermore, it is easy to see how those who use different concepts in respect to the same metaphysical and/or phenomenal given, can be justified in their beliefs. In fact, we can see how in such a situation they can both come to hold true beliefs or have knowledge, though in respect to very different propositions.

This last point might entail some sort of pluralism or perspectivalism in terms of metaphysical descriptions of the world (i.e., there is no one, uniquely true story of the world), but that should not prove problematic if we keep in mind the following factors: [1] such a view does not entail a plural notion of what it is for a proposition to be true; [2] while putatively true propositions from various frameworks may apparently conflict (e.g., concerning the number of kinds of objects present), such propositions must be indexed to the conceptual frameworks in which they emerge (and there may be extensive overlap between these frameworks); [3] descriptive pluralism does not entail that there is no world in itself or things in themselves, since that world may be such that it can successfully sustain the truth of many different descriptions; and [4] that does not mean that anything goes in terms of metaphysical descriptions since the world cannot successfully sustain just any propositions nor can it sustain intra-framework contradictions and the like. (For more arguments in favor of the coherence of pluralism in respect to metaphysical descriptions, see Michael Lynch's dissertation, 1995.)

Third, even if there are various conceptual frameworks that affect the formation of perceptual beliefs, that does not mean that sense perceptual or any other doxastic practices are identical with conceptual frameworks. There are various reasons not to make the identification: sense perceptual practices may span across several frameworks, one framework may employ or be involved in many different practices (memory, perceptual, rational intuition, etc.), and practices are largely individuated by input, output, and overrider types, not conceptual types per se (though the functions between inputs and outputs are concept generated).

Let us now move on, deepening our account concerning quasi-inference and conceptual competence, and examining how such considerations come into play in various ways in various doxastic practices.