INTRODUCTION
It is common in ordinary and philosophical discourse to talk about how things seem or appear. Recent epistemologists have suggested that these ‘seemings’ ought to play a central role in epistemology.Footnote 1 Michael Huemer, specifically, has been instrumental in defending Phenomenal Conservatism or PC: if it seems to S that P, S thereby has some degree of defeasible justification for believing P (Reference Huemer2007: 30). But what are these seemings? Answering this question is crucial for determining what (if any) epistemological role such a state can play.
Many have adopted a kind of non-reductionism that construes seemings as intentional states that cannot be analyzed in terms of more familiar mental states such as beliefs or sensations.Footnote 2 In this paper I aim to show that reductive accounts need to be taken more seriously by illustrating the plausibility of identifying seemings and conscious inclinations to form a belief. I close the paper by briefly considering the implications of such an analysis.
We must first note that “seems”, “appears,” and their cognates are used equivocally. Chisholm, for instance, famously distinguished epistemic, comparative, and non-comparative uses of ‘appears’ language (Reference Chisholm1957: Ch. 4). On the epistemic use, these words are used to assert that we have justifying evidence for a particular belief. In their comparative use, these words describe an experiential state as being similar to the experience we'd have under certain conditions. Finally, the non-comparative use picks out the intrinsic and sensory qualities of an experience.
I doubt Chisholm's list is exhaustive. Sometimes “seems” or “appears” words are also used to indicate merely that we have a belief without the implication that we have justifying evidence. Other times we place a special emphasis on “seems” to suggest that the evidence that exists for P is misleading. Still other times we use “seems” to indicate that the available evidence favors P but, for practical purposes, we cannot appropriately ignore the possibility that ~P.Footnote 3
I won't attempt to produce an exhaustive list of the uses of “seems” but it's crucial to remember these varied uses. If seemings are to do explanatory work in epistemology then we need to be careful to pick out – and keep our attention on – the relevant phenomena. If we use “seems” in Chisholm's epistemic sense then it's trivial that S has justification for believing P when it seems to her that P. However, this use of “seems” presupposes such justification and thereby fails to offer any explanation. Focusing on one particular use of “seems” rather than another might also tend to lead one to particular views regarding the epistemic role that seemings or appearances can play. Sellars (Reference Sellars, Feigl and Scriven1956) focused on the comparative use of “appears” in a way that motivated his version of coherentism. Alternatively, Chisholm's defense of a non-comparative use of “appears” that picks out the intrinsic and sensory qualities of experience was crucial to his reply to Sellars and his defense of foundationalism. Lastly, appreciating the varied uses of “seems” is crucial since advocates of non-reductionism present worries for reductive accounts that can be accommodated by appealing to these varied uses of “seems” rather than positing a sui generic propositional attitude.
This leads to our focus: a particular kind of mental state that “seems” and “appears” statements are sometimes used to pick out. Perceptual, memorial, introspective, moral, and intellectual uses of “seems” nicely gesture at this state. It perceptually seems to me that there is a red table in front of me. It memorially seems to me that I ate eggs for breakfast. It introspectively seems to me that my pain is more of a stinging than a throbbing pain. It morally seems to me that it's wrong to take pleasure in another's pain. And it intellectually seems to me that 2+2=4. In each of these statements, “seems” is consistently used to pick out a peculiar mental state with which we're all intimately familiar. Two aspects of such seemings are especially noteworthy.
First, these seemings are representational states in the minimal sense of having accuracy conditions. We're interested in seemings that something is the case. My seeming that there is a red table in front of me is accurate iff there is a red table in front of me.Footnote 4 Second, a crucial component of a seeming is its phenomenology. There is something-it's-like for a proposition to seem true. A key phenomenological quality of seemings identified by Tolhurst (Reference Tolhurst1998) is their “felt-veridicality”: “seemings have the feel of truth, the feel of a state whose content reveals how things really are” (pp. 298–9). This notion is a bit obscure but becomes less so when we compare seemings to imaginings or desires. Imagining or desiring that the Tigers win the World Series represents this state-of-affairs but it doesn't strike one as actually being the case. It's in this sense that a seeming feels as if something about the world is being revealed to its subject.
1. SENSATIONS, BELIEFS, AND CONSCIOUS INCLINATIONS
Having honed in on our target of analysis, can these mental states be identified with any more familiar mental states or must we construe them as a sui generic and irreducible propositional attitude?
First, let's distinguish seemings and sensations. There are perceptual, memorial, introspective, moral, and intellectual seemings but it's controversial whether there are memorial, moral, or intellectual sensations. Any reason to doubt the existence of said sensations is thereby a reason to distinguish seemings and sensations.
Even if you accept the existence of the controversial sensations, we can distinguish seemings and sensations by considering perceptual constancies. The contents of sensations are more fine-grained than those of seemings. When I look at a red table in normal lighting conditions I have a particular sensation and it thereby seems to me that the table is red. As I move around the table or as the lighting dims, my sensation is constantly changing in incredibly subtle ways but the perceptual seeming remains constant. Thus, the sensation and the seeming must be distinct states.
A distinction between seemings and sensations is also supported by the possibility of two subjects sharing the same perceptual sensations while things perceptually seem different. Consider a novice and expert birdwatcher. Each might have the same visual and auditory sensations when looking at a Goldfinch while only the expert has the perceptual seeming that the bird is a Goldfinch.
However, even if sensations aren't seemings they often give rise to seemings. The expert birdwatcher has a variety of visual and auditory sensations that thereby give rise to her perceptual seeming. A natural thought then is to identify seemings and beliefs. Beliefs are representational states distinct from sensations that are nevertheless often caused by sensations. Moreover, unlike desires and imaginings, the contents of a belief strike us as actually being the case.
Unfortunately, known illusions pose a difficulty for identifying seemings and beliefs. When looking at a straight stick in water it seems that the stick is bent. It seems (to many) that the Müller–Lyer lines are different lengths. When looking at a white table illuminated by red lights it seems that the table is red. Such seemings persist even after we've become privy to the illusion and we no longer hold the relevant beliefs. Seemings are not beliefs.Footnote 5
What about an inclination to form a belief? We often have an inclination to form a belief even though our awareness of countervailing evidence prevents us from actually forming the relevant belief. Known illusions therefore pose no threat to a reductive account of seemings in terms of inclinations to believe.
One difficulty, however, is that it's likely that I'm unconsciously inclined to believe an infinite number of propositions that I have yet to so much as consider. Such propositions don't yet appear true, they don't have “the feel of truth.” We wouldn't describe these inclinations as seemings until they rose to the level of consciousness. This leads to the view I want to defend:
The Conscious Inclinations View (CIV)
P's seeming true to S is S's having a conscious inclination to believe P.Footnote 6
Before motivating CIV and defending it from objections, I should clarify the crucial notion of a conscious inclination to form a belief.
There are two aspects to a conscious inclination. First is the actual disposition. A sugar cube has the disposition to dissolve when placed in water. This is true because there is a categorical state of the sugar cube – weak molecular bonds – that causes the cube to dissolve when placed in water provided no stronger countervailing causal factors are present. Similarly, it's true that S has an inclination to believe P when there is a categorical state X that has the power to produce in S a belief that P provided no stronger countervailing causal factors are present.
A second and equally important aspect of a conscious inclination to form a belief is a phenomenology that could occur absent the underlying dispositional state. Consider an analogy with courageousness. A subject is courageous when she is disposed to behave in certain ways under specific conditions. However, a subject could feel as if she is courageous even though she isn't; she could feel courageous even though she is actually disposed to run away and hide when a loved one's life is threatened. You could also be in the phenomenological state of feeling inclined to believe a proposition even though you don't actually have the relevant inclination.
Various philosophers understand the structure of phenomenal properties differently and these would lead to different understandings of the phenomenological component of a conscious inclination. One possibility is that phenomenological properties are monadic properties of a subject's consciousness (a kind of adverbial theory). The monadic property associated with being inclined to form a belief could be present despite the actual inclination being absent; such a feeling would be illusory. If you adopt this account of phenomenal properties then a conscious inclination would be identified with the state of having a veridical feeling as of being inclined to form a belief.
An alternative account of phenomenological states as relational would lead to a subtly different account of conscious inclinations. Some hold that phenomenal properties are best understood as states where a subject stands in a real relation of acquaintance with an object.Footnote 7 On such a view, S's veridical feeling as of being inclined to believe P consists of S being acquainted with her inclination to form said belief. What about the illusory case? Here the advocate of the relational view can take a note from direct realist theories of perception and claim that, in the bad case, one is merely acquainted with a state subjectively indistinguishable from the inclination to believe P.
In this paper I remain neutral on this issue. What I want to stress is that CIV identifies seemings with a complex mental state. A conscious inclination to believe P has both the dispositional state and its associated phenomenology as components. On the monadic view the conscious inclination is constituted by: (a) the dispositional state and (b) the monadic property of a subject's consciousness associated with the presence of such an inclination. Alternatively, on the relational approach the conscious inclination is constituted by one's standing in a relation of acquaintance with the relevant inclination – note that this state includes the inclination itself as a constituent.Footnote 8
It's also important to note that S's having a conscious inclination to believe P shouldn't be understood in terms of S's representing (or being able to report) that she is inclined to believe P. A young child with the concepts <table> and <red> can have a conscious inclination to believe that there is a red table in front of her – and thereby have this seeming – even if the child lacks the more complicated concepts <belief> and <inclination>. Borrowing a distinction from Ned Block (Reference Block1995), in CIV I identify seemings with phenomenally conscious rather than access conscious inclinations.
Finally, it's important to distinguish CIV from a similar view. Tolhurst (Reference Tolhurst1998) suggests that a seeming is a mental state M with propositional content, felt-veridicality, and that is responsible for S's feeling an inclination to form a belief endorsing the propositional content of M.Footnote 9 Tolhurst identifies seemings not with the conscious inclination but, rather, with the mental state that causes this inclination. Cullison (Reference Cullison2010) correctly points out that Tolhurst's view is therefore consistent with (but doesn't require) construing the seeming (i.e. the ground of the conscious inclination) with an irreducible propositional attitude.Footnote 10 Tolhurst's discussion of seemings is one of the most insightful and illuminating, but it's neutral on whether seemings can be identified with more familiar mental states. It's actually inconsistent with the particular reductive account I've proposed with CIV.
2. SPECIES OF SEEMINGS
Seemings are a genus that includes many species. Noting differences between kinds of seemings will put us in a better position to defend CIV from the supposed counterexamples discussed in §4.
First, seemings differ in strength. A seeming that P is sometimes strong and other times very weak, subtle, and difficult to notice. There is nothing mysterious about this on CIV since conscious inclinations come in degrees. Notice here that there is an interesting interaction between the two aspects of conscious inclinations. The disposition can be stronger or weaker and these differences give rise to phenomenological differences in one's conscious experience of that inclination.
Second, seemings differ in what I will call their mode. We can have perceptual, introspective, memorial, moral, and intellectual seemings. Both the etiology of the inclination and the content of the belief toward which one is inclined serve to individuate these modes. Perceptual seemings are conscious inclinations towards beliefs about the external world directly grounded in one's sensations (e.g. visual or tactile). By “directly grounded” I mean that the sensation produces the conscious inclination in an inferentially opaque manner (see Bonjour Reference BonJour1985 and Lyons Reference Lyons2008); S's conscious inclination isn't the result of any introspectively accessible process of inference.Footnote 11 Introspective seemings are conscious inclinations towards beliefs about your own mental life that are directly grounded by your mental states. Memorial and intellectual seemings can also be individuated in terms of the content of the belief towards which one is inclined along with this inclination's etiology.
In characterizing those different modes of seemings we've been focusing on a single aspect of conscious inclinations: the inclination itself. I take this to be metaphysically fundamental for individuating modes of seemings. However, it's important to note the phenomenological differences as well. Different features of the inclinations of which one is conscious give rise to differences in the phenomenology involved in being conscious of the inclination. Consider differences in the content of the belief toward which one is inclined. It feels different to be conscious of a perceptual inclination to believe that a bird is in front of you than it feels to be conscious of a perceptual inclination to believe that a Goldfinch is in front of you. Now consider the etiological differences. Consciousness of an inclination to believe your car is in the west parking lot feels different when it's produced by perceptual mechanisms than when it's produced by memory mechanisms. These phenomenological differences associated with our conscious experience of different inclinations helps explain our ability to introspectively categorize seemings by their modes.Footnote 12
Despite these phenomenological differences we should be careful to properly appreciate the phenomenological similarity of the various modes of seeming truth. Each mode involves a similar feeling as of being compelled to form a belief. This is part of what unites the genus.
Lastly, seemings can be differentiated in terms of their experienced directness. Consider the following scenario:
Jill, an expert birdwatcher, has taken Jack bird watching for the first time. They each look through a pair of binoculars at the same Goldfinch. Jill and Jack have the same kind of visual sensation. Given the differences in their training, it seems to Jill that there is a Goldfinch even though Jack lacks such a seeming. However, Jill then provides Jack with testimony that the bird is a Goldfinch.
If Jack trusts Jill then it's natural to suppose that there is some sense in which it will begin to seem to him that there is a Goldfinch; it feels as if the truth that there is a Goldfinch is being revealed to him via Jill's testimony. Jack's seeming is importantly different than Jill's, however. Jill's background and non-occurrent beliefs about how Goldfinches look bring about a seeming that feels (from her own perspective) as if it arises directly from her visual sensation. I'll therefore refer to the kind of seeming Jill has as a direct seeming. Jack's seeming that there is a Goldfinch, on the other hand, feels different. His seeming is existentially dependent on his seeming that Jill provided the relevant testimony, and (from his perspective) it's experienced as mediated. I'll therefore refer to this latter kind of seeming as an indirect seeming.
Incorporating the direct vs. indirect distinction is straightforward on CIV. Both Jack and Jill have a conscious inclination to believe the same proposition. However, due to her sub-personal beliefs about how Goldfinches look, Jill's conscious inclination arises from her visual experience absent any introspectively accessible process of inference. Alternatively, Jack's conscious inclination to believe that the bird is a Goldfinch is explained by the presence of a distinct inclination in a way that is introspectively accessible to Jack.
The inclination itself is once again metaphysically fundamental in individuating direct and indirect seemings. Nevertheless, the nature of the inclination is often responsible for phenomenological features that are present when one is conscious of that inclination. When one is conscious of direct and indirect inclinations there is a phenomenological similarity such that they both deserve to be classified as seemings; each involves the feeling that a proposition's truth is being revealed. However, direct seemings have a spontaneity included in their phenomenology that is lacking in the case of indirect seemings. The former feel as if they're direct responses to experience and thereby strike us as cases of a proposition's truth being directly revealed. Alternatively, the latter feel as if they arise from an (often very subtle) inferential process and thereby strike us as cases where a proposition's truth is only indirectly revealed.Footnote 13
This distinction hasn't been much discussed but it's important to make if one is going to build an epistemological theory that relies heavily on seemings. Discussions of PC, for instance, tend to focus on foundational justification while largely ignoring inferential justification. Distinguishing direct and indirect seemings raises a number of interesting issues when we move to a discussion of PC and inferential justification.
Consider an argument with premises P 1-PN and the conclusion C. A view such as PC should admit that S's running through this argument only provides her with inferential justification for C if the premises seem true to S. But when we consider a few possible cases we quickly see that there are a variety of other factors that need to be considered.
Imagine that P 1-PN seem true to S and, as a matter of fact, these premises make it probable that C. Is this sufficient for P 1-PN to provide S with inferential justification for believing C? If one believes, as many advocates of PC do, that the phenomenology of a seeming plays a crucial role in the explanation of why it provides justification then you might be worried about the possibility that P 1-PN all seem true, make probable that C, but one's consideration of this argument doesn't make C seem at all true. Perhaps one should add the requirement that S's consideration of P 1-PN causes S to have a seeming that C is true. But how should we distinguish between the case where C's seeming true provides foundational justification and those where it provides inferential justification? Here the distinction between direct and indirect seemings becomes very helpful. Direct seemings provide foundational justification and indirect seemings provide inferential justification. This would allow the defender of PC to continue to insist that the phenomenology of a proposition feeling as if its truth is being revealed is a crucial aspect for why seemings provide justification without being committed to the idea that the conclusions of our arguments must seem true in the same way that foundationally justified propositions seem true.
However, things get more complicated when we notice some additional possibilities. If you place a large explanatory burden on the phenomenology of seemings in the case of foundational justification then shouldn't we treat inferential justification similarly? So what should you say about the following cases: (a) premises P 1-PN seem true, P 1-PN do not make probable C, but S's consideration of P 1-PN still produces an inferential seeming that C; and (b) the same conditions but add that S also has a direct (illusory) seeming that P 1-PN make probable C. A third and final type of interesting cases are those where an argument's premises seem true, seem to support a conclusion C (perhaps this is even a veridical seeming), yet one not only lacks a seeming that C but one actually has a seeming that C is false (good skeptical arguments might be examples of this phenomena).
I'm not going to argue what those attracted to PC ought to say regarding each of these cases, nor am I going to propose my own requirements governing inferential justification. My point is that distinguishing direct and indirect seemings becomes important when we move from discussing the roles seemings could play in the context of foundational justification to discussing seemings and inferential justification – for similar reasons, a consideration of the direct/indirect distinction would also become important when considering the notion of defeat.
3. MOTIVATING THE CONSCIOUS INCLINATIONS VIEW
What motivations are there for identifying seemings and conscious inclinations? One motivation is simply that it captures the phenomenology well. We posit the existence of seemings partly to capture a particular phenomenological phenomenon with which we're all familiar. Ernest Sosa explains:
What are these seemings? It's helpful to compare deliberation on a choice or the pondering of a question, where we “weigh” reasons pro or con. Switching metaphors, we feel the “pull” of conflicting considerations. No matter the metaphor, the phenomenon is familiar to us all. There is something it's like to feel the pull of contrary attractions as we deliberate and ponder. (Sosa Reference Sosa2007: 47)
The metaphor of pushes, pulls, and attractions towards belief is almost irresistible when describing the phenomenology of seemings. Thus, it's natural to simply identify the seeming with a subject's conscious pushes and pulls towards various beliefs, i.e. conscious inclinations.
Moreover, consider statements of the following form: “It seems (or appears) to me that P is true but I don't feel at all compelled to actually believe P.” Such statements sound infelicitous. On the non-reductive view of seemings, however, statements of this form might be true. In fact, if the non-reductive view is correct, it should be possible that we have an incredibly strong seeming that P without feeling at all inclined to believe P. Surely this is an implausible consequence. It's possible to have a seeming when the conscious inclination to form the relevant belief is weak, subtle, and difficult to notice, but if we have absolutely no conscious inclination to believe P then it isn't clear that there is any sense in which P has the feel of its truth being revealed. The simplest explanation of this is that seemings are conscious inclinations.
A second motivation for CIV are cases where we naturally describe ourselves as hosting a seeming but the only introspectively accessible states present that we could plausibly identify as the seeming are the sensation and the resulting conscious inclination. Since we've ruled out sensations as seemings, we must identify the seeming with the conscious inclination. I'll refer to this as the introspective poverty (IP) argument for CIV.
The IP argument for CIV can be most forcefully pressed within the context of introspective seemings. Consider a case where it introspectively seems to you that you're in searing pain. In such a case it's easy for me to introspectively locate the sensation and the resulting conscious inclination to believe that I'm in pain. However, on the non-reductive theory, there is a third entity: the introspective seeming. According to this view a subject has a pain sensation, this sensation produces a seeming that you're in pain, and this seeming then produces a conscious inclination to believe that you're in searing pain. Such a view strikes me as phenomenologically inadequate – I'm not alone, similar sentiments have been expressed by both Goldman (Reference Goldman and Dougherty2011) and Smithies (Reference Smithies, Smithies and Stoljar2012). When I introspect I can only find the pain and the conscious inclination to believe that this sensation is accurately described as a pain. I simply cannot find any third mental state that acts as an intermediary between the sensation and my inclination.
One non-reductivist strategy for responding to the IP argument is to claim that introspective seemings are a special case where we can successfully identify the seeming with the sensation itself. Huemer (Reference Huemer2007) considers adopting such a proposal when he mentions the plausibility of the hypothesis that my introspective belief that I'm in pain is justified by the pain itself and not by some intermediary mental state. Huemer doesn't explicitly endorse the view but its plausibility leads him to express sympathy for the idea that the pain sensation is a self-representing state and so is identical to the seeming that one is having a pain sensation.
Even if it's possible to make sense of the idea of self-representational mental states, there are two difficulties for this response to the IP argument. First, if we identify sensations of pains, color experiences, etc. with the introspective seemings that we have these sensations then it's hard to make sense of the fact that even introspective seemings can be illusory. There are interesting things one might say in response to this worry but there is an even more worrisome problem for the self-representational response to the IP argument. This second problem derives from a case analogous to the birdwatcher case used to distinguish perceptual seemings and sensations. It's possible for two subjects to have the same sensation but have distinct introspective seemings nonetheless. Consider a case where two subjects have the same kind of color sensation that's midway between red and orange. To one subject it may introspectively seem that she is having an orange color experience and to the other subject it (the qualitatively identical sensation) might introspectively seem that he is having a red color experience. If the sensation is of the same kind but the introspective seeming is distinct then we cannot identify the seeming with the sensation. Thus, given the IP argument, we should identify the seeming with the conscious inclination. By doing this we make sense of the possibility of erroneous introspective seemings since an itchy sensation could give rise to a conscious inclination to believe that one is having a very slight pain experience. We also make sense of the fact that two subjects could have the same color sensation but have distinct introspective seemings about the color sensation. The same sensation might give rise to the conscious inclination to believe that one is having a red sensation in one subject but give rise to a conscious inclination to believe that one is having an orange sensation in another subject. This IP argument and the various concerns I've mentioned are closely related to a final motivation for CIV.
If CIV can do all the same work as a non-reductive account then one of the main reasons to prefer CIV is ontological simplicity. CIV commits us to fewer kinds of entities than the non-reductive view.Footnote 14 With CIV in hand we need only admit the existence of beliefs, conscious inclinations to believe, and sensations. On the non-reductive account one is committed to each of these same entities plus an additional sui generis propositional attitude.Footnote 15
These considerations suggest that, ceteris paribus, we should prefer CIV to non-reductive accounts of seemings. The initial plausibility of CIV is high enough that a strong case will have been made for the view if it can be defended from prominent objections.
4. SUPPOSED COUNTEREXAMPLES
Huemer (Reference Huemer2007) suggests a possible counterexample to CIV. Huemer claims that it's possible for P to seem true to S even though S isn't at all inclined to believe P. Given the apparent infelicity of claims of the form “It seems to me that P but I'm not at all inclined to believe it,” we should be careful to consider a concrete case. Huemer goes on to explain that, “[o]ne might be so convinced that [a seeming] was illusory that one was not even inclined to believe its content. One could even be convinced in advance that one was going to experience an illusory [seeming], so that there would be no time at which one had the relevant inclination to believe” (Reference Huemer2007: 31). Huemer appears to have something like the following case in mind:
Expected Hallucination (EH)
Dillon knows pink elephants don't exist, and he knows that he is under the influence of hallucinogens. Dillon then has the same kind of phenomenal experience that he would usually have if there were a pink elephant in front of him.
In EH Dillon would have a seeming that there is a pink elephant in front of him but, according to Huemer, his background knowledge would prevent him not only from forming the belief that there is a pink elephant but it would also prevent the conscious inclination to form this belief.
In responding to this objection it's helpful to focus on the two aspects of a conscious inclination separately. First, let us consider the inclination itself. It's important to remember that a proper characterization of a disposition or inclination requires a ceteris paribus clause. S has an inclination to believe P iff S is in a state X that has the power to produce a belief that P provided that more powerful countervailing causal factors are not present. I see no reason to doubt that this condition is met in these cases where one preemptively expects to suffer from a hallucination or illusion. Dillon's visual sensation produces an inclination to believe that there is a pink elephant, but this is overpowered by his knowledge that pink elephants don't exist and that he is under the influence of a hallucinogenic drug.
This only shows that in EH one of the two components of a conscious inclination is present. One might therefore object that the second (i.e. the phenomenological) component of a conscious inclination would be absent in EH, i.e. Dillon wouldn't feel this inclination in this situation. Remember that conscious inclinations and seemings come in degrees. The conscious inclination to believe that there is a pink elephant produced by Dillon's sensation is likely very weak and subtle. That the inclination still shows up subtly in Dillon's consciousness is supported by the fact that he could likely introspectively recognize that he would have believed a pink elephant was in the room had he not known that pink elephants don't exist and had he forgotten that he took hallucinogens. If Dillon didn't have at least a very weak consciousness of the relevant inclination then there would be no introspective reason for him to accept the truth of such a counterfactual.
Also relevant to this kind of case is the following underappreciated fact about seemings: one can simultaneously have a seeming that P and a seeming that ~P (even when both are construed as direct seemings). Certain aspects of my mental life might produce a conscious inclination to believe P while other aspects produce a conscious inclination to believe ~P. A nice example of this might be the waterfall illusion. You might be tempted to describe this illusion as having the paradoxical content such that the riverbank seems to be both moving and stationary. A more accurate description, however, seems to be that it both seems that the riverbank is moving and it seems that the riverbank is stationary. Certain features F of one's sensation produces a conscious inclination to believe the riverbank is moving and other features F* produce a conscious inclination to believe the riverbank is stationary. Dillon is in an analogous situation. He's conscious of pushes and pulls in both directions. Some aspects of his mental life produce a conscious inclination to believe there is a pink elephant but other aspects produce an opposing conscious inclination to believe there is no such elephant. As it turns out, the conscious inclination in this latter direction is much stronger.
It is of course logically possible for a person to become so familiar with an illusion that they no longer have the relevant conscious inclination. But, in such a situation, it doesn't appear to be implausible that the proposition no longer seems true. Consider again the apparent inconsistency in the claim, “It seems to me that P is true but I don't feel at all compelled to believe P.” Or consider the particular example of the Müller-Lyer illusion. It's mysterious to me what someone has in mind when they claim that the subject doesn't feel at all inclined to believe that the lines are different lengths but that the proposition nevertheless has the feel of its truth being revealed to the subject.Footnote 16
This contention can be further supported by empirical findings – see McCauley & Henrich (Reference McCauley and Henrich2006) for an overview of these empirical studies – that susceptibility to the Müller–Lyer illusion varies across cultures. A leading empirical hypothesis is that, during the perceptual system's development, exposure to sharp carpentered corners causes a bias in our visual systems that assumes certain angles are indicative of depth. When an individual with such biases encounters the Müller–Lyer lines her visual system assumes the wedges indicate depth and thereby incline the subject to judge that the lines are different in length (as the different wedges on each line suggest 90° angles going off in different directions). Most of you reading this article have probably grown up in the carpentered environments that create these biases that produce a conscious inclination to judge the lines are different lengths. As McCauley & Henrich (Reference McCauley and Henrich2006) point out, our experiences after approximately 20 years of age have little impact on the biases of our perceptual systems. Thus, anyone reading this article is likely at a point in your life where no matter how familiar you become with the Müller-Lyer lines you cannot get rid of the conscious inclination to judge that the lines are different lengths. On the other hand, subjects whose perceptual systems develop without very much exposure to carpentered corners don't have the relevant bias. Such subjects are not susceptible to the illusion; it doesn't perceptually seem to these subjects that the lines are different lengths.
Since most of us cannot rid ourselves of the conscious inclination to judge the Müller–Lyer lines to be different lengths, better examples are more mundane “illusions” like a circular coin seen from an angle. We are so familiar with phenomena of perspective that in this kind of case most of us lack even a weak or subtle conscious inclination to believe the coin is elliptical. In this case it is natural to say that it does not perceptually seem to me that the coin is elliptical, however. This is why many philosophers considered it misleading to categorize this as a genuine illusion. When we do not feel at all inclined to form the relevant belief there is no introspectively accessible mental state that we pick out as the inaccurate seeming.Footnote 17 Thus, there isn't a case of genuine illusion.
We've been discussing a challenge to CIV that presents a supposed case where a seeming is present but the conscious inclination is absent. Cullison (Reference Cullison2010) also challenges CIV by appealing to supposed counterexamples but he works in the opposite direction. He presents cases where, supposedly, a conscious inclination is present but the seeming is absent.
Cullison's first case concerns Sam. Sam's wife recently underwent a religious conversion and he believes she will divorce him if he doesn't adopt her religious beliefs. Cullison claims that certain religious propositions might seem false to Sam, yet “he might feel an attraction toward believing them” due to his love for his wife (pp. 264–5). Thus, he concludes that P's seeming true to S is distinct from S's having a conscious inclination to believe P.
We must be careful when discussing “attractions” towards states-of-affairs. We sometimes speak of a desire for state-of-affairs Y as if the desire itself is an attraction toward Y. This is because desires are the kinds of things that often incline us to bring about the relevant state-of-affairs. Thus, when we have a conscious desire for Y we often say that we feel attracted to Y. When Sam has a conscious desire to believe the religious propositions we might thereby say that he feels an attraction toward believing the religious propositions. However, feeling an attraction in this sense just means that Sam is conscious of a certain state X (e.g. a desire) that tends to produce an inclination to believe P. This isn't the same as being conscious of the inclination produced by X. So one could feel an attraction (i.e. be conscious of a desire) towards believing P in some sense without having a conscious inclination to form that belief. CIV can therefore accommodate the claim that the religious propositions seem false to Sam even though, in some loose sense, he feels an attraction toward believing them.
I must again admit the logical possibility that Sam's conscious desire produces not only an inclination but also a conscious inclination to believe the religious propositions. In such a situation, however, it's plausible that P really does seem true to Sam. Consider the fact that when I have an experience as of imagining P it doesn't thereby seem or feel as if P's truth is being revealed to me. Nevertheless, we should admit that when my imagining that P produces a conscious inclination to accept P, P does thereby feel or seem true. Isn't this how we often think of intuition or intellectual seemings? Analogously, in a case where a desire that P produces a conscious inclination to believe P we construe the seeming as a manifestation of wishful thinking. We should remember, however, that even in this case the seeming is to be identified with the conscious inclination rather than the conscious desire causally responsible for the inclination.
Another feature of this case that helps defend CIV is that Sam's conscious inclinations to believe the religious propositions are the result of his seeming that his wife will divorce him if he doesn't form these beliefs. These are indirect seemings and don't feel as if they are a direct revelation of the proposition's truth.Footnote 18 Moreover, Sam's other mental states might produce much stronger direct and indirect conscious inclinations to believe that the religious propositions are false. Thus, Sam's utterance, “It seems to me that such-and-such religious proposition is true” could be true while striking us as inappropriate since, given conversational norms, this utterance has the implicature that this is Sam's strongest (all things considered) conscious inclination. It wouldn't strike us as odd if Sam were to say, “It kind of seems to me that such-and-such religious proposition is true but these other considerations Y and Z make it seem even more so that the proposition is false.” As I explained earlier, a subject can simultaneously have a seeming that P and a seeming that ~P. The fact that the religious propositions seem false does not entail that they don't also seem true.
Cullison's second case is autobiographical. He explains that his younger self was consciously inclined to accept his father's testimony regarding various political propositions. He goes on to explain that his attitude towards these propositions changed as an adult who had thought about the issues more carefully. He claims that his current state feels as if he's “actually apprehending the truth.” His current seeming is “phenomenally different” from his previous conscious inclination (2010: 265–6).
Cullison is appealing to a method of phenomenal contrast and we can summarize the argument as follows:
1. P seems true to adult Cullison.
2. Adult Cullison's psychological relation to P is phenomenologically distinct from any of Child Cullison's attitudinal states directed at P.
3. Therefore, Child Cullison didn't have a seeming that P.
4. Child Cullison had a conscious inclination to believe P.
5. Therefore, seemings aren't conscious inclinations.
The problematic inference is from 1 and 2 to 3. Such an inference is legitimate only if we assume that seemings directed at the same proposition are phenomenologically identical.
My earlier discussion illustrates the implausibility of the required assumption. Differences in the nature of the inclination of which one is conscious manifest themselves in differences in the phenomenology. There is a phenomenological difference between being conscious of a weak and a strong inclination to believe that P. Moreover, Cullison admits that he's had more experiences and learned more about the issues. Differences in the etiology of the inclination of which one is conscious can also lead to phenomenological differences. The mental states currently responsible for Adult Cullison's conscious inclination are different from those that produced Child Cullison's conscious inclination and therefore provide further explanation of why the states are phenomenologically distinct despite being directed at the same proposition.
Cullison notes, however, that a difference in the strength of the conscious inclination fails to explain why he only interprets his current state as actually “apprehending the truth,” and he might say something similar about my appeal to a difference in the etiology of the conscious inclinations. To explain this difference we need only note that Adult Cullison has a better and fuller understanding of the political propositions under consideration. Insofar as his current grasp of the relevant propositions is less muddled, it's unsurprising that he interprets his current conscious inclinations as if only they are a genuine apprehension of truth. Additionally, Child Cullison was inclined to accept the political propositions on the basis of his father's testimony. His current conscious inclinations are produced by his own consideration of the topics. Child Cullison had indirect seemings but Adult Cullison has direct seemings. This provides further means for explaining why Cullison would take the current (as opposed to the previous) state to be an actual apprehension of the proposition's truth without abandoning CIV.Footnote 19
5. THE EXPLANATORY OBJECTION
The final objection to CIV is the most worrisome: the explanatory objection. Huemer gives a succinct presentation:
Another reason for distinguishing [seemings] from dispositions to believe is that [seemings] may provide a non-trivial explanation for what we are disposed to believe. I am disposed to accept that there is a white cat on the couch because that is the way things [seem] to me, and this is not just to say that I am disposed to accept that there is a white cat because I am so disposed. (Huemer Reference Huemer2007: 31)
We must remember our earlier lesson that “seems” is used equivocally. If we use “seems” in two different senses so as to pick out two distinct states, then the fact that it seems to me that P in the sense of CIV can be “non-trivially explained” by the fact that it seems to me that P in some other sense. Of course, if we're to successfully defend CIV then the second use of “seems” must pick out neither the conscious inclination nor an irreducible propositional attitude.
When we explain a conscious inclination to believe P by citing a seeming, we're slipping into a varied use of seems discussed at the outset of this paper. In this context we use “seems” to pick out a particular sensation as the explanans of the fact that it “seems” to us that P in the sense that has been our primary focus (see the discussion in the introduction). We're not positing an irreducible seeming. At a linguistic level we can think about our act of explaining. When I explain to an interlocutor that my conscious inclination to believe that there is a white cat on the couch by asserting that it seems to me that there is a white cat on the couch, I'm citing the occurrence of a particular sensation and asserting that it produces my conscious inclination. Second, at an ontological level, we can discuss the explanatory relations themselves. The sensation is causally responsible for, and thereby stands in an explanatory relation to, the conscious inclination. Such an explanation isn't the trivial “I have a conscious inclination to believe P because I'm consciously inclined to believe P.” Nothing in this account requires positing an irreducible seeming.
However, there is another way to understand Huemer's explanatory objection that challenges the idea that a sensation can stand in the required explanatory relation to a conscious inclination. Why would a sensation explain a subject S's conscious inclination to form a belief with the propositional content P rather than Q unless the sensation had the same or similar propositional content?Footnote 20
Consider again the case of the Müller–Lyer lines. As we've seen, for subjects of some cultures the sensation gives rise to a conscious inclination but for subjects in other cultures the sensation gives rise to a conscious inclination to judge that the lines are the same length. So the sensation isn't sufficient to explain the presence of the conscious inclination to believe P as opposed to the conscious inclination to believe ~P. Alternatively, citing a seeming that P would (presumably) offer a satisfactory explanation. A non-reductionist about seemings can tell a nice story about the Müller–Lyer lines. For subjects who are prone to the illusion, the sensation produces a seeming that the lines are different lengths, which in turn produces a conscious inclination to form the mistaken judgment that they're different lengths. For subjects who aren't prone to the illusion, the sensation fails to produce the relevant seeming and, therefore, the conscious inclination is also absent.
Such considerations are unsuccessful in establishing a need to posit irreducible seemings. These considerations only show that a sensation can only explain the existence of a conscious inclination when taken in conjunction with various background conditions. When we use “seems” to pick out a sensation that explains our conscious inclination we assume that this sensation is a necessary part of a more complicated set of conditions that are together sufficient for the existence of the conscious inclination. We also assume that these additional conditions are usually present in our environment even if we don't know what these conditions actually are. Similarly, we might explain the occurrence of a fire by citing the fact that someone left the oven on even though leaving the oven on is insufficient for starting a fire. Other enabling conditions must be present. We might not know what these conditions are, but citing the fact that the oven was left on still picks out a fact that provides an informative explanation of why the fire started. Thus, using “seems” to cite a sensation can still provide an informative explanation of a conscious inclination given our assumption that the conditions of our environment, whatever they may be, are relatively stable, similar for different subjects, and usually produce such an inclination.
This does raise the question of what additional conditions must be met in a full explanation for the existence of a conscious inclination to believe that P. If it includes an appeal to an irreducible seeming with the same propositional content then the explanatory objection still goes though. My response to this kind of challenge will be broken up into three parts.
First, this is an empirical question and failing to answer this question doesn't thereby speak against CIV. Why a sensation produces a conscious inclination to believe P rather than Q isn't itself given in experience. We naturally assume that sensations play an essential causal role in producing our conscious inclinations to form beliefs but how they do so is often beyond our ken. The full causal explanation is a complicated empirical fact for cognitive science to investigate. That I plead ignorance of the intricate causal processes, a complicated evolutionary story, or complicated facts about neural networks involved in the mechanisms that lead from a sensation to a conscious inclination to form a belief in a particular proposition isn't reason to reject CIV. Our assumption that the sensations play a vital causal role is sufficient to make sense of why we can appeal to these sensations to explain a conscious inclination to form a belief.
Second, I want to provide a ‘just-so’ story of how a sensation could be part of a more complicated explanation of a conscious inclination to believe P rather than Q without any appeal to an irreducible seeming that P. The key here is to consider the way in which we pick out a sensation when we attempt to provide an explanation of a conscious inclination to form a belief. I want to suggest that we pick out the sensation via Chisholm's comparative use of “seems.”
In Chisholm's (Reference Chisholm1957) comparative use, “seems” is used to compare our current sensation to the kind of sensation we would usually have when looking at things of kind K. “That table (comparatively) seems red” means something akin to “the table currently looks the way red things look.” In other words, we're asserting that the table is producing a sensation in us that is of the same kind as the sensation we usually have when looking at red things in normal conditions. So when I explain the presence of a conscious inclination to believe that there is a white cat on the couch by asserting, “it (comparatively) seems to me that there is a white cat on the couch” I am both citing a particular kind of sensation and expressing a belief that associates this kind of sensation with a white cat's being on the couch. The background condition needed in order for my sensation to explain the existence of my conscious inclination is the presence of a previous psychological association of that sensation with the relevant state-of-affairs.
We can bolster this kind of response by modifying an example from Chisholm (Reference Chisholm1957). I'm looking down railroad tracks and an interlocutor asks why I feel inclined to believe the tracks are parallel. I go on to explain, “Because the tracks seem to converge.” The interlocutor and I both have background beliefs that associate one's looking at parallel lines from the current angle with the presence of a sensation similar to the kind one would have when looking at converging lines. As such, my response cites a sensation and an associative belief that together adequately explain the existence of the conscious inclination to form the belief that the tracks are parallel. Notice, however, that the content used to describe my sensation is logically inconsistent with the content of the belief towards which I feel inclined. In order to understand my response as citing features of my mental life that explain the presence of the conscious inclination we need to interpret this as a comparative use of “seems.” The comparison or association rather than any irreducible propositional attitude is what does the explanatory work.
At this point one could ask what explains the presence of my association between a sensation and the relevant state-of-affairs. One possibility is to appeal to concept acquisition or training. Consider again the case of the novice (Jack) and the expert (Jill) birdwatcher. Both look at the same bird, have the same kind of visual sensation, but only Jill's sensation produces a conscious inclination to believe the bird is a Goldfinch. Why? It's natural to appeal to associative mechanisms utilized during Jill's training. During her training, a certain kind of sensation is produced in Jill by showing her actual Goldfinches, photographs of Goldfinches, or drawings of Goldfinches. After producing the sensation in Jill, the trainer (or the book Jill is reading) provides testimony so as to bring it about that Jill associates salient features of her current sensations with the presence of a Goldfinch. Her training thereby provides a nice explanation of why this particular sensation gives rise to a conscious inclination to believe that there is a Goldfinch rather than a Canary present. There is no need to appeal to the claim that the sensation produces the conscious inclination because it (or some intermediary state that it produces) has the same or similar propositional content as the belief to which she is consciously inclined.
Similar remarks apply mutatis mutandis to how associative mechanisms utilized in concept acquisition can provide a general explanation of the conditions under which a sensation would give rise to the conscious inclination that it does. Children acquire concepts such as <grey>, <elephant>, etc. because (a) adults produce in the child a sensation with salient similarities to the kind of sensation that would be present in cases where these properties are actually instantiated and then (b) the adults utter the words “grey”, “elephant”, etc. If we were to try to teach a child <white> by presenting a white ball illuminated with red lights and then we uttered the word “white” we would no doubt set the child up for mass confusion later. The processes we use to facilitate concept acquisition appeal to associative mechanisms that thereby explain why a sensation gives rise to a conscious inclination to believe the proposition that it does. Yet again, there is no need to appeal to the claim that the sensation has (or produces an intermediary irreducible seeming that has) the same or similar propositional content.
Finally, this version of the explanatory challenge isn't unique to CIV. Analogous questions arise on the non-reductive account of seemings and an advocate of CIV can piggyback on the answers given by the non-reductivist. More importantly, such piggybacking will always produce a simpler explanation of the conscious inclination that will show an appeal to irreducible seemings to be superfluous. Let me explain.
We can always ask why a state S gives rise to a state S* with the propositional content P. Eventually our explanation will have to appeal to a state that lacks propositional content. We've already seen that there are good reasons to distinguish sensations and seemings and many advocates of the non-reductive view accept such a distinction – see Cullison (Reference Cullison and Tucker2013) and Tucker (Reference Tucker2010, Reference Tucker2013). How does a sensation cause a conscious inclination to believe P according to these philosophers? Presumably the sensation produces the irreducible mental state of its seeming to S that P which then ‘non-trivially’ explains the conscious inclination to believe that P. But now we can ask: why does the sensation give rise to an irreducible seeming with propositional content P rather than Q? A natural suggestion would be to appeal either to something like my previous story about an association produced by past training or to give some sort of evolutionary story. Either way, whatever story is given can be co-opted by the advocate of CIV so as to explain why the sensation gives rise to a conscious inclination to believe that P while bypassing any appeal to an irreducible seeming.Footnote 21 Piggybacking on the explanation offered by the non-reductivist circumvents a need to posit an irreducible seeming as a middleman and thereby produces a simpler explanation of the conscious inclination.
6. IMPLICATIONS
I want to briefly consider some implications of CIV. Specifically, I want to consider the following questions: Is Phenomenal Conservatism (PC) plausible given CIV? What kind of epistemological role (if any) can seemings play? Given that my main goal is to defend reductive accounts of seemings by illustrating the plausibility of CIV, my discussion will be brief and inevitably require development in future work.
According to PC the mere occurrence of a seeming that P provides a subject with some degree of defeasible justification for believing P. I admit that PC has pre-theoretic plausibility. Moreover, PC can provide a unified account of how a lot of different kinds of beliefs are justified; this is a rather attractive feature. However, as Cullison (Reference Cullison2010) notes, some people might worry about PC because they don't have an adequate grasp on what seemings are, and to be told that they are irreducible or sui generis isn't very helpful (p. 273). In this regard, CIV might be seen as a helpful amendment to PC.
Moreover, my distinction between direct and indirect seemings can be helpful when philosophers attracted to PC consider the epistemic role of seemings in inferential justification. Despite these ways that my development of CIV could help advocates of PC, however, our previous discussions and an acceptance of CIV can also make PC seem less attractive.
It might turn out that the existence of a seeming is always accompanied by prima facie justification. Or, more cautiously, the existence of a direct seeming might always be accompanied by prima facie (foundational) justification. Even if true, advocates of PC need to explain why seemings necessarily provide this justification. Unfortunately, once we realize that a seeming is a conscious inclination to believe, PC loses much pre-theoretic plausibility as an account of justification. Why think the mere fact that I have a conscious inclination to believe P necessarily provides justification? The mere fact that I have a conscious inclination to believe P doesn't appear to make it probable that P. Is the fact that I have a conscious inclination to believe P really an epistemic reason to believe P is true? A conscious inclination to believe P doesn't appear to be the kind of thing that can provide me with an epistemic (as opposed to a merely psychological) assurance that P is true.
Moreover, conscious inclinations can be produced in any number of ways. It's natural to think that what matters most for justification is not the conscious inclination itself but rather the mental states responsible for such an inclination. Is my conscious inclination to believe P produced by evidence that supports P or something completely unrelated to P's truth? My inclination might be the result of an epistemic vice such as wishful thinking, self-deception, or some random cognitive malfunction. Surely we don't want to grant that a conscious inclination to believe P provides justification when it's the manifestation of wishful thinking or self-deception. Worries surrounding this kind of cognitive penetration problem have been discussed extensively in the literature – see Markie (Reference Markie2005, Reference Markie2006) and various papers in Tucker's (Reference Tucker2013) anthology. However, the problem of seemings with an inappropriate etiology is even more worrisome once we've adopted CIV. If CIV is correct then it's natural to identify the basis of a subject's belief not with the conscious inclination itself but with the state that produces the inclination.Footnote 22 This would mean that in cases of wishful thinking a subject S doesn't base her belief on a seeming produced by a desire. In such cases, S bases her belief on the desire itself. A desire that P cannot justify believing that P absent some reason to think a subject's desires are reliably correlated with their contents being true.
I'm not arguing that seemings never provide justification. My point is that if we adopt CIV then it's natural to think the seeming must meet some further condition in order to be epistemically efficacious. Perhaps the seeming only provides justification when it's the manifestation of a reliable cognitive virtue. Sosa (Reference Sosa2007) appears sympathetic to this kind of view regarding a priori justification.
Of course, most epistemologists interested in seemings are appealing to such states in an attempt to develop a plausible form of internalism. Philosophers with internalist sympathies obviously won't like this reliabilist suggestion for when seemings are epistemically efficacious. More internalist friendly proposals include: (i) the seeming must be produced by evidence for the proposition; (ii) the seeming must be accompanied by evidence that the conscious inclination is reliable; (iii) the best explanation available to S for the existence of her conscious inclination is that the belief towards which she is inclined is true; or (iv) a conjunction or disjunction of these three options. I'll note my internalist sympathies but I won't take a stand on exactly when a seeming is accompanied by prima facie justification. Either way, PC no longer states a necessary truth about epistemic justification. The intuitive appeal of PC as an account of justification is lost on CIV.
Even if I'm correct that seemings don't necessarily provide justification, this leaves open the possibility that they are still of epistemological significance. The important point is that a seeming doesn't do the justificatory work unless other enabling conditions are met. In order to explain the justificatory power of a seeming we need to complete the honest toil of explaining how the mental states which produce the conscious inclination support the relevant belief, identifying the evidence we have for believing that our conscious inclinations (or some subset of them) are generally reliable, or illustrating why the existence of the conscious inclination is best explained by the belief's truth.Footnote 23
Finally, to close the paper I want to suggest that an appeal to seemings might be fruitful for shedding light on epistemological controversies where they haven't yet been utilized. Some of the things I've said in this section hint at the possibility that a central epistemological role for seemings isn't as evidence or proper bases of our beliefs but rather in an account of the basing relation itself. In future work I hope to defend a view of basing in the spirit of the following: a belief Bp is based on E iff: (a) E causes Bp and (b) either E produces a seeming that P or we have a seeming that E supports P. Defending such an account, however, must await future work. My point in mentioning this possibility is that even if CIV prevents seemings from doing the heavy lifting in epistemology that some have hoped, it doesn't follow that there aren't important roles for seemings to play in a fully developed epistemology.Footnote 24
