1. The Intuition Picture
Intuitions are widely assumed to play a crucial epistemic role in philosophy. In particular, intuitions are thought to play a crucial evidential role in the testing of philosophical theories by the method of cases, for instance in supporting Gettier's (Reference Gettier1963) counterexamples to the JTB theory of knowledge. Thus, many philosophers assume:
(E1) Intuitions are assigned an important evidential role in philosophy.
The assumption that E1 is true is usually common ground between sceptics and defenders of armchair philosophy, who go on to disagree about the further claim that intuitions should be assigned this role:
(E2) Intuitions should be assigned this evidential role in philosophy.
Philosophers who endorse E1 and E2 typically endorse the claim that intuitions are distinct from mere beliefs:
(E3) The intuitions that are assigned this evidential role in philosophy are importantly distinct from any mere belief or inclination to believe.
Some hold that intuitions are a special kind of doxastic state with certain distinctive properties (e.g. Ludwig 2007; Sosa Reference Sosa2007). Others hold that intuitions are sui generis and cannot be analysed in terms of more familiar mental states (e.g. Bealer Reference Bealer, DePaul and Ramsey1998). However, proponents of both kinds of views will often attribute many of the same properties to philosophical intuitions and, in particular, endorse closely related versions of the following claim:
(E4) The intuitions that are assigned this evidential role in philosophy are rational intuitions.
I will label the view shared by those philosophers who endorse all of E1–E3 the intuition picture of philosophical methodology – and in the discussion to follow I will be concerned with versions of this view that also endorse E4. I call this view a ‘picture’ because it is obviously under-described and could be developed in many different ways, depending on how one conceives of this supposed evidential role and the nature of the intuitions that play that role. However, the intuition picture does capture key commitments of a number of important accounts of philosophical methodology, including those given by Bealer (Reference Bealer, DePaul and Ramsey1998), Chudnoff (Reference Chudnoff2010), Goldman and Pust (Reference Goldman, Pust, DePaul and Ramsey1998), Grundmann (Reference Grundmann2007), Huemer (Reference Huemer2005), Ludwig (2007), Pust (2000) and Sosa (Reference Sosa2007).Footnote 1 And, arguably, the intuition picture is implicitly endorsed by many of the numerous philosophers who employ the method of cases.
Importantly, not all defenders of armchair philosophy embrace the intuition picture, with perhaps the most notable exception being Williamson (Reference Williamson2004, Reference Williamson2007). Williamson's most central objection to intuition-based accounts of philosophical methodology is that they are committed to a psychologistic conception of philosophical evidence that encourages scepticism about the armchair judgements relied upon in philosophy. The purpose of this paper is to defend the intuition picture against this psychologisation objection. (There are other less direct objections to the intuition picture suggested by Williamson's work on philosophical methodology. For reasons of space I will not examine these objections here, although see below for an indication of how I think they can be addressed.Footnote 2)
2. The Psychologisation Objection
What do philosophers mean when they claim that one's evidence that the JTB theory is false is one's intuition that the Gettier subject has non-knowledgeable justified true belief (henceforth, NKJTB)? Williamson holds that when philosophers say such things what they mean is that one's evidence consists in the psychological fact that one has the intuition that the Gettier subject has NKJTB. More generally, Williamson (Reference Williamson2004: 119) says that this
is the uneasy conception which many contemporary analytic philosophers have of their own methodology. They think, that, in philosophy, our ultimate evidence consists only of intuitions. Under pressure, they take that not to mean that our ultimate evidence consists of the mainly non-psychological putative truths that are the contents of those intuitions. Rather, they take it to mean that our ultimate evidence consists only of intuitions.
According to Williamson, this psychologistic view arises in response to a sceptical challenge to the judgements that philosophers rely on. Williamson (Reference Williamson2004: 117–18) draws a close analogy between this challenge and traditional challenges to our perceptual knowledge of the external world. The sceptic about perception tries to narrow one's evidential base to one's current mental state. Suppose one looks outside and see that it is raining. One might think that one's evidence now includes the fact that it is raining. However, the sceptic contends that one's evidence only includes a fact about one's own mind, namely, that it perceptually appears to one that it is raining, for sometimes what perceptually appears to be is not the case. The sceptic challenges one to justify one's taking this psychological fact to be good evidence that it is raining. Similarly, after reading Gettier's article one might think that one's evidence includes the fact that the Gettier subject has NKJTB. However, the judgement sceptic contends that one's evidence only includes the fact that it intellectually appears to us (perhaps the members of some restricted group) that this is so, for sometimes what intellectually appears to us is not the case. The sceptic challenges one to justify one's taking this psychological fact to be good evidence that the Gettier subject has NKJTB.
One way of responding to the perceptual sceptic would be to agree that one's evidential base is merely the fact that it appears to one that it is raining, and then try to meet their challenge of showing how one can reason outwardly from this psychological fact to the proposition that it is raining. Williamson (Reference Williamson2004: 17) thinks that proponents of the intuition picture like Bealer are making the parallel response to the judgement sceptic: ‘If scepticism about judgment is treated by analogy with scepticism about perception, then its evidential base will be described as intellectual appearances, somehow analogous to perceptual appearances. George Bealer has defended just such an account of intuitions as intellectual seemings.’
Williamson holds that in both cases this kind of response encourages the very scepticism it seeks to avoid. In the case of the judgement sceptic philosophers are left with the unenviable task of trying to argue for the ‘hoped-for correlation between having an intuition that P and its being true’. For example, in the Gettier case they are left with ‘the challenge of arguing from a psychological premise, that I believe or we are inclined to believe the Gettier proposition, to an epistemological conclusion, the Gettier proposition itself. That gap is not easily bridged’ (2007: 211).
According to Williamson (Reference Williamson2004: 118), philosophers psychologise their evidence because they are committed to an operational standard of evidence, according to which one must always be in a position to know what one's evidence is.Footnote 3 This commitment leads philosophers to conclude that our evidence must be the same in both the ‘good case’ (where it perceptually or intellectually seems to one that p and p is true) and the ‘bad case’ (where it still seems to one that p but p is false) because: ‘typically, if one is in a sceptical scenario, one is in no position to know that one is not in the corresponding non-sceptical scenario, and is therefore in no position to know that one's evidence differs from what it would have been in that non-sceptical scenario’ (2004: 118). The proper response to both challenges is to resist the operational standard. Once we acknowledge that we are not always in a position to know what our evidence is, we can hold that in the good case one's evidential base can include the non-psychological content of one's perceptual/intellectual seeming even if one would lack that evidence in the bad case.
Suppose proponents of the intuition picture do endorse a psychologistic view of the evidential role of intuitions. Furthermore, suppose this view does encourage scepticism about the armchair judgements that proponents of the intuition picture hope to defend. Given these assumptions one might be tempted to press the following argument against the intuition picture:
The Psychologisation Argument
(1) The Psychologisation Premise: Proponents of the intuition picture psychologise our evidence in philosophy (in an attempt to answer the judgement sceptic's challenge).
(2) The Gap Premise: Psychologising our evidence in philosophy encourages scepticism about the armchair judgements relied upon in philosophy.
Therefore,
(3) The Sceptical Conclusion: Proponents of the intuition picture encourage scepticism about the armchair judgements relied upon in philosophy.
The argument is valid, and no proponent of the intuition picture will want to embrace its conclusion, or at least they will want to claim that their version of the intuition picture does not encourage scepticism about the armchair judgements relied upon in philosophy. Faced with this argument then, a proponent of the intuition picture will want to do one of two things. First, they could resist the psychologisation premise by claiming that at least their version of the intuition picture does not psychologise our evidence in philosophy. Or, alternatively, they could accept the psychologisation premise whilst denying the gap premise. That is, they could embrace a psychologistic view of the relevant philosophical evidence, but then contest Williamson's claim that if true such a view would have sceptical consequences for armchair philosophy. The latter response is worthy of exploration but I will not develop such a response here.Footnote 4 Rather, I shall contest the psychologisation premise. More precisely, I will argue that there is no strong connection of any kind between endorsing the intuition picture and adopting a psychologistic conception of philosophical evidence.
3. Resisting the Psychologisation Premise
Let us distinguish two questions. Does a proponent of the intuition picture need, in principle, to psychologise philosophical evidence? Do actual proponents of the intuition picture psychologise philosophical evidence? The answer to the first question is clearly no. Take the claim that, after reading Gettier's article, one's evidence that the JTB theory is false is one's intuition that the Gettier subject has NKJTB. This claim, as many have noted,Footnote 5 is ambiguous. It could be interpreted as saying either: (i) that one's evidence consists in the psychological proposition I have the intuition that the Gettier subject has NKJTB; or (ii) that one's evidence is the non-psychological content of one's intuition, namely the proposition the Gettier subject has NKJTB.Footnote 6 The second interpretation is perfectly legitimate, and what it shows us is that the intuition picture can, in principle, be interpreted in a way such that it does not psychologise our evidence in philosophy.
As we have seen, Williamson implicitly acknowledges the possibility of a non-psychologistic interpretation of the idea that intuitions are evidence in philosophy but he thinks that in fact proponents of the intuition picture are endorsing (i), and not (ii), when they say that our evidence in the Gettier case is our intuition that the Gettier subject has NKJTB. This brings us to our second question. Do actual proponents of the intuition picture endorse the psychologistic interpretation?
Bealer is Williamson's sole example of a philosopher who supposedly endorses a psychologistic view of philosophical evidence in response to pressure from the judgement sceptic. Williamson is right that on Bealer's view the intuitions that play an evidential role in philosophy are intellectual appearances or seemings.Footnote 7 But Bealer (Reference Bealer, DePaul and Ramsey1998) couldn't be clearer in explicitly disavowing a psychologistic interpretation of this evidential role: ‘When I say that intuitions are used as evidence, I of course mean that the contents of the intuitions count as evidence’ (Bealer Reference Bealer, DePaul and Ramsey1998: 205; emphasis original). More generally, it is quite hard to identify proponents of the intuition picture that commit themselves to a psychologistic view of philosophical evidence – Goldman and Pust (Reference Goldman, Pust, DePaul and Ramsey1998) being a notable exception.Footnote 8 What one often finds instead is that when proponents of the intuition picture make claims like ‘intuitions are used as evidence in philosophy’, they simply do not clarify whether these claims should be interpreted in a psychologistic or a non-psychologistic way.
There are also reasons for thinking that a psychologistic interpretation of such claims would be uncharitable. This is because most proponents of the intuition picture hold that the knowledge one gains on the basis of having the Gettier intuition is a priori. But how could this be if it is based on evidence consisting of a contingent fact about one's psychology? Presumably such evidence will typically be gained via introspection, which is normally thought to be a source of a posteriori not a priori knowledge. Indeed, we will see in §5 that when proponents of the intuition picture characterise rational intuitions they usually stipulate that they are not based on introspection.
The situation then is this. The psychologisation premise is false if it is interpreted as a claim about all possible versions of the intuition picture, as Williamson in effect acknowledges. It is also false if it is interpreted as a claim about all actual versions of the intuition picture, as the Bealer quote demonstrates. The premise also looks implausible if it is interpreted as a generalisation. For often proponents of the intuition picture do not explicitly commit themselves to either a psychologistic or a non-psychologistic interpretation of their claims about the evidential role of intuitions, and there are reasons to think that they are implicitly committed to a non-psychologistic interpretation of such claims. All we are left with is the fact that the premise is true if it is interpreted as the mere claim that some proponents of the intuition picture psychologise philosophical evidence. But the truth of the existential claim gives us no reason to think that there is any strong connection between accepting the intuition picture and endorsing a psychologistic conception of philosophical evidence.
4. The Partiality and Revenge Problems
Can we now dismiss the psychologisation argument? Yes, but at this point Williamson might well argue that there will be new and serious objections to be made against the non-psychologistic interpretation of the intuition picture. To see why, we need to think about how a non-psychologistic version of the intuition picture might be developed. We know that the proponent of such a view will say that, after reading Gettier's article, the propositional content of the Gettier intuition will be a part of one's evidence, and that this evidence can be used as part of an argument against the JTB theory. But the mere claim that the contents of our intuitions are used as evidence in philosophy is close to a trivial truth. Presumably, any proponent of the intuition picture will want to add that there is a substantive connection, of some kind, between the fact that after reading Gettier's article one comes to have the Gettier intuition and the fact that the content of that intuition also comes to be part of one's evidence.
On a psychologistic interpretation of the intuition picture it is obvious what the connection is: the fact that one has the intuition that the Gettier subject has NKJTB is one's evidence for the truth of the proposition that the Gettier subject has NKJTB. But what is the connection on a non-psychologistic interpretation of the intuition picture?
Probably the most obvious suggestion to make would be to offer a view according to which simply having an intuition that p is itself a sufficient condition for one's having the proposition that p as a part of one's evidence, as opposed to being one's evidence for the proposition that p. That is, one could endorse the following claim:
I → E: If S (rationally) intuits that p, S's evidence includes p.Footnote 9
My impression is that Bealer is tempted to endorse something like I → E, although he never explicitly does so. Suppose, for the sake of argument, that Bealer did endorse I → E. Given Bealer's views on the nature of intuitions it would follow that he is committed to denying the following claims:
E → T: If S's evidence includes p, p is true.
E → B: If S's evidence includes p, S believes that p.
The reason Bealer would have to reject E → T and E → B is that he holds that one can have an intuition that p even when p is false and that one can have an intuition that p even when one does not believe p.Footnote 10 And most proponents of the intuition picture join Bealer in endorsing at least the first of these two claims. But if non-psychological versions of the intuition picture are committed to denying that evidence is factive and belief-entailing, Williamson should take this to be a reductio of such views, given that knowledge is factive and belief-entailing and his view that E = K:Footnote 11
E = K: S knows that p if, and only if, S's evidence includes p.
The non-psychologistic interpretation of the intuition picture (henceforth, simply ‘the intuition picture’) appears then to be committed to very controversial theses about the nature of evidence and evidence possession – what I will call the partiality problem. Importantly, the problem is not merely that the intuition picture is inconsistent with the views of one particular philosopher. For one thing, Williamson's views on evidence have been highly influential and have led other philosophers to accept E = K. And, more generally, many philosophers who reject E = K still accept E → T and/or E → B.
Furthermore, if a proponent of the intuition picture endorses I → E Williamson should think that the resulting view will still encourage the judgement sceptic – I will call this the revenge problem. This is because Williamson's response to the judgement sceptic rests on the claim that one's evidence in the good case (where one intuits that p and p is true) is different from one's evidence in the bad case (where one intuits that p and p is false). It is only in the good case that one's evidence includes the non-psychological proposition that is the content of one's intuition, but if I → E is true it follows that one's evidence includes this proposition in both cases. In which case, a proponent of I → E cannot join Williamson in appealing to an asymmetry in one's evidence between the good and bad cases to support the claim that one knows that p in the good but not the bad case.
How should proponents of the intuition picture respond to the partiality and revenge problems? They could dismiss these supposed problems by providing arguments against E → T, E → B and Williamson's response to the judgement sceptic. And this would be a reasonable strategy, given that the plausibility of Williamson's response to the perceptual sceptic (on which his response to the judgement sceptic is modelled), and the status of theses like E → T and E → B, are themselves issues that are (at least) as controversial as the intuition picture itself is.Footnote 12 However, other things being equal, it would surely be preferable for proponents of the intuition picture to show that the intuition picture is not committed to any particular stance on these highly contested matters. Furthermore, if this can be done, I think it would deepen our understanding of the intuition picture to see that, while individual proponents of this view often appear to be committed to rejecting E = K and Williamson's response to the judgement sceptic, there is actually no tension between the intuition picture and the central tenets of his ‘knowledge-first’ epistemology.
For these reasons, in §6 I will respond to the partiality problem by showing how proponents of the intuition picture can (if they choose) endorse E → T and E → B.Footnote 13 My strategy will be to assume that E = K is true and then show how a version of the intuition picture can be made consistent with this assumption. In which case, it will follow that the resulting version of the intuition picture will also be consistent with E → T and E → B.Footnote 14 It will also be clear that a proponent of this view can employ Williamson's strategy for responding to the judgement sceptic, hence avoiding the revenge problem. However, before turning to this response to the partiality and revenge problems it will be useful to first take a closer look at what proponents of the intuition picture mean when they claim that philosophy relies not just on intuitions but on rational intuitions.
5. Rational Intuition
My aim now is to unpack the notion of a rational intuition using Bealer's prominent view as our starting point. Bealer claims that ‘For you to have an intuition that A is just for it to seem to you that A’ (Reference Bealer2000: 30), where these seemings are sui generis and cannot be reduced to any kind of doxastic state.Footnote 15 Bealer qualifies this claim by saying, ‘of course, this kind of seeming is intellectual, not sensory or introspective (or imaginative)’. This claim in turn is also qualified as applying only to rational intuitions which are distinguished from physical intuitions like one's ‘intuition that, when a house is undermined, it will fall’ (1998: 207). Putting these ideas together we get the following claim:
To have a rational (as opposed to physical) intuition that p just is to have an intellectual (as opposed to sensory, imaginative, or introspective) seeming that p.
The idea that philosophical intuitions are rational intuitions (or intellectual seemings), and the negative etiological characterisation of these as intuitions that do not have their source in perception or memory or introspection etc, can be found in the work of a number of theorists. A positive etiological characterisation is also usually added, namely, that rational intuitions are based on one's understanding of the intuited content, perhaps together with some process of reasoning or thinking about that content.Footnote 16
Importantly, these characterisations are also endorsed by proponents of the intuition picture who hold that intuitions are a special kind of doxastic state. Sosa (Reference Sosa2007), for example, holds that intuitions are a special sort of intellectual seeming where, unlike Bealer, this is simply an attraction to assent, as opposed to an experiential-type state that attracts one to assent. But Sosa does offer etiological characterisations of rational intuitions that are closely related to those given by Bealer. On Sosa's view intuitions are distinguished from other attractions to assent by the fact that the ‘rational basis for attraction to assent to the propositional content lies in nothing more than [one's] conscious entertaining of that content’ (2007: 53). Furthermore, an intuition is a rational intuition only if it has no ‘reliance on introspection, perception, memory, testimony or inference’ (2007: 61).
Rational intuitions are sometimes also characterised in terms of their phenomenology. Bealer, for example, holds that seemings, unlike guesses or hunches, in some sense ‘present’ their propositional contents as being true.Footnote 17 Bealer also suggests that rational intuitions present themselves as not merely true but necessarily true.Footnote 18
To sum up, we have identified four key claims that Bealer uses to characterise the nature of rational intuitions/intellectual seemings:
The Negative Etiological Thesis: Rational intuitions are not based on perception, testimony, memory or introspection.
The Positive Etiological Thesis: Rational intuitions are based on one's understanding of the intuited content (perhaps together with some process of reasoning or thinking about that content).Footnote 19
The Presents as True Thesis: Rational intuitions, like all intuitions or seemings, have a presentational phenomenology.
The Presents as Necessarily True Thesis: Rational intuitions present their contents as being necessarily true.
The etiological theses are endorsed by most, if not all, proponents of the intuition picture. But proponents of doxastic versions of the intuition picture, like Ludwig and Sosa, tend to reject or downplay the significance of the kind of phenomenological claims made by the presentational theses.Footnote 20
6. Intuition, Insight and E = K
Now we have a clearer grasp on the notion of a rational intuition we can return to the partiality and revenge problems. Recall, the challenge now is to show how a proponent of the intuition picture can maintain that intuitions play an important evidential role in philosophy given the assumption that E = K. The proponent of what I will call the knowledge version of the intuition picture (KI) meets this challenge by appealing to the idea that intuiting is a way of knowing. This idea may seem hopeless for reasons already raised, namely, that knowledge is factive and belief-entailing whereas intuition is not. To address this worry we need to adopt an account of what ways of knowing are. Two well-developed accounts are: Williamson (Reference Williamson2000: ch. 1) and Cassam (Reference Cassam2007). Given our engagement with Williamson and E = K, I will focus more on Williamson's account but it will also be useful to note some distinguishing features of Cassam's account.
Williamson and Cassam on ways of knowing
Williamson's account of ways of knowing is intimately related to his view that knowledge is the most general factive stative attitude (2000: 33–48). Williamson tells us that a propositional attitude is factive, iff, necessarily, it is an attitude one has only to true propositions. Not all factive attitudes are mental states. For example, forgetting is a mental process rather than a mental state. The factive stative attitudes include such mental states as seeing or perceiving that something is so, remembering that something is so and knowing that something is so. The claim that knowledge is the most general factive mental state amounts to the claim that knowledge is the factive stative attitude one has to a proposition if one has any factive stative attitude to that proposition at all.
To illustrate this idea, Williamson draws an analogy between the state of knowing and the property of being coloured. Williamson (Reference Williamson2000: 34) writes: ‘If something is coloured then it has some more specific colour property; it is red or green or … Although that specific colour property may happen to lack a name in our language, we could always introduce such a name, perhaps pointing to the things as a paradigm.’ The analogy is then drawn between the property of being coloured and the mental state of knowing: ‘Similarly, if one knows that A then there is a specific way in which one knows; one can see or remember or … that A. Although the specific way may happen to lack a name in our language, we could always introduce such a name, perhaps pointing to the case as a paradigm’ (2000: 34). And, in the other direction, if one sees, or remembers, that A, it follows that one knows that A. As Cassam (2009: 348) points out, the implication of Williamson's analogy is that the relationship between a specific way of knowing and knowledge is a determinate-determinable one, just like the relationship between being coloured and being some specific colour.
Williamson also claims that corresponding to some specific ways of knowing there are also mental processes that aim at those mental states: ‘While a belief aims at knowledge, various mental processes aim at more specific factive mental states. Perception aims at perceiving that something is so; memory aims at remembering that something is so.’ As part of his ‘knowledge-first’ approach to epistemology, Williamson rejects the project of trying to analyse knowledge as some special kind of belief. Instead, he advocates taking knowledge to be a primitive that can then be used to ‘elucidate’ other important epistemic and mental notions. For example, Williamson offers a characterisation of belief according to which ‘believing p is, roughly, treating p as if one knew p’, and mere believing is treated as ‘a kind of botched knowing’ (2000: 47). Similarly, on this approach one might think of a perceptual appearance that p as an apparent seeing that p, and a mere perceptual appearance as a kind of botched seeing.
On Cassam's account, ways of knowing are not determinates of the determinable that is knowing. Rather, Cassam (Reference Cassam2007: 340) gives an explanatory account according to which Φ-ing that p is way of knowing that p just if it is possible to satisfactorily explain how S knows that p by pointing out that S Φs that p. One worry with Williamson's account is that regretting is a factive stative attitude but, intuitively, it is not a way of knowing. An attractive feature of Cassam's account is that it predicts this result given that normally one cannot explain how one knows that p by citing the fact that one regrets that p. Another important difference between the accounts is that on Williamson's view ways of knowing entail knowledge, but Cassam rejects this restriction because, for example, that Bill read that James was born in Akron is a possible explanation of how Bill knows that James was born in Akron even though ‘S read that p’ does not entail ‘S knows that p’.
KI and rational insights
The proponent of KI claims that as well as seeing and remembering, etc., there is at least one other distinctive way of knowing, one that plays a crucial evidential role in philosophy. What is this way of knowing? Given that intuitions are non-factive it seems that having a rational intuition cannot be the relevant way of knowing. In which case, the proponent of KI might have to admit that we lack a natural language term for this way. But, as Williamson notes, we could always introduce a term for it, perhaps by pointing to paradigm cases. For example, if we assume that we do know that the Gettier subject has NKJTB then the proponent of KI could point to this as an instance of this distinctive way of knowing. The proponent of KI might say that in this case one has a genuine rational insight that the Gettier subject has NKJTB, rather than just a mere rational intuition or seeming.Footnote 21
Now, as a matter of fact, I think a proponent of KI could make a reasonable case that there is a factive sense of ‘intuition’ and its cognates that picks out this putative way of knowing. For example, if one looks up the verb ‘intuit’ in a dictionary the definitions one finds often seem to be appealing to something like the notion of a genuine rational insight:
2 a. intr. or absol. To receive or assimilate knowledge by direct perception or comprehension.
2 b. trans. To know anything immediately, without the intervention of any reasoning process; to know by intuition.Footnote 22
transitive verb: to know, sense, or understand by intuitionFootnote 23
Definitions like these with their use of phrases like ‘to know by direct comprehension’ appear to support the hypothesis that there is a factive sense of ‘intuits’. If there is such a sense of ‘intuits’ the proponent of KI who accepts E = K could appeal to it and endorse both I → E and the claim that intuiting is a way of knowing, as long as they added the clarification that ‘intuits’ in these claims should be interpreted in the factive way.Footnote 24
However, the important issue is not whether there is a factive sense of ‘intuits’. The important issue is whether a proponent of KI can identify this supposed way of knowing. And a proponent of KI will not only think that they can do this by pointing to paradigm cases, they will also think they can offer a partial characterisation of what unites those cases. In particular, the proponent of KI can utilise any of the standard characterisations of rational intuitions. For example, echoing the etiological theses, the proponent of KI can say that a rational insight is knowledge, the source of which is not perception, or memory, or testimony, etc., and that it is knowledge gained simply on the basis of one's understanding of the known proposition and one's reasoning capacities. The proponent of KI might also echo the presentational theses and claim that this way of knowing has a distinctive presentational phenomenology, and that when one has a rational insight that p this proposition is presented as being not merely true but necessarily true.
It is true that such characterisations do not distinguish rational insights from mere rational intuitions. However, identifying such properties of rational insights can still be a way of distinguishing this distinctive way of knowing from other ways of knowing. Of course, if the proponent of KI adopts not only Williamson's account of ways of knowing, but also his knowledge-first approach to epistemology, then they will want to deny that one can analyse what it is to have a rational insight that p in terms of a rational intuition that p and certain further conditions. But this non-analysability assumption is consistent with the idea that we can usefully distinguish rational insight from other ways of knowing by pointing to certain features that it shares with rational intuition.
Proponents of the intuition picture frequently endorse various analogies between perception and intuition. Similarly, the proponent of KI will not only want to claim that having a rational insight is a way of knowing but that it is a way of knowing that is more closely analogous to perceptual ways of knowing than other ways of knowing. For example, unlike remembering, and like seeing, having a rational insight is most naturally thought of as a way of coming to possess knowledge rather than a way of retaining it. One could also use the notion of a rational insight to elucidate the notion of a rational intuition, in a way analogous to the Williamsonian elucidation of perceptual appearances given above. One might say that a rational intuition is an apparent rational insight and that a mere rational intuition is a kind of botched rational insight. Similarly, one could characterise the mental process of rational intuition as that process which aims at having a rational insight that something is the case. And a proponent of KI can adopt any of the more familiar analogies between perception and intuition, for example, that they are both basic sources of evidence (Bealer Reference Bealer, DePaul and Ramsey1998) and that perceptual states and intuitions share a presentational phenomenology (e.g. Bealer Reference Bealer, DePaul and Ramsey1998; Huemer Reference Huemer2005; Chudnoff Reference Chudnoff2010).
KI and the role of intuitions
The imagined proponent of KI will characterise rational insights in something like the way just described. To see how they would characterise their evidential role let us return to E1 and E2:
(E1) Intuitions are assigned an important evidential role in philosophy.
(E2) Intuitions should be assigned an important evidential role in philosophy.
The proponent of KI can endorse both claims. But, given that these claims are ambiguous, we should specify the particular interpretations of these claims that they will endorse:
(E1*) Intuited propositions are assigned an important evidential role in philosophy.
(E2*) Intuited propositions should be assigned an important evidential role in philosophy in those cases where they are known to be true (where typically the way philosophers know such propositions is by having a rational insight).
E1* clarifies that KI is a non-psychologistic version of the intuition picture. The proponent of KI claims only that intuited propositions – as opposed to propositions about what intuitions we have – play an important evidential role in philosophy. Of course, the proponent of KI can allow that philosophers often have, as part of their evidence, psychological propositions of the form ‘S has the intuition that p’, and they could even allow that such propositions are sometimes used as evidence for philosophical conclusions. But they will not think that the method of cases crucially relies on the use of such psychological propositions.
E2* tells us that intuited propositions should be used as evidence in philosophy, but it adds the qualification that the intuited propositions need to be known to be true. The bracketed clause is also added to reflect the assumption that the standard way we come to know such propositions is by having the appropriate rational insight. (Note a proponent of KI can allow that we can sometimes gain such knowledge by other means, e.g. by testimony.)
Importantly, the proponent of KI will not merely claim that the content of the Gettier intuition is our evidence that the JTB theory is false, as they will also claim that there is a substantive connection between one's having the Gettier intuition/insight and the fact that its content is now a part of one's evidence. This connection falls out of three assumptions: (i) that E = K; (ii) that having a rational insight is a way of knowing; and (iii) that one's Gettier intuition is a genuine rational insight as opposed to a mere rational intuition.
If the proponent of KI adopts Williamson's account of ways of knowing, they will think that the rational insight that the Gettier subject has NKJTB is a determinate of the determinable that is knowing that the Gettier subject has NKJTB, that is, it is a specific way of knowing this truth. On the other hand, if they adopt Cassam's account they will think that the rational insight is a state the possession of which can explain one's knowledge that the Gettier subject has NKJTB, but they need not also think that it is itself a state of knowledge or a state that entails knowledge.
KI and the nature of intuitions
In a similar way to E1 and E2, the proponent of KI can endorse E3 and E4 when they are interpreted like so:
(E3*) The intuited propositions that are assigned this important evidential role in philosophy are the objects of intuitions which are importantly distinct from any mere belief, judgement or inclination to believe or judge.
(E4*) The intuited propositions that are assigned this important evidential role in philosophy are the objects of rational intuitions.
In endorsing E3* and E4*, and in claiming that having a rational insight is a distinctive way of knowing, the proponent of KI clearly distinguishes their position from Williamson's deflationary view of the nature and role of intuitions in philosophy. On this deflationary view, E3* and E4* would be rejected because Williamson claims that what philosophers call ‘intuitions’ are simply beliefs or inclinations to beliefs that are not distinct in any interesting way from mere beliefs or inclinations to believe in general.Footnote 25 KI does not collapse into this deflationary view because it incorporates the central claims about the nature of intuitions made by prominent proponents of the intuition picture, including E3 and E4 as well as the etiological and presentational theses.
The partiality and revenge problems
KI illustrates one way that a proponent of the intuition picture could meet the challenge of accommodating E = K by appealing to the notion of a rational insight and Williamson's account of ways of knowing. In so doing, KI shows us that the partiality problem is only an apparent problem by illustrating the flexibility of the intuition picture with respect to different conceptions of the nature of evidence. Furthermore, a proponent of KI can endorse Williamson's response to the judgement sceptic and hence avoid the revenge problem.Footnote 26 This is because, given their endorsement of E = K, a proponent of KI will agree with Williamson that it is only in the good case (where the content of one's intuition is true) that one's evidence can include the non-psychological proposition that is the content of one's intuitions.
It is worth emphasising that appealing to the notion of a rational insight is not an ad hoc or ‘whatever-it-takes’ response to the assumption that E = K. Consider an analogy. Stephanie is a philosopher of mathematics who has developed a novel theory according to which perception is (correctly) assigned an evidential role in certain kinds of mathematical reasoning that were previously thought to be purely a priori. Being antecedently committed to a non-factive conception of evidence, when Stephanie develops her view she appeals to a non-factive notion of perceptual appearances. But then Stephanie attends an epistemology conference at Oxford and comes away convinced that E = K. Should Stephanie now abandon her theory? It is easy to imagine the case turning out in such a way that the answer is no. After all, the subject of Stephanie's theory is the epistemology of mathematics, not the nature of evidence in general. It would not be surprising then to find out that her core ideas about the evidential role of perception in mathematics could be made consistent with different theories about the nature of evidence, including E = K. And in meeting this challenge, Stephanie will presumably have reason to appeal to a factive notion of perceiving that p, rather than just the non-factive notion of it perceptually appearing to one that p.
Similarly, most proponents of the intuition picture appear to presuppose a non-factive view of evidence. In which case, it is not surprising that they appeal to a non-factive notion of rational intuitions or intellectual seemings when developing their views. If we assume that E = K such philosophers do face a prima facie challenge of showing how the core of their views can be restated so as to be made consistent with this assumption. But given that the core of their view concerns the evidential role of intuitions in philosophy, and not the nature of evidence in general, we should not be surprised if they can meet this challenge. And, in addressing this challenge, it would be natural for them to appeal to a factive notion of having a rational insight that p. It would also be natural to appeal to the notion of a way of knowing. Recall that, on the most obvious way of developing a non-psychologistic model of the intuition picture, simply having the Gettier intuition is itself a way of possessing the proposition that the Gettier subject has NKJTB as evidence. But note, if E = K, knowing is effectively the way of possessing evidence. In which case, a proponent of the non-psychologistic model who accepts E = K will quite naturally think that having the Gettier intuition/insight is a way of possessing the proposition that the Gettier subject has NKJTB as evidence, because it is a way of knowing that proposition to be true.
Conclusion
The intuition picture is consistent with the non-psychologistic view of philosophical evidence advocated by Williamson, and even with his view that E = K. This shows us that the intuition picture is not wedded to any particular positions on the nature of the evidence provided by thought experiments, or the nature of evidence in general. One should not expect then to be able to reject the intuition picture by appealing to different theoretical positions on such matters. Furthermore, this shows us that proponents of the intuition picture are not wedded to any particular response to sceptical challenges to the armchair judgements relied upon in philosophy. In which case, if they choose to, proponents of the intuition picture can avail themselves of Williamson's response to such challenges, without adopting his deflationary view of intuitions.Footnote 27
