Como citar:
Cómo citar: Espejo-Serna, J.C. (2019). Against cognitive homelessness. Humanitas Hodie, 2(1), xx-xx. https://doi.org/ 10.28970/hh.2019.1.a 4

Introduction

A cognitive home is not unlike a regular one: a place where you feel comfortable, a familiar or usual environment where you know your way around. Saying that we have cognitive homes is saying that there are certain aspects of our minds with properties similar to those our regular homes have in terms of our familiarity and closeness to them. Simply put, the idea is that in one’s cognitive home, if one is in a cognitive state one knows one is in such state (I’ll provide a fuller definition later on). On the other hand, claiming that we are cognitive homeless seems to mean that there is no such relation with our own minds; i.e., our mind is not a special place where things are open to plain sight in this familiar way. Without further clarification, the claim that we are cognitive homeless seems to run against our lived experience. It seems to me that if I have a pain, I know that I have a pain; if I feel cold, I know I feel cold; if I believe that the earth is flat, I know that I believe that the earth is flat, etc. The claim that we are cognitive homeless goes against this. Williamson (1996, 2000) defends the claim that in fact we are cognitive homeless, and for most aspects of our cognitive life it is not the case that if we are in the mental state S we know that we are in the mental state S. He characterizes cognitive homes in terms of what he calls luminosity.

Luminosity: a condition is luminous if the condition obtains and one is in a position to know that it obtains in such case.

Williamson argues that given the characterization of luminosity offered, an argument showing that we can generally claim that mental states do not have luminous conditions can be constructed. Therefore, we are cognitive homeless.

Williamson’s argument shows that luminosity is not something to be taken for granted. I agree. But I do not share Williamson’s conclusion that we have no cognitive homes. How I see it, our cognitive homes are like our real homes in much more than just the familiarity. In a tidy, well-kept regular house, things can be easily found. The owner might not know every single fact about how things are in her house, but she knows how to find out. She might not know if there is any milk left, but she can put herself in the position to know through some procedure (usually one that requires hard work). Things can get misplaced in our house but they are never completely lost; it is possible to set some order and find them. In a similar manner, in our cognitive home things can also be misplaced or hidden in plain sight (in line with Williamson’s diagnosis) but it is possible to relate to them in such a way that they are made luminous.

The luminosity of our mental states is not a given but an achieved property. So, the brute fact that a mental state occurs is not enough for obtaining luminosity.

Luminosity crucially depends on the guise of the mental state, that is, on the ways different conditions are presented and not on the brute fact that it obtains. This means that there is a non-trivial sense in which mental states can be luminous and can provide us with a cozy cognitive home.

In order to show this, I will first present argument and the key aspects of Williamson’s view. I think the argument has strong points, so I will then consider some objections raised in the literature to Williamson’s view and attempt to show their failure to address the real problem. In the end, I will try to explain why I take it to be the case that, although Williamson is right in his critique of the view according to which the contents of the mind are luminous, it depends on a general view of the mind one need not accept. I will suggest how there are other ways to see the mind and non-trivial ways of talking about the luminosity of its contents.

Williamson’s argument against luminosity

Williamson’s argument against luminosity takes the form of a proof by contradiction. To accept the luminosity of any given condition yields a contradiction, and therefore luminosity should be denied. Let us consider the argument.

First, take the condition of feeling cold. Prima facie, if one feels cold then one knows that one feels cold. What could count as evidence in favor of my knowledge that I feel cold if my feeling cold is not evidence enough? It seems very plausibly that if I feel cold, I am in a position to know that I am cold. Thus, feeling cold looks like a good example of a luminous condition.

Now, consider a morning in which you wake up feeling cold and you warm up during the day. At 7 am, you clearly feel cold. At around 10 you start feeling less cold. By noon, you are not sure whether you are cold or not, and roughly around 3 pm you start feeling warm. Take a series of times t0 ... tn at a millisecond interval between 7 am and 3 pm. Suppose that at ti (0-i-n) you feel cold, and since it is a luminous condition, you know that you feel cold.

The main premise of the argument is that if at time ti one knows (or is in the position to know) that p then at ti +1 at least it must be the case that p obtains. Let us call this the reliability condition: if at time t one knows that p, at a time t’ very close to t it is true that p. According to this premise, if one knows that one feels cold at ti , at ti+1 one feels cold. In turn, if one feels cold at ti+1 then one knows that one feels cold at ti+1, and then one feels cold at ti+2. This way of reasoning can be repeated x times to show that at ti+x (where i + x=n) one feels cold, which according to the initial characterization of the case is false —because at 3 pm you no longer feel cold. We have, then, that accepting luminosity and the reliability condition yields a contradiction.

Astraightforward way of resisting the conclusion is to attack what we have called the reliability condition on knowledge (Brueckner & Fiocco, 2002; Steup, 2009). One line of argument (Steup, 2009) attempts to show that there is no problem in saying that at time ti there is knowledge and that there is no knowledge at ti+1 by presenting a simple and intuitive case in which in just a millisecond there is a change that yields a change in knowledge without implying anything about knowledge at a time before:

Suppose you are in a pitch-dark room. Someone turns on the light. It seems plausible to me that here we have [Know](dark) at ti and ∼[Believe](dark) at ti+1 where ti 1 I will drop the qualification from now on but bear in mind that it should always be present. And ti+1 are only one millisecond apart. Or suppose it is pleasantly quiet now but then a screeching smoke alarm goes off. Again, I do not see why we should not think that you can pass within one millisecond from K(quiet) to ∼B(quiet). Likewise, if we assume maximum warm-up speed and make the contrast between feeling cold and feeling less cold maximally extreme, why should we not assume that a one-mil-lisecond change of feeling less cold than before is accompanied by awareness? (Steup, 2009, p. 229)

This kind of reply attempts to counter Williamson’s argument by showing that there is no problem with a change from knowing at one time and not knowing at a very short time later. The example presents a very intuitive case in which in just a millisecond there can be a big change that thus leads to a change in knowledge.

But, as I will suggest in the following section, we can do without the identification of conditions through time if we manage to present the conditions in a way in which the changes in the subject and the world are controlled in another manner. Time is not important as such but only in so far subject and world change regarding to time: changes in time that are not accompanied with changes in subject or world do not produce significant changes in the condition. Williamson’s argument does not depend on a mere small change of time (one millisecond) but how much the other aspects (subject and world) have changed. Identifying a series of times at a very small interval is not important in itself but in so far as it also accompanied with very small changes in the subject and the world, thus very small changes in the condition. Selecting cases only one millisecond apart is one way of getting very similar cases because there is not much change in either subject or world in that interval.

This understanding of Williamson’s argument according to which small changes in time are only relevant as they help specify small changes in the condition fits the characterization of reliability provided later in Knowledge and its Limits:

Reliability resembles safety, stability, and robustness. These terms can all be understood in several ways, of course. For present purposes, we are interested in a notion of reliability on which, in given circumstances, something happens reliably if and only if it is not in danger of not happening. That is, it happens reliably in a case α if and only if it happens (reliably or not) in every case similar enough to α. In particular, one avoids false belief reliably in α if and only if one avoids false belief in every case similar enough to α. (Williamson, 2000, p. 134)

Accordingly, we can reframe the reliability condition on knowledge without (direct) reference to time: if knowledge of p is reliable in a certain case, it must be the case that in similar cases p also obtains. This presentation of the reliability condition on knowledge shows why the alleged counter-example fails to grasp the role played by the appeal to reliability: a case in which there is darkness and a case in which there is light simply are not similar cases—they are in fact opposite cases—regardless of their closeness in time. The case does not show genuinely similar conditions in which there can be knowledge in one but not in the other. To generate a genuine counterexample to Williamson, there should be similar cases where the reliability condition is false. Another line of argument (Brueckner & Fiocco, 2002) against Williamson attempts just this.

The idea is that there can be two very similar cases in which we would deny that there is knowledge at ti+1 but accept that there is knowledge at ti , thus providing a counter example to the reliability condition. Consider the following:

Suppose that at time t0 Joan is staring at a dead parrot on the floor and continues to do so for five hours. In that time interval it seems that we can say that Joan knows that there is a dead parrot on the floor. At tn after five hours Joan blinks. One millisecond after Joan has blinked, she opens her eyes and sees a dead-parrot hologram. At any tw (0 ≤ w ≤ n), Joan can have knowledge that there is a dead parrot. Yet at tn+1 Joan cannot have knowledge that there is a dead parrot on the floor simply because there is no dead parrot on the floor, there is just a hologram.

This example seems to meet the demand of examining the relevant similar cases. Though, as we will see, it depends on how exactly we individuated the cases. And it also seems acceptable that Joan does have knowledge that p at ti , for the subject is in fact seeing a dead parrot and there is a dead parrot. But since in case ti+1 the subject merely has a visual experience of a dead parrot but there is only a dead parrot hologram, there is no such knowledge. These two premises make it a stronger stance against Williamson.

We can again respond on Williamson’s behalf by showing that the conditions are insufficient. The critique relies on the seemingly simple continuous condition that the subject is seeing a dead parrot from t0 to tn+1.

One possible way out is to deny that the example provides two very similar cases by claiming that in between t0 to tn there is veridical perception while at tn+1 there is no veridical perception. Hence there is a substantive difference. This amounts to claiming that there is nothing in common between veridical and non-veridical cases of perception, i.e., some kind of disjunctive theory of perception. (Although for the present matters it is enough to show them not to be similar cases, that is, merely a non-conjunctive account. See Soteriou [2016] for different forms of disjunctive theories of perception). There is a relevant difference between the cases, namely, that before tn+1 there is a factive propositional attitude while afterwards there is a non-factive propositional attitude (if there is an attitude at all). If one were to accept that we can only see that p when p is true, the counter example would fail.

Williamson’s critics might resist such account of the nature of seeing and dig their heels in. The issue would then hang on what we decide about the nature of visual perception, that is, whether a hallucination is similar to a veridical perception. Likewise, Williamson might claim that there is no such similarity. This line of argument seems to land us on a stalemate.

But the line of argument above lets us appreciate an advantage of Williamson’s original way of characterizing similarities. The original example uses closeness in time as a way of stating similarity of cases without espousing any particular way of individuating differences and similarities. In the case of feeling warm, we can get a good enough sense of what it is for two cases, close by in time, to be relevantly similar. Likewise, for any condition that gradually changes across time, we can make sense of the idea that those moments close in time are, though not the same, sufficiently similar. Unless this line of argument criticizes such notion of similarity or defends a non-disjunctive view perception, the point against Williamson has little traction.

Other replies to Williamson take less direct paths: a possible answer to the argument against luminosity by presenting an aspect of the mind for which it cannot be instantiated (Tennant, 2009). Such opponent claims that semantic qualia (something that it is like, for example, understanding a sentence as expressing a thought) guarantee that when one knows that an expression has meaning, one knows that one knows. Others (like Weatherson, 2004) criticize Williamson’s account of knowledge by claiming that it is because of it that problematic claims like the reliability condition seem to hold. They argue that Williamson’s account of knowledge fails to capture the appropriate role of safety in knowledge.

The success of these replies depends heavily on additional theories—and not just simple counterexamples. It depends on either a detailed examination of the problems in Williamson’s complete account of knowledge as presented in Knowledge and its Limits or a substantive defense of the existence of highly problematic entities like semantic qualia. Again, on their own, these lines of argument have little traction.

In what follows I attempt to issue a different kind of criticism. For this, we need to look back at Williamson’s understanding of luminosity.

The idea of luminosity

I want to consider some noteworthy points about the way luminosity has been characterized. First, luminosity has been presented fundamentally as a property of a condition. The reference to a subject, a time, and a world play a role but only as a way to identify the condition, i.e., as ways to determine the identity of a condition in a given context. Williamson talks about a context in which a condition obtains, without pointing out the elements that conform the context aside from the subject, the world, and the time. I have avoided this way of presenting the issue because in the view I am trying to set forward there are fundamental and different roles played by the elements of this context, especially of the particular subject, as I will try to show ahead.

The determination of the condition’s subject is crucial because for some subjects a determinate condition might obtain while for others will not; thus, reference to a subject is needed. That there is pain in a leg might be true of a subject at a time and in a possible world while in another not true for another subject. Likewise, reference to a particular world is needed because a condition might obtain for a subject in some possible worlds but not in others. That there is pain in a leg might be true of a subject at a time and a possible world but not in another possible world.

But time, unlike subject and world, does not seem to play a role in determining if a condition obtains. Time seems to be needed in order to specify a context because a change in time might imply a change in subject or world. A change in the subject is enough to change the context and likewise for a change in the world: fixing the other elements that conform the context while changing either subject or world will be enough to introduce a change that can alter the condition. But a change in time that is not accompanied with a change in either subject or world does not change the context significantly. If we imagine that there are no changes in the world, nor in the subject, everything stood still while time continued going on, there is nothing that could change in the condition. (Unless, of course, it is a condition defined in terms of times. For sake of simplicity, I will simply choose conditions that are not thus time dependent.) Determining a context by reference to a time helps to exclude any change in either world or subject. Thus, it seems that reference to time is not as important as reference to a world and a subject to specify the context of a condition.

This means that if we manage to find a way to determine a particular subject and world such that it leaves no open space for major change, time could be dispensed with. So, when looking at Williamson’s argument we are to pay special attention to the role each of these elements play and examine if other specifications of luminosity could still guarantee correct identification. (This is especially important in light of the objections presented above.)

Another important thing to have in mind about luminosity as characterized is that someone (and perhaps the luminosity advocates quoted by Williamson are among them) might consider luminosity not as a property that a condition could have for any subject that happens to be in the condition—and this would not be a matter of their having some psychological pathologies or physical impediments.

Luminosity proponents might claim that a condition is luminous when it obtains for a subject at a time and in a world, and a given subject (and not just anyone) is in a position to know that the condition obtains.

For example: subject S at time T in world W is thinking about Venus. This condition is luminous if and only if S knows that he is thinking about Venus. Now this condition obtains if a subject is thinking about the Morning Star, but it could only be luminous if the subject knows that the Morning Star is Venus. A subject that does not know that the Morning Star is Venus would still be thinking about Venus when she is thinking about the Morning Star but the condition of thinking about Venus would not be luminous. In this case, additional knowledge of the subject makes the difference, which suggests that cognitive particularities of the subject might play a role in the luminosity of mental states.

The additional reference the general mental state of the subject do not seem well accounted for when the conditions are individualized in terms of a context that does not consider the particular subject involved and their particular cognitive abilities. Mentioning a subject and a world does not need to characterize this difference. Williamson very briefly considers a similar case of different presentations of the same condition: the case of a condition of drinking water and the condition of drinking H2O. In his words:

If the condition that one is drinking water is the condition that one is drinking H2O, because they obtain in the same cases, it does not seem to follow that one knows that the condition that one is drinking water obtains if and only if one knows that the condition that one is drinking H2O obtains, for one may not know that water is H2O. Fortunately, in a context in which the only relevant presentation of the condition C is as the condition that one is F, knowing that C obtains can be identified with knowing that the condition that one is F obtains, which is in turn only trivially different from knowing that one is F. We can therefore often leave the reference to guises tacit. (Williamson, 2000, pp. 94-95)

Williamson simply states that there is one and only one guise in which the condition is going to be taken along the process (see also Williamson, 2000, p. 108), ignoring any reference to different ways the condition could present to the subject. But precisely what my earlier example suggests is that at least in some cases it is important to consider how the condition is presented to the subject in order to determine if it is luminous or not. The point is that, with respect to our cognitive home, guise matters.

The example might be far from conclusively presenting a point against Williamson by itself, but at this point my interest with it is to show that there might be some other ways of characterizing luminosity as a property of conditions in which restriction about the kind of subject and the kind of condition involved make a difference, and the example provides initial support for the alternative. Accordingly, we should pay attention to these matters when examining the argument against luminosity. In the following I explain how.

Three examples of luminosity

Descartes, Wittgenstein and Dummett are presented by Williamson as endorsing some version of luminosity. Descartes is taken to hold that the condition of thinking is luminous because just by thinking one can know that one thinks. Wittgenstein is taken to claim that all of what is in interest of philosophy is luminous. And Dummett is taken to say that the meaning of words is luminous because if, for a speaker, two words have the same meaning, the speaker knows that they have the same meaning.

All of them are presented as showing that certain conditions, just by the mere fact of being the case, entail knowledge that the condition obtains. Although this is in a sense correct, it is very misleading because it disregards how those claims are obtained. It is central to their position that the conditions regarded are luminous because they are presented in a certain way. To consider the conclusion of a condition’s luminosity independently of the philosophical development that leads to it in each case fails to adequately capture the heart of the kind of philosophical inquiry of the philosophers in question.

Williamson’s remarks against Descartes and Wittgenstein are rather short. Against Dummett, Williamson is a little bit more explicit. In what follows, I will comment the treatment they receive from Williamson as a way of showing how luminosity, as understood by these three philosophers, is not simply a matter of the brute fact that a condition obtains but rather a matter of an achievement.

Take Descartes’s Meditations (2008 [1641]). The First meditation guides the reader through a path of different kinds of doubts that ultimately leads to the wellknown I think, I am. The claim that thinking is luminous is put forward as the result of a method, i.e., the method of doubt explained in the Meditations.³ We get to see that thinking is luminous because through the exercise of the methodical doubt we gain a further understanding of how uncertain our general knowledge is. The meditator comes to understand that the method of doubting can prove to be a very powerful tool. And this understanding that strengthens doubt also strengthens thought because we come to realize that doubting is just a way of thinking. When the meditator comes to realize that doubting is thinking, they are but one step short of seeing the luminosity of their own thinking. What could it mean that thought is not luminous if the tool used to attempt to undermine it turns out to be just one expression of it? It looks as if nothing fits the bill because thought, thanks to the meditators exercise, has been placed under the spotlight and is showered in light. If the brute fact that one thinks were enough to guarantee that one knows that one thinks, the whole plan of Descartes would be idle. Why not simply start off the Meditations with the purportedly luminous condition of thinking? If we are going to take Descartes at his word and accept that he is genuinely proposing a method, we should take into account the path taken as well as his conclusions: method matters. Showing that the condition of thinking can be known does most of the work in making the condition luminous because an advance in the understanding yields light into the condition. For Descartes, thinking becomes luminous as it withstood the meditators doubts. Luminosity is not simply a given.

The case with Wittgenstein is very similar. When Wittgenstein says that what is of the interest in philosophy lies open to sight, he is claiming that philosophical problems are not ‘deep’. The activity of philosophy is seen by Wittgenstein as a child’s constant questioning why, in that it is not a request for more information but an expression of puzzlement. To answer those questions we need to see things in a different manner. The work of the philosopher is to dissolve, rather than solve, problems; not through scientific discovery of new information but through re-description, “finding and inventing intermediate links” (PI §122) between concepts: “Philosophy just puts everything before us, and neither explains nor deduces anything” (PI §122). What is of interest to philosophy, according to Wittgenstein, is made luminous by the very philosophical method. Luminosity is not something given, otherwise there would not be any philosophical problems. Rather, the kind of luminosity that Wittgenstein is looking for is the kind one has to work for. The philosopher needs help finding his way out of the problem an thus the goal of philosophy is to go from a puzzling non-luminous situation to one in which the answer is evident: it is a process of illuminating its subject of study. Luminosity is not simply a given.

Contrary to the very short references Descartes and Wittgenstein receive (they are only mentioned in passing as examples of luminosity), Williamson discusses Dummett’s position more thoroughly. With Dummett there is a clear statement of what is going to be considered as luminous, namely, the identity and difference of meaning.

For Dummett, knowledge of meaning is knowledge of reference, but is more than just bare knowledge of the reference (Dummett, 1978, p. 130). What matters is the way of knowing the reference. This knowledge has special luminous conditions because of the way it is acquired (Dummett, 1993, pp. 94-95). Just as in the examples of Descartes and Wittgenstein, what matters in meaning is the mode of presentation of the condition: the kind of knowledge. Perhaps assertability conditions fail to capture the kind of knowledge of the reference of an expression that speakers have, but that would be because they fail to capture the kind of knowledge that speakers have. Thus, we can reasonably say that for Dummett too luminosity depends on particularities of the condition, that is, on the special kind of knowledge it is and the way of being obtained. Not all knowledge of reference is the same; the kind of knowledge of the condition we might have will depend on the very way in which the condition is presented. Luminosity is not simply a given.

I do not expect to have established the correct understanding of the philosophical stance that Descartes, Wittgenstein, and Dummett have towards the issues they come to consider luminous. All that I hope this brief discussion has managed to convey is that for each of them there is an issue about the way luminosity is understood. A luminous condition is not to be taken as something that can be fully identified just by quoting the subject, the world and the time, as Williamson attempts. At least, it might involve their mode of presentation, the way they are described and their relations with other conditions; it might involve the way in which we come to know them, i.e., if by doubt or conceptual clarification. It depends.

In claiming that we are cognitive homeless, Williamson is only dealing with mental events as conditions and the particular situation in which they obtain. But it seems that the advocates of luminosity seem to agree that the brute fact that a condition obtains is not enough to its produce luminosity. This seems to point that Williamson is fighting against a view of luminosity that even his own quoted examples do not hold.

Conclusion

Take a very rough picture of the mind under which we think the contents of the mind as pixels in a digital photograph. The pixels are all in the image, open in plain sight, and everything in the image is a pixel. When looking at the image for one millisecond, we might not have knowledge of all the pixels because we did not pay attention to each and every single one of them. But were we paying attention to a particular pixel, there would not have been anything stopping us from having knowledge about it (provided we zoom in to the appropriate distance, have good sight, are well rested, and other usual physiological conditions). The pixels are completely accessible to us: there is nothing preventing us from being in the correct position to see them, and there is nothing more to the pixels that what can be seen in plain sight.

Analogously, one might think the contents of the mind are perfectly visible in plain sight: it takes no particular effort to access any of them and once we access them, we get complete access (there is nothing more to access that what can be accessed at one time). In this picture the contents of our mind are luminous because for every case, if in that case the subject is in a mental state S, the subject is in a position to know that she is in S. More importantly: the whole nature of the mental state can be simply grasped; there is nothing to the mental states that cannot be captured in a single intake.

This is the picture of the mind that Williamson seems to be arguing against when he claims that mental states are not luminous. He refuses to think of the mind as a place where everything is open in plain sight. But this image, as I shown in this section, does not fit adequately with what the advocates of luminosity seem to be talking about. Descartes, Wittgenstein, and Dummett, the three examples Williamson considers to be advocates of luminosity, do not seem to accept this image.

Now consider another rough picture of the mind in which its contents are like the objects in a room. In a room all objects are not equally visible at the same time. Some are very well lit and in the foreground. Others are partially occluded by bigger objects, and some other objects are completely covered (or perhaps light just happens not to get to them). This means that it is not possible to have knowledge of all the objects of the room at a single glance. Even if one was looking in the right direction, paying attention and without any physiological or psychological abnormalities, there can still be objects that one could not see completely. But there is nothing that in principle impedes us from seeing them.

In the case where an object is not visible (partially or completely) we can change its position, modify our point of view or otherwise arrange things so that it becomes visible. Objects in a room can also be of different kinds: some are bigger and brighter; some are small and blend in with the color of the carpet; some are like soap bubbles that only last for a time and then disappear never to be seen again; some are used daily while some others left almost untouched for months. The kind of element and how it is arranged make a difference in their “easiness” to be seen. If my keys (a persistent medium-small object of daily use) are in my room but I do not know that my keys are in the room, I can tidy up my room and improve the order so that when I leave my keys in the room, I know that they are in the room.

Likewise, the contents of the mind might be said to have different degrees of luminosity. Not all the contents of my mind can be seen at one time. Even if I am currently paying attention to a mental condition, the fact that it obtains might still elude me. But that does not mean that I cannot change the way the condition is presented to me and engage with the contents of my mind in a way that throws light to said condition. The complete nature of the contents of my mind might not be available to me at any given moment but I can modify the way I relate to them so that other aspects become visible.

A very crude example: think about a half-empty glass and then think about a half-full glass. The glass has the same amount of water but the way it is being presented changed. Another example: one can hope that p without realizing that one is hoping that p. Through reflection upon my current beliefs, my previous actions, my desires and other mental states, I change the way that things are being presented to me and find out something about my mind: I come to notice that I hope that p. I can get to see the situation in a different manner, so that it is now evident for me that I am at this state of mind. A final example: think about the person who is reading this text and think about yourself. Both of these ways may be expressing a thought about the same person. But one way of thinking elicits connections that the other one does not: thinking that the reader of the text is hungry does not move you, by itself, to take some action but thinking that you are hungry can, ceteris paribus, move you to go for a snack (cf. Perry, 1979).

The modes of engagement with one’s thought (reflection, meditation, philosophical understanding…) do not alter the condition, at least according to Williamson’s way of individualizing conditions in terms of subject, world, and time. (A condition, we have said, is individuated by the situation in which it obtains, namely, the state of a subject, a world, and a time.) In the examples of different ways of achieving luminosity presented above, the subject and the world do not change. The only thing that changes is the time, since the reflection effectively has a span. This change in time would not be accompanied by a respective change in the subject or the world; and according to what we said above (in section 4), a mere change in time is not enough to produce a change in a condition. Yet, in another sense, something has changed in the condition: the way it is being presented but this is not an aspect that can be captured in terms of the subject, the world, and the time involved. Proof that it is an important change is that the way the condition is presented determines the degree of luminosity.

Luminosity seems to be a property that conditions can achieve through different means. So, there can be some conditions that are luminous for me because I have come to see them as such, thus building myself a place where I can feel at home. Cognitive homes, much like regular ones, are not given, but not unachievable.

Notas:

¹This is a version of a paper I wrote in 2010 but never quite finished for publication. If I were to write it today, the paper would have a different tone and scope. I am nevertheless fond of it because this paper was part of the application process to get me into the Ph.D. programme at The University of Warwick. Many thanks to Bill Brewer for comments on the first version. I would also like to thank the two anonymous reviewers and the copyeditor at Humanistas Hodie who helped me give this paper its current form.

²Ph.D. Philosophy. Assistant Professor, Universidad de la Sabana. ORCID 0000-0002-1717-6630. E-mail: Juan.espejo1@unisabana.edu.co

³The importance of the path and method followed by Descartes is skillfully presented by Frankfurt (2007). The following understanding of Descartes’s involvement in claims about luminosity is based on Frankfurt’s reading of the Cartesian project.

Bibliography

Brueckner, A. & Fiocco, M. (2002). Williamson’s anti-luminosity argument. Philosophical Studies, 110, 285-293.

Descartes, R. (2008 [1641]). Meditations on First Philosophy: With Selections From the Objections and Replies. Oxford University Press UK.

Dummett, M. (1978). Truth and Other Enigmas. Harvard University Press.

Dummett, M. (1993). The Seas of Language. Oxford University Press.

Frankfurt, H. (2007). Demons, Dreamers, and Madmen: The Defense of Reason in Descartes’s “Meditations”. Princeton University Press.

Perry, J. (1979). The problem of the essential indexical. Noûs, 13(1), 3-21. Soteriou, M. (2016). Disjunctivism. Routledge.

Steup, M. (2009). Are mental states luminous? In P. Greenough and D. Pritchard (Eds.), Williamson on Knowledge (pp. 217-236). Oxford University Press.

Tennant, N. (2009). Cognitive phenomenology, semantic qualia, and luminous knowledge. In P. Greenough and D. Pritchard (Eds.), Williamson on Knowledge (pp. 237-256). Oxford University Press.

Weatherson, B. (2004). Luminous margins. Australasian Journal of Philosophy, 83, 373-383.

Williamson, T. (1996). Cognitive homelessness. The Journal of Philosophy, 93(11), 554-573.

Williamson, T. (2000). Knowledge and its Limits. Oxford University Press.

Wittgenstein, L. (2009). Philosophical Investigations (4th ed.) Hacker and Schulte (Trans.). Wiley-Blackwell.