Descartes is widely portrayed as the arch proponent of “the epistemological transparency of thought” (or simply, “Transparency”). The most promising version of this view—Transparency-through-Introspection—says that introspecting (i.e., inwardly attending to) a thought guarantees certain knowledge of that thought. But Descartes rejects this view and provides numerous counterexamples to it. I argue that, instead, Descartes’s theory of self-knowledge is just an application of his general theory of knowledge. According to his general theory, certain knowledge is acquired only through clear and distinct intellection. Thus, in his view, certain knowledge of one’s thoughts is acquired only through clear and distinct intellection of one’s thoughts. Introspection is a form of intellection and it can be clear and distinct. Ordinarily, however, introspection isn’t clear and distinct but is instead confused with dubitable perceptions of bodies. To make introspection clear and distinct, we need to “sharply separate” it from all perceptions of bodies by doubting all perceptions of bodies. Without such radical doubt, introspection remains confused and we lack certain knowledge not just of the specific features of our thoughts, but even of the minimal claim that a thought exists. Far from being the high priest of Transparency, Descartes is radically opposed to it.

KEYWORDS: cogito, consciousness, Descartes, doubt, introspection, reflection, self-knowledge, transparency

Before Descartes, there had been Skeptics, but they were merely Skeptics. Descartes taught his age the art of making Skepticism give birth to philosophical Certainty. (Para du Phanjas 1779: Vol. 1, xx*)

Read almost any article, textbook chapter, or reference entry on self-knowledge and you will learn that Descartes is the arch proponent, and indeed the progenitor, of a doctrine Margaret Wilson has dubbed “the epistemological transparency of thought or mind” (1978: 50)—or simply, “Transparency.” Transparency is associated with the claim that we each have “privileged access” or “first person authority” regarding our own thoughts, and there are different versions of it. The simplest version—Transparency-through-Having—says that merely having a thought is sufficient for knowing with certainty that you have that thought. According to this view, you don’t need to do anything to earn this knowledge; in fact, there is nothing you can do to avoid it. You are inevitably omniscient about everything happening within your own mind all of the time.

Many readers have taken Descartes to espouse Transparency-through-Having because he asserts that having a thought is sufficient for being conscious of that thought. The assumption behind this reading is that consciousness is sufficient for certain knowledge. But Descartes shows otherwise with counterexamples: cases where people are ignorant of their beliefs, emotions, and other thoughts, even while they are, in his view, necessarily conscious of them.

In light of these counterexamples, some commentators acknowledge that Descartes rejects Transparency-through-Having. They usually read Descartes as holding what I regard as a variant of Transparency—Transparency-through-Introspection—which says that introspecting (i.e., inwardly attending to) your thought guarantees certain knowledge of that thought. But Descartes provides counterexamples to this view, too: cases where people are ignorant of their thoughts, even while introspecting them.

In this essay, I aim to advance the discussion of Descartes’s supposed commitment to Transparency in four ways. Even among scholars who acknowledge the aforementioned counterexamples, there is still the lingering idea that Descartes must be committed to some restricted version of Transparency that applies only to some privileged subset of truths concerning one’s own mind. My first aim is to show otherwise: Descartes does not rely on Transparency for any self-knowledge, not for the generic claim that I am thinking, and not even for the minimal claim that a thought exists. He rejects Transparency completely.

Second, I identify the theory of self-knowledge Descartes actually holds. It is widely recognized that according to Descartes’s general theory of knowledge—namely, Rationalism—certain knowledge is acquired only through clear and distinct intellection. I argue that Descartes’s theory of self-knowledge is an application of his general theory: certain knowledge of one’s thoughts is acquired only through clear and distinct intellection of one’s thoughts.

Third, I use Descartes’s Rationalist theory of self-knowledge to explain why he rejects Transparency. He rejects Transparency-through-Having because having a thought does not entail having clear and distinct intellection of that thought. He does hold that having a thought entails being conscious of that thought. But consciousness does not guarantee certain knowledge, precisely insofar as consciousness is not a form a clear and distinct intellection. Transparency-through-Introspection is much closer to Descartes’s actual view because introspection is a form of intellection and it can be clear and distinct. Ordinarily, however, it isn’t. Introspection is ordinarily confused with dubitable perceptions of bodies. So, in order to acquire self-knowledge, we need to make introspection clear and distinct.

Fourth, I explain how this is supposed to be done. To make introspection clear and distinct, we need to “sharply separate” introspection from all perceptions of bodies—by doubting all perceptions of bodies (i.e., through radical doubt). Transparency entails that doubt is not required in this way. What all versions of Transparency have in common—the essence of Transparency, if you will—is the notion that self-knowledge is easy. I mean “easy” in the sense that, if Transparency were correct, you could acquire self-knowledge without employing any special method, much less Descartes’s method of radical doubt. Transparency-through-Having entails that self-knowledge is not just easy but inevitable: you cannot have a thought without knowing with certainty that you have it. Transparency-through-Introspection does not make self-knowledge inevitable, but it still makes it easy. You don’t need to employ a special method to acquire self-knowledge. You certainly don’t need Descartes’s method of radical doubt. (How could that help, anyway?) All you have to do is look within. This alleged easiness of self-knowledge is exactly what Descartes denies. Self-knowledge, on his view, is a hard-won achievement. A special method is required to “prepare the mind” to perceive itself correctly—that is, clearly and distinctly—and that method is precisely the method of radical doubt. Doubt about the external world is not merely contrasted with certainty about the self—it’s the only way of gaining such certainty in the first place.

Here is the plan. In Section 1, I clarify what we need to account for as we interpret Descartes’s theory of self-knowledge. In Sections 2 and 3, I survey Descartes’s counterexamples to Transparency-through-Having and Transparency-through-Introspection. In Section 4, I argue that, instead of Transparency, his actual theory of self-knowledge is just an application – and indeed his paradigm illustration – of his general theory of knowledge: Rationalism. In Section 5, I show that, in his view, introspection ordinarily doesn’t deliver any self-knowledge because it isn’t clear and distinct, and that it becomes clear and distinct only through radical doubt. In Section 6, I close by noting how Descartes’s real view of self-knowledge might pique our interest today.

1. Cartesian Self-Knowledge

Let’s clarify our terms. By “knowledge” I do not mean ordinary knowledge, but a grade of knowledge (cognitio) that Descartes describes as “clear,” “evident,” “indubitable” or “certain”. Certain knowledge is characterized by “absolute,” or “metaphysical certainty,” and precludes even “slight or so to speak metaphysical” doubt. By “ignorance” I mean lack of certain knowledge.

By “self-knowledge” I mean certain knowledge of contingent truths concerning one’s mind, or one’s self qua mind. The primary examples of such truths are what I call (true) thought-claims, i.e., contingent truths concerning one’s thoughts. We can classify thought-claims into three grades of complexity, according to whether they are claims about:

  • the existence of a thought: A thought exists.
  • the (first-person) possession of a thought: I am thinking.
  • the specific features of a thought, like what kind of thought it is (e.g., I am doubting), or its content (e.g., I think that I am walking), or its character (e.g., I have a clear perception), or all of the above (e.g., I clearly conceive of a triangle).[2]

As far I’m aware, Descartes never asserts the first claim—that a thought exists—on its own; but it is, importantly, the minimal thought-claim, the one that is entailed by all other thought-claims. When a thought-claim is true, it is made true by at least one token thought existing in a particular mind. My current thought that it’s warm outside does not make it true that it’s warm outside, but it does make it true that I think it’s warm outside. Like every other token thought I have, it also makes it true that I am thinking. And like every other token thought, it makes it true that a thought exists.

In addition to true thought-claims, there is one other truth that can be the content of self-knowledge in the relevant sense, namely the contingent truth affirming one’s existence: I exist (or I am).[3]

In Descartes’s view, certain knowledge of one’s existence is inferred from certain knowledge of one’s thoughts, as in his famous cogito argument: “I am thinking, therefore I exist” (cogito, ergo sum). He often substitutes the generic premise, I am thinking, with specific thought-claims (in the first-person present)—I am doubting; I think I am walking; I think I am breathing; etc.—to infer the conclusion, I exist.[4] So, in asking about Descartes’s account of self-knowledge, our principal question is this: How, in Descartes’s view, does one acquire certain knowledge of thought-claims (i.e., contingent truths concerning one’s thoughts)?

It is significant that claims of self-knowledge, so defined, always affirm or entail the actual existence of a given thought and the actual existence of the mind (or a self qua mind) which has that thought. Since our thoughts and our minds exist only contingently, claims of self-knowledge can only be contingently true. Modal claims (concerning what is necessary or merely possible) are not claims of self-knowledge in the sense we are investigating here, not even when they mention thoughts, the mind, or the self. This is important because commentators sometimes conflate claims of self-knowledge with modal claims that mention the self. Let me explain. Descartes often prefaces the cogito—by which I mean the whole argument: I am thinking, therefore I am—with phrases like “this proposition,” “this truth,” and “this first principle,” treating it is as one proposition. It is, nevertheless, one complex proposition, comprising three smaller propositions. It counts as a claim of self-knowledge because two of its constituent propositions—the explicit premise and conclusion (I am thinking and I exist)are claims of self-knowledge. But it also comprises a third proposition, at least implicitly, which is not a claim of self-knowledge in the relevant sense:

Necessarily, if I am thinking, then I exist.

This proposition expresses the validity of the cogito argument, the fact that the premise entails or necessitates the conclusion.[5] Descartes indicates that it is easy to acquire certain knowledge of this truth; we only have to consider it to intuit it clearly and distinctly.[6] But importantly, it is merely conditional. You can have certain knowledge that if you are thinking, then you must exist, without having certain knowledge that you are thinking, or that you exist. In other words, you can have certain knowledge that the cogito argument is valid, without having certain knowledge that it is sound. Indeed, I will argue that, in Descartes’s view, we are prone to be in exactly this situation when we consider the cogito argument without making our introspection clear and distinct through radical doubt. Moreover, we are prone to mistake the mere conditional for the cogito argument as a whole, as evinced by the fact that commentators often refer to them interchangeably as “the cogito.” As a result, when we consider the cogito, we are apt to think we have self-knowledge when in fact all we know with certainty is the mere conditional.

Another modal truth which is not a claim of self-knowledge in the relevant sense is the fact that the proposition, I am thinking, is “incorrigible”: one cannot think it falsely. The incorrigibility of I am thinking can be expressed in terms of impossibility: It’s impossible for me to think ‘I am thinking,’ while I am not thinking. Or it can be expressed in terms of necessity:

Necessarily, if I think ‘I am thinking,’ then I am thinking.

The proposition, I exist, is also incorrigible: It’s impossible for me to think ‘I exist,’ while I don’t exist. In other words:

Necessarily, if I think ‘I exist,’ then I exist.

Again, these claims, expressing the incorrigibility of I am thinking and I exist, are merely conditionals. You can have certain knowledge that a proposition is incorrigible without having certain knowledge that it’s true.[7]

Finally, having self-knowledge, in the sense at issue here does not include the two claims—positive and negative—that Descartes makes about the nature (or essence) of the mind:

Positive: The mind is essentially thinking, which entails that necessarily, if x is a mind, then x is thinking. (Established after the cogito in Meditation Two.)[8]

Negative: The mind is not essentially embodied, which entails that it’s possible for x to be a mind while x is not embodied. (Established by the argument for the “real distinction” between mind and body in Meditation Six.)[9]

Neither of these claims about the nature of the mind entail the existence of a mind. And the meditator achieves self-knowledge through the cogito before going on to establish these two additional claims.[10]

2. Transparency-through-Having?

Many readers have assumed that Descartes’s account of self-knowledge is Transparency-through-Having: the view that whenever you have a thought, you have certain knowledge of that thought. This reading is based on Descartes’s definition of “thought” (cogitatio, pensée):

Thought. I use this term to include everything that is within us in such a way that we are immediately conscious [conscii] of it. (2R, 2:113/AT 7:160*)

By the term ‘thought’ I understand everything which we are conscious [consciis] of happening within us, in so far as we have consciousness [conscientia] of it. (Pr. i.9, 1:195/AT 8A:7*)

As we use the term today, “thinking” connotes reasoning, but Descartes’s usage is more expansive as it applies to every conscious operation of the mind, including sensations: “Hence, thinking is to be identified here not merely with understanding, willing, and imagining, but also with sensory awareness” (Pr. i.9, 1:195/AT 8A:7*). In accord with this definition, Descartes says repeatedly,

[… We] cannot have any thought of which we are not conscious [conscii] at the very moment when it is in us. [… We] are always actually conscious [conscios] of the acts or operations of our minds. (4R, 2:171–172/AT 7:246*; cf. 1R, 2:77/AT 7:107; 4R, 2:162/AT 7:232; 6R, 2:288/AT 7:427; M3, 2:33–34/AT 7:49; Passions, i.19, 1:335/AT 11:343)

These texts assert that consciousness is exhaustive in the sense that we are always conscious of everything happening in our minds, and not just potentially but “actually.” As the saying goes, consciousness, for Descartes, is “the mark of the mental.” A bit more formally:

If S has a thought, then S has consciousness of that thought.[11]

As a claim about consciousness, this is a psychological thesis. We might say it asserts a kind of psychological transparency of the mind. But our topic—Transparency with a capital “T”—is what Margaret Wilson rightly calls “the epistemological transparency of thought or mind” (1978: 50), which is a claim about certain knowledge. The version we are presently considering is

? Transparency-through-Having
If S has a thought, then S has certain knowledge of that thought.[12]

(The “?” on the left flags that, on my reading, Descartes rejects the view in question.) According to Transparency-through-Having, “the mind is naturally ‘given’ to itself” (Rorty 1980: 97). Indeed, if this view were correct, it would be impossible to be ignorant about anything in your mind. Complete self-knowledge would not just be given, but a gift you literally cannot refuse.

Descartes provides a litany of counterexamples to this view, cases where people lack certain knowledge of their thoughts. Here are just two, for starters:

  • Beliefs. In the Discourse, Descartes says, “In order to discover what opinions [people] really held I had to attend to what they did rather than what they said.” This is not only because people are reticent to share what they believe, but also because “many people do not know what they believe. Believing something and knowing that one believes it are different acts of thinking, and the one often occurs without the other” (1:122/AT 6:23). What a remarkably Freudian thing for Descartes to say. Not only can you be ignorant of your beliefs, but someone else might better surmise what you believe, by watching what you do.
  • Passions or emotions. In the Passions of the Soul, Descartes says that “those who are the most strongly agitated by their passions are not those who know them best.” What they lack, he adds, is “evident knowledge” (i.28, 1:339/AT 11:350), which is what is at issue here. 

In both of these cases, people have thoughts—beliefs or passions—while lacking certain knowledge of them. Descartes describes many other counterexamples to Transparency-through-Having, but I will postpone them until the next section, because they are also counterexamples to the more promising version of Transparency that we turn to next.

3. Transparency-through-Introspection?

Commentators are increasingly aware that Descartes rejects Transparency-through-Having.[13] But if having thoughts doesn’t ensure certain knowledge of thoughts, in Descartes’s view, then what does? The alternative that scholars usually offer is introspection or what Descartes calls “reflection,” the act of inwardly attending to, concentrating on, observing, or noticing one’s thoughts. According to these scholars, then, Descartes is committed to

? Transparency-through-Introspection
If S has a thought and introspects that thought, then S has certain knowledge of that thought.[14]

Note that you cannot introspect a thought unless you are the one who has that thought. This means that Transparency-through-Having entails Transparency-through-Introspection, and the arguments against the latter, which we’re about to see, also count against the former. Note also that Transparency-through-Introspection invokes ordinary introspection—not introspection purified through radical doubt.

To be fair to proponents of the Transparency-through-Introspection reading, let us consider the best case for ascribing that view to Descartes. To do so, we should emphasize that introspection, for Descartes, is different from the kind of consciousness that, as we saw in Section 1, does not provide certain knowledge. Introspection is a kind of consciousness—an inner awareness of one’s own mind—but Descartes distinguishes between two kinds (or orders or levels) of consciousness. What I will call “basic consciousness” is the kind of consciousness, introduced in Section 1, which exhaustively accompanies all of our thoughts. Basic consciousness comes along with every thought because it’s built into the very nature of thought, as one of its essential intrinsic features. Just as thoughts are representational (in Descartes’s lingo, they have “objective reality”) “by their nature”, they are also conscious, by their nature. Alison Simmons explains it well:

[Basic] consciousness is a kind of reflexive property of the first-order thought itself, so that every thought effectively has two objects: in virtue of having objective reality the thought has, say, celery as its object; and in virtue of having a reflexive property [i.e., basic consciousness] it also has itself as an object. In thinking about celery, a thinker thus becomes aware at once of both the celery and her act of thinking, but through different features of the first-order thought, viz., representation and [basic] consciousness, respectively. (2012: 6)[15]

In addition to basic consciousness, Descartes allows for what he calls “reflective consciousness” or “reflection”—or what I call “introspection.” Unlike basic consciousness, reflection on a thought is not built into that thought. It’s a separate, higher-order thought that stands over and above the first thought—an act of thinking about thinking. Descartes sometimes describes the contrast by saying that basic consciousness is “immediate” or “direct” as opposed to “reflective.” In the 4th Set of Objections, Arnauld mentions infants in the womb as a counterexample to Descartes’s claim that “the soul always thinks.” In reply, Descartes insists that “the mind begins to think as soon as it is implanted in the body of an infant, and that it is immediately conscious of its thoughts” (4R, 2:171–172/AT 7:246–267*). Arnauld evidently isn’t convinced, as he presses Descartes on this issue again, seven years later, in letters they exchange in the summer of 1648. In one letter, Descartes writes:

Finally, we make a distinction between direct and reflective thoughts corresponding to the distinction we make between direct and reflective vision, one depending on the first impact of the rays and the other on the second. I call the first and simple thoughts of infants direct and not reflective — for instance the pain they feel when some wind distends their intestines, or the pleasure they feel when nourished by sweet blood. But when an adult feels something, and simultaneously perceives that he has not felt it before, I call this second perception reflection, and attribute it to the intellect alone, in spite of its being so linked to sensation that the two occur together and appear to be indistinguishable from each other. (To Arnauld, 29 July 1648, 3:357/AT 5:221)

This is one of several of passages where Descartes uses terms for “feels” (sentit) interchangeably with terms for “is conscious of” (conscios) (cf. Pr. i.9, 1:195/AT 8A:8; Passions i.26, 1:338/AT 11:348-9); thus, he goes on to say that just as infants always “feel” their current sensations, we are always “conscious of” our current thoughts, even during sleep. What he is saying above, then, is that the infant’s feeling (or consciousness) of its pleasures, pains, and other sensations is “immediate” or “direct” in the sense that it comes with these sensations themselves, instead of proceeding through a “second perception” which mirrors or “reflects” the first. When we as adults feel a sensation, we may simultaneously reflect on it (and thereby perceive, in some cases, that we “have not felt it before”). So the basic consciousness that continually pervades the mind from the moment of conception is not the higher-order reflection that we have, occasionally, on some of our thoughts, in maturity.[16] Reflection is due “to the intellect alone,” Descartes says, even when you are reflecting on a sensation. The fact that infants have sensations without such reflection shows that reflection is not built into the nature of thought, like basic consciousness is. Minds necessarily think and necessarily have basic consciousness of their thoughts, but “it cannot in any way be regarded as essential” that minds “reflect on their thinking” (7R, 2:382/AT 7:559). While basic consciousness of a thought is an intrinsic feature of that very thought, introspection or reflection on a thought is a separate, higher-order thought whereby you attend to a lower-order thought.

The most salient cases of introspection are when you voluntarily turn your attention inward. As Descartes says to Burman, “the soul … has the power to reflect on its thoughts as often as it likes [i.e., at will], and to be conscious of its thought in this way” (16 April 1648, 3:335/AT 5:148–149*). Reflection or introspection is a way of being conscious of one’s thoughts, and we can engage in it voluntarily. Voluntary introspection is a particularly important case of introspection for Descartes, because it’s required for meditation: from the beginning to the end of the Meditations, you have to voluntarily reflect on your beliefs and other thoughts so as to examine them systematically. But Descartes needn’t insist that introspection is always voluntary. Just as an external stimulus like a fire alarm can grab your sensory attention, Descartes could allow that an internal state like anxiety can grab your reflective attention. Indeed, if the alarm is loud or the anxiety is intense, it may not only grab your attention but also keep it against your will, making it hard to concentrate on anything else. Attention can be directed either voluntarily or as an automatic response to a salient event. But either way, what matters about introspection is that it’s attentive.

Proponents of the Transparency-through-Introspection reading don’t always distinguish between basic consciousness on the one hand and reflection or introspection on the other. But it would help their case to do so, because it would allow them to make the following points. Descartes says that we never have thoughts “of which [we] are not in some way conscious” (1R, 2:77/AT 7:106*)—that is, the basic way—but what we need for self-knowledge is to be conscious of our thoughts in another way—that is, the reflective way, through introspection. Also, the fact that thoughts often occur without introspection fits with the passage in the Discourse where Descartes explains how people are often ignorant of their beliefs: “believing something and knowing that one believes it are different acts of thinking, and the one often occurs without the other” (Discourse, 1:122/AT 6:23).

Nevertheless, Descartes offers many counterexamples to Transparency-through-Introspection. We’ve already seen that Descartes rejects Transparency-through-Having when he says that people are often ignorant of their beliefs and passions. Those two cases are not clear counterexamples to Transparency-through-Introspection, because they do not portray such people as introspecting their thoughts. On the other hand, it’s not at all obvious that Descartes thinks such people could remedy their ignorance just by glancing inward, as Transparency-through-Introspection would imply. If Descartes thought self-knowledge were that easy to acquire, it wouldn’t be a problem worth grappling with; he would simply tell us to look within.[17] But even setting our first two cases aside, Descartes describes plenty of other cases where people are ignorant of their thoughts even while they are introspecting them. These cases refute Transparency-through-Introspection. And since you cannot introspect a thought unless that thought is yours, these cases also refute Transparency-through-Having.

In three of these cases, people form false beliefs about their thoughts, even while they are introspecting them:

  • Conceiving vs. imagining. Introspecting your act of conceiving, you may mistakenly believe that it’s an act of imagining. (To Mersenne, July 1641, 3:186/AT 3:395; see Radner 1988; Rozemond 2006)
  • Strength of belief. Introspecting your beliefs, you may overestimate their strength, particularly in matters of faith. (To Huygens, 10 October 1642, 3:216/AT 3:798–799; see Curley 1978: 177–178)
  • The clarity, distinctness, evidentness, or certainty of perceptions. “There is some difficulty,” Descartes says, “in recognizing which are the things that we distinctly conceive” (Discourse, 1:127/AT 6:33). Introspecting her sensory perceptions in the Third Meditation, the meditator concedes, “I previously accepted as wholly certain and evident many things which I afterwards realized were doubtful” (M3, 2:24/AT 7:35). In particular, she adds, the proposition that her sensations are caused by external objects which fully resemble her sensations is something which “through habitual belief I thought I perceived clearly, although I did not in fact do so” (M3, 2:25/AT 7:35; cf. Pr. i.44 1:207/AT 8A:21; cf. AT 8B:352). So, before undergoing the method of doubt, the meditator introspects her sensory perceptions and judges them to be “wholly certain,” “evident,” “clear,” or “distinct” though in fact, as she learns later, they are not. (See M. Wilson 1978: 155.) This point is of particular significance for the present study because, as I will argue, it extends to introspection as well. We are prone to think that our introspection is wholly certain, clear, evident, or distinct in many cases where it is not.

In two other cases, people form no belief about (part of) the content of their thoughts, even while they are introspecting them:

  • Sensory content. Introspecting one of your sensory ideas, you may suspend judgement about what it represents. (M3, 2:30/AT 7:43–44; 4R, 7:232–234; see Cottingham 2006: 181–182; Simmons 2012: 12)
  • Intellectual content. Introspecting one of your intellectual ideas, you may fail to judge that various propositions are “contained within it,” as part of its content. For example, Descartes holds that your idea of a triangle contains the proposition that its three angles equal to two right angles, and your idea of God contains the proposition that God necessarily exists. But your ideas have these contents only “implicitly,” insofar as you do not notice them. Introspecting, in such cases, is like looking at a treasure chest without seeing all the treasures within. (M5, 2:46/AT 7:63–64; 3:183–184/AT 3:383; see Simmons 2012: 11–12; M. Wilson 1978: 155)

In all five of these cases, the subject lacks certain knowledge of specific features her thoughts, even though she is the one who has those thoughts, and even though she introspects those thoughts. These cases therefore belie both Transparency-through-Having and Transparency-through-Introspection. More precisely, these cases show that neither version of Transparency secures certain knowledge of all of the specific features of one’s thoughts. One might wonder whether Transparency secures some certain knowledge, at least of the minimal thought-claim, that a thought-exists. We’ll see in Section 5 that Descartes rejects even this minimal form of Transparency.

But for now, let us consider another way of taming Transparency, not by restricting the range of truths it is supposed to reveal but by weakening the quality of the epistemic state it is supposed to account for. Remember that both versions of Transparency, as defined here, are claims about what it takes to acquire certain knowledge, which is infallible and indubitable. As noted, that is the only kind of knowledge Descartes is concerned with in the relevant texts. But when people say that Descartes is committed to the epistemological transparency of thought, it isn’t always clear that certain knowledge is at issue. We should be especially mindful of this point when contemporary epistemologists use Descartes as a foil, for they are generally theorizing about ordinary knowledge, which they generally take to be fallible and dubitable. Insofar as they treat Descartes as an interlocutor who is concerned with the same thing as they are, then, they portray him as being committed to a weaker analog of either of our two forms of Transparency, according to which either having a thought or introspecting a thought is sufficient for ordinary knowledge of that thought. This reading of Descartes is doubly mistaken. First, it misidentifies Descartes’s epistemic target. Second, in all five of the cases just considered, the subject lacks a true belief, and in the last two cases, the subject lacks a belief, concerning some feature of her introspected thoughts. True belief is necessary for ordinary knowledge. So, these cases show that introspecting a thought doesn’t even guarantee ordinary knowledge concerning all the features of that thought.

Some commentators refrain from attributing any form of Transparency to Descartes.[18] However, they have not explained why Descartes rejects Transparency, nor have they identified the theory of self-knowledge Descartes embraces instead.

4. Rationalism for Self-Knowledge

I propose that Descartes’s theory of self-knowledge is an application of his general theory of knowledge. His general theory of knowledge—commonly referred to as “Rationalism”—is well-recognized, but it will be helpful to review the basics.

The Cartesian mind has two basic faculties: intellect and will. The intellect (in the broad sense of the term) is the faculty of mental representations, which Descartes calls “perceptions” or “ideas.” There are three kinds of perceptions, due to three sub-faculties of the intellect: senses, imagination, and intellect in the narrow sense of the term (i.e., “the pure intellect,” also known as “reason,” “understanding,” or “the natural light”).

The will is the faculty of judgment. A perception by itself is not a judgment. A perception, due to the intellect, provides the content for a possible judgment. The will then either assents to the perception (forming a judgment), or withholds assent (in a state of doubt).

In order for a judgment to constitute certain knowledge (cognitio), it must be based on (i.e., formed by assenting to) a perception that is of such a kind as to be indubitable (cannot be doubted) and infallible (cannot be false). The only kind of perception that fits the bill is clear and distinct perception.[19] Thus, in order for a perception to provide certain knowledge, it must be clear and distinct as opposed to obscure or confused. Discussing “the most perfect certainty,” Descartes says:

A perception which can serve as the basis for a certain and indubitable judgement needs to be not merely clear but also distinct. (Pr. i.45, 1:207/AT 8A:21–22)

It is clear that we do not have this kind of certainty in cases where our perception is even the slightest bit obscure or confused; for such obscurity, whatever its degree, is quite sufficient to make us have doubts in such cases. (2R, 2:103/AT 7:145)

Sense-perception can be “very clear and in its own way distinct” (2:57/AT 7:83), and it can be “clear and distinct enough” for practical purposes (2:57/AT 7:83).[20] But only pure intellection can be clear and distinct in the strict sense of being indubitably clear and distinct, as is required for certain knowledge:

Again, we do not have the required kind of certainty with regard to matters which we perceive solely by means of the senses, however clear such perception may be. … Accordingly, if there is any certainty to be had, the only remaining alternative is that it occurs in the clear perceptions of the intellect and nowhere else. (2R, 2:104/AT 7:145; cf. Preface to French Principles, 1:182/AT 9B:7)

Clear and distinct intellection is not only necessary for certain knowledge; it is also sufficient. The authors of the 2nd Set of Objections read Descartes as holding that one must have certain knowledge of God in order to have certain knowledge of anything else. In reply, Descartes explains that this is not his view:

The fact that an atheist can ‘clearly know [clare cognoscere] that the three angles of a triangle are equal to two right angles’ is something I do not dispute. … (2R, 2:101/AT 7:141*)

One does not need knowledge of God to have clear knowledge (cognitio) of other things. Using her intellect, an atheist can have clear and distinct intellection of mathematical truths, for example, and that is all that is required for clear knowledge of those truths.[21]

Putting all of this together, we arrive at the central tenet of Descartes’s epistemology, standardly referred to as “Rationalism” because of the primacy it accords to the intellect or reason:


S has certain knowledge that p if and only if S has a clear and distinct intellectual perception that p.[22]

Descartes holds that we can have clear and distinct intellection, and thus certain knowledge, of necessary truths: for example, 2+3=5; what is done cannot be undone; a body is essentially an extended thing; a mind is essentially a thinking thing; etc. As noted, he also holds that we can have self-knowledge: certain knowledge of contingent truths concerning our own minds. In keeping with Descartes’s Rationalism, then, certain knowledge of contingent truths concerning one’s own mind must be acquired the way certain knowledge of any truth is acquired: through clear and distinct intellection.

Rationalism for Self-Knowledge

Where p is a contingent proposition concerning S’s own mind,

S has certain knowledge that p if and only if S has a clear and distinct intellectual perception that p.

There is a curious evasion in the literature. While commentators are aware of Descartes’s Rationalism, they generally make no appeal to it when writing on his account of self-knowledge. “Clear and distinct intellection” makes no appearance in formulations of Transparency. Commentators routinely acknowledge that clear and distinct intellection is what provides knowledge of necessary truths, but self-knowledge is treated as an exception to the rule, secured instead by Transparency.[23]

The irony is that self-knowledge is the farthest thing from an exception to Descartes’s Rationalism. It is his paradigm example. In the Discourse, he points to the cogito to consider “what this certainty consists in,” and he writes,

I observed that there is nothing at all in the proposition ‘I am thinking, therefore I exist’ to assure me that I am speaking the truth, except that I see very clearly that in order to think it is necessary to exist. (1:127/AT 6:33)

Likewise, he says in Meditation Three,

I am certain that I am a thinking thing. […] In this first item of knowledge there is simply a clear and distinct perception of what I am asserting. (2:24/AT 7:35)

Recalling the cogito again in Meditation Four, he says,

I have realized that from the very fact of my raising this question [whether anything exists] it follows quite evidently that I exist. I could not but judge that something which I understood so clearly was true. (2:41/AT 7:58)

Both in the Discourse and the Meditations, Descartes uses his first certainty—an instance of self-knowledge—as the exemplar which illustrates “in general what is required of a proposition in order for it to be true and certain” (1:127/AT 6:33) or “what is required for my being certain about anything” (2:24/AT 7:35). It is precisely by generalizing from that example that he comes to propose “the general rule” that certain knowledge is gained only through clear and distinct perception (2:24/AT 7:35). He has more work to do to confirm this proposal, but what suggests it to him in the first place is the shimmering exemplar of self-knowledge with the cogito.

In addition to being Descartes’s actual view, Rationalism explains why he rejects Transparency. Start with Transparency-through-Having. As we saw, readers often suppose that since Descartes says that having a thought is sufficient for having (basic) consciousness of that thought, he must hold that having a thought is sufficient for having certain knowledge of that thought. The assumption behind this reading is that basic consciousness is sufficient for certain knowledge.[24] Given Descartes’s Rationalism, this assumption implies the following:

? Basic consciousness is always a form of clear and distinct intellection.

Descartes denies this. In the text cited earlier where Descartes says that people can lack “evident knowledge” of their passions, he also explains why: “the passions are to be numbered among the perceptions which the close alliance between the soul and the body renders confused and obscure” (1:339/AT 11:350). You are always conscious of your passions, in his view, but you perceive them only obscurely and confusedly, by default, because of your embodied condition. When he asserts that “those who are most strongly agitated by passions are not those who know them best” (1:339/AT 11:350), he means that the more you are consumed by an emotional frenzy (like a fit of rage), the harder it is for you to see your passion clearly and distinctly for what it is: simply a thought, a state of the mind, an “‘emotion’ of the soul” (1:339/AT 11:350). Basic consciousness doesn’t guarantee certain knowledge, because it doesn’t meet Rationalism’s requirement of being a form of clear and distinct intellection.

How about Transparency-through-Introspection? Given Rationalism, introspection would guarantee certain knowledge only if the following were true:

? Introspection is always a form of clear and distinct intellection.

This proposal is more promising. Descartes indicates that introspection (reflection) is always a form of intellection:

[W]hen an adult feels something, and simultaneously perceives that he has not felt it before, I call this second perception reflection, and attribute it to the intellect alone, in spite of its being so linked to sensation that the two occur together and appear to be indistinguishable from each other. (To Arnauld, 29 July 1648, 3:357/AT 5:221, my bold, italics in CSM)

When, for example, we are asleep and notice [advertimus] that we are dreaming, we need imagination in order to dream, but to notice [advertamus] that we are dreaming we need only the intellect. (5R, 2:248/AT 7:358*)

In both of these quotes, even though the first-order thought is from the senses or imagination, the second-order, introspective act of reflecting on or noticing the first thought is from the intellect alone. So, introspection is always intellectual, even when it is focused on a state of sensation or imagination.

But is introspection always clear and distinct? The prevailing tendency in the literature is to assume that, for Descartes, these two conditions are mutually entailing: being intellectual and being clear and distinct (in the strict sense required for certainty).[25] Descartes endorses the entailment in one direction:

If a perception is clear and distinct, it is intellectual.

That claim is the heart of his Rationalist view that the intellect or reason is superior to the senses. But I will argue that, contrary to what is usually assumed, Descartes denies the converse claim:

? If a perception is intellectual, it is clear and distinct.

Indeed, he goes out of his way to stress that intellection is not always clear and distinct, and that in fact it’s often obscured by, and confused with, perceptions from the senses and imagination. This is one lesson of the famous wax passage in the Second Meditation. The meditator begins by viewing the wax obscurely and confusedly as something with a particular sensible shape, size, texture, color, taste, etc. She then notes that the wax can persist without any of these particular sensible qualities, and through the pure intellect she comes to conceive of the wax clearly and distinctly as merely an extended thing, which can take on countless different shapes and sizes. In addition to this metaphysical lesson about the nature of the wax, she draws an epistemological lesson about how she grasps—and how, unbeknownst to her, she had always grasped—the nature of the wax and other bodies, namely, not through her senses or imagination but through her intellect alone: “And yet, and here is the point, the perception I have of [the wax] is a case not a vision or touch or imagination – nor has it ever been, despite previous appearances – but of purely mental scrutiny; and this can be imperfect and confused, as it was before, or clear and distinct as it is now” (2:21/AT 7:31). Her grasp of the wax was always intellectual, even when it was obscured by and confused with sensory images.

In the wax example, the instance of intellection that is initially obscure and confused targets a necessary truth about the nature of the wax. I will argue that intellection can also be obscure or confused when it takes the form of introspection, targeting contingent truths about one’s thoughts. Indeed, introspection is ordinarily confused, even with respect to the minimal thought-claim, that a thought exists.

5. Minimal Transparency?

We now set our sights on the most cautious, most promising version of Transparency—the version entailed by every other version. This minimal version escapes the counterexamples above. What all of those counterexamples show is that even when you introspectively attend to one of your thoughts, you may still remain ignorant of specific facts about that thought, like what all of its contents are, or how strongly you hold it (if it’s a belief), or whether it’s an instance of conceiving or imagining, or whether or not it’s clear and distinct. But surely, you might think, Descartes must hold that introspection guarantees knowledge at least of the generic fact that you are thinking, or at least of the generic, impersonal fact—the minimal thought-claim—that a thought exists. Call this view “Minimal-Transparency-through-Introspection,” or, for short,

? Minimal Transparency
If S introspects a thought, S has certain knowledge that a thought exists.

Descartes accepts at least this bare minimum of Transparency—doesn’t he?

Actually, no. In the next two sub-sections, I will show that Minimal Transparency has two implications that Descartes rejects. First, given Descartes’s Rationalism, Minimal Transparency entails that introspection is always clear and distinct at least in presenting the minimal truth that a thought exists—but it isn’t. Second, Minimal Transparency entails that radical doubt isn’t needed to acquire certain knowledge that a thought exists—but it is.

5.1. The Need for Clear and Distinct Introspection

The time has come to tackle what Transparency readings curiously avoid: the central concept in Descartes’s philosophy—clear and distinct perception. Though a full treatment of that topic requires at least an essay of its own,[26] we can isolate the points that are needed here. Keep in mind that obscurity is the opposite of clarity, and confusion is the opposite of distinctness.

The only place where Descartes explains what he means by the terms ‘clear’ and ‘distinct’ is in Principles i.45. He begins with clarity: “I call something ‘clear’ when it is present and open <manifest> to the attentive mind […]” (1:207/AT 8A:22*). Then he turns to distinctness:

I call a perception ‘distinct’ when, as well as being clear, it is so sharply separated from all other perceptions that it contains within itself only what is clear. (1:207–208/AT 8A:22)

Clarity is more fundamental. Distinctness is defined in terms of clarity. Moreover, distinctness is defined negatively, not as clarity combined with some additional feature, but as clarity in the absence of—“sharply separated from”—anything unclear. Notice the double-negation: a perception is distinct to the extent that it’s not contaminated by what is not clear. Distinctness is simply the purest case of clarity, obtaining when a perception “contains within itself only what is clear”—when it is wholly or thoroughly clear. This bears emphasis:

(1) A distinct perception is just a thoroughly clear perception.

Within Pr. i.45, Descartes also says:

(2) "A perception which can serve as the basis for a certain and indubitable judgement needs to be not merely clear but also distinct.” (1:207/AT 8A:22*)

If I have interpreted distinctness correctly, then, it should follow that:

(3) A perception which can serve as the basis for a certain and indubitable judgement needs to be thoroughly clear. [from 1, 2]

And that is indeed what Descartes says:

It is clear that we do not have this kind of certainty in cases where our perception is even the slightest bit obscure or confused; for such obscurity, whatever its degree, is quite sufficient to make us have doubts in such cases. (2R, 2:103–104/AT 7:145)

A perception provides certainty only when it is thoroughly clear, unmarred by even the slightest degree of obscurity.

This point is crucial because, as Descartes says in the very next article, “a perception can be clear without being distinct” (Pr. i.46, 1:208/AT 8A:22)—that is, clear but confused. In other words, a perception can be clear—more precisely, it can be relatively clear, or even very clear—without being thoroughly clear. Even when a perception is very clear, it may be confused with an obscure perception, such that it is not distinct, not thoroughly clear, in which case it does not provide certain knowledge. Descartes gives an illustration of this point, and, tellingly, the example he chooses is a perception that would deliver self-knowledge if it were clear and distinct—the ordinary perception of one’s own pain:

A perception can be clear without being distinct. For example, when someone feels an intense pain, [c] the perception he has of this pain is indeed very clear, but is not always distinct. For people commonly confuse this perception with [o] an obscure judgement they make concerning the nature of something which they think exists in the painful spot and which they suppose to resemble the sensation of pain; but in fact it is the sensation alone which they perceive clearly. (Pr. i.46, 1:208/AT 8A:22)

Two perceptions are “commonly” (vulgō) or “ordinarily” (ordinairement) confused—literally “fused together” (confusio). One of them is (c) a very clear perception of pain. Importantly, the object of this perception—pain—is a sensation, which, for Descartes, is a kind of thought, existing only in the mind. Pain and other sensations are caused by the body but they are not in the body, so the perception of pain is not a perception of the body. Rather, it’s an inner perception of something within one’s own mind, a perception which Descartes goes on to identify as “inner consciousness [intimè conscii]” (i.66*, 1: 216/AT 8A:32). In this example, one’s (c) very clear perception of one’s mind is not distinct, because it’s fused with (o) an obscure perception of one’s body. Descartes calls the obscure perception an “obscure judgment” here, but a judgement is a perception, in his view: it’s a perception “with an additional form,” provided by the will’s assent (M3, 2:25–26/AT 7:37). The contents of the perceptions involved may be formulated as follows:

(c) very clear perception: a pain exists.

(o) obscure perception: something in my foot exists.

(c) and (o) fused together: a pain in my foot exists.

To explain this, I will adapt an expository device from an excellent paper by Alan Nelson (1997) and use diagrams as follows: each circle is an act of perception; what is printed inside each circle is the content of that perception; and a white interior indicates clarity while shades of grey mark degrees of obscurity. If the perception of pain were clear and distinct, it would be sharply separated from the obscure perception of the foot so as to be wholly clear:

But this is not how we ordinarily perceive the pain. We don’t just perceive pain and something in the foot. We perceive pain as something in the foot:

We generally regard [pain] not as being in the mind alone, or in our perception, but as being in the hand or foot or in some other part of our body. (Pr. i.67, 1:217/AT 8A:32–33)

Our perception of the pain and our perception of something in the foot are not separated but are fused into an indiscriminate whole:

With the two perceptions “mixed together” in this way, the obscurity of one “lessens the clarity” of the other.[27]

Now consider a corollary of (3), as it applies to claims about the existence of things:

(4) S cannot have certain knowledge that x exists unless S has a clear and distinct intellectual perception of x. [corollary of 3]

This claim follows from (3) given a very plausible principle of compositionality: insofar as a proposition involves two or more elements, how well you perceive that proposition is at least partly a function of how well you perceive each of its elements. For example, I cannot have a clear and distinct perception that the interior angles of a triangle must add up to 180 degrees while I have an obscure or confused conception of triangle or of degrees. Likewise, I cannot have a clear and distinct perception that x exists while I have an obscure or confused conception of x or of existence. The challenge is not with existence; it’s with x, the thing that exists—regardless of what that thing is. In fact, Descartes holds that

(5) People in general don’t have clear and distinct intellectual perceptions of their own minds or thoughts.

This deficit is not only typical, but natural; it begins in infancy as a result of our embodiment:

In our childhood the mind was so immersed in the body that although there was much that it perceived clearly, it never perceived anything distinctly. (Pr. i.47, 1:208/AT 8A:22; cf. 6R, 2:297/AT 7:441)

In infancy, none of our perceptions are purely intellectual, so none of them are clear and distinct enough for certainty. We can form clear and distinct intellectual perceptions in maturity, but most people continue to rely on their senses for all of their judgments and to conceive of everything imagistically (in terms of sensory qualities) and thus in a confused way:

Since […] there is nothing whose true nature we perceive by the senses alone, it turns out that most people have nothing but confused perceptions throughout their entire lives. (Pr. i.73, 1:220/AT 8A:37)

There are, however, few people who correctly distinguish between what they in fact perceive and what they think they perceive; for not many people are accustomed to clear and distinct perceptions. (7R, 2:348/AT 7:511)

Indeed, there are very many people who in their entire lives never perceive anything with sufficient accuracy to enable them to make a judgment about it with certainty. (Pr. i.45, 1:207/AT 8A:21)

There are two broad domains in which Descartes holds that we can achieve clear and distinct perception: mathematics and pure logic on the hand, and metaphysics on the other. Taken in context, when Descartes says that most people never perceive “anything” clearly and distinctly, he means anything in metaphysics. As he says to Mersenne: “there are few who are capable of understanding metaphysics” (16 October 1639, 3:139/AT 3:596). In the broad sense of the term, metaphysics includes all of ontology, all (true) claims about the existence of things, mental and physical. Mental things include God as well as the human soul (or mind, or self qua mind) and its properties (thoughts). Mental things are literally “metaphysical” in a narrow sense of the term: beyond the physical. But broadly construed, metaphysics includes claims about the existence of physical things (bodies), too. As such, any claim that affirms the existence of something—A body exists, God exists, I exist, and any thought-claim, including the minimal one, that a thought exists—is a claim of metaphysics. Self-knowledge is foundational in Descartes’s order of metaphysical discovery. You begin by establishing the existence of your own thoughts and self through the cogito—“the first principle” of Cartesian metaphysics (1:127/AT 6:32; 1:195/AT 8A:7; 1:184/AT 9B:10)—and only then can you go on to confirm the existence of God and of bodies. So, because people don’t perceive the existence of their own thoughts and selves clearly and distinctly, they can’t go on to perceive the existence of anything else clearly and distinctly.

Part of the problem is that people normally rely on sense-perception rather than pure intellection to judge their own existence. Discussing “those who have not philosophized in an orderly way”—that is, people in general—Descartes writes,

Although they believed that they were more certain of their own existence than of anything else [Et quamvis sibi certius esse putârint, se ipsos existere, quàm quidquam aliud], they failed to notice that they should have taken ‘themselves’ in this context to mean their minds alone. They were inclined instead to take ‘themselves’ to mean only their bodies – the bodies which they saw with their eyes and touched with their hands, and to which they attributed the power of sense-perception […]. (Pr. i.12, 1:196–197/AT 8A:9*)

Descartes does not say that people generally are certain of their own existence but are just mistaken about what they are. He says that people “believe” that they are certain of their own existence (and more certain of this than of anything else). And the whole thrust of the passage is to suggest that they are mistaken in this belief. Insofar as they identify themselves with their bodies, they judge that they themselves exist on the sensory perception of their bodies, “which they saw with their eyes and touched with their hands,” and since sensory perceptions do not provide certainty, they are not actually certain, as they take themselves to be. Descartes identifies the root of the problem in the preceding sentence, where he says that people in general have “have never taken sufficient care to distinguish the mind from the body,” (Pr. i.12, 1:196/AT 8A:9)—that is, to render their perceptions of their minds distinct from their perceptions of their bodies. In general, these perceptions are not distinct but confused. Altogether then, Descartes is saying that people generally believe they have certain knowledge of their own existence, but they don’t, precisely because they haven’t made their perceptions of themselves (their minds) distinct.

Elsewhere, Descartes says we have a capacity to acquire a certain knowledge of I exist and God exists through pure intellection, but we often fail to exercise this capacity because “such knowledge is somewhat obscured by the soul’s mingling with the body.” Tellingly, he makes this point explicitly with reference to his cogito argument:

You will surely admit that you are less certain of the presence of the objects you see than of the truth of the proposition ‘I am thinking, therefore I exist.’ Now this knowledge is not the work of your reasoning or information passed on to you by teachers; it is something that your mind sees, feels and handles; and although your imagination insistently mixes itself up with your thoughts and lessens the clarity of this knowledge by trying to clothe it with shapes, it is nevertheless a proof of the capacity of our soul for receiving intuitive knowledge [une connoissance intuitive] from God. (To [Silhon], March or April 1648, 3:331/AT 5:136–137)

People have an inborn capacity to gain intuitive knowledge through “I am thinking, therefore I am,” but instead of using this capacity they tend to imagine everything, “clothing it with shapes,” even when the thing in question, the self, is an immaterial thing that cannot be imagined.

Because of our embodied nature, we are prone to conceive of minds in a confused way, beginning with our own:

I had from my earliest years conceived of my mind and body as a unity of some sort (for I had a confused awareness that I was composed of mind and body). (6R, 2:299–300/AT 7:445)

Commentators almost universally recognize that Descartes believes that our awareness of the mind is ordinarily confused with our awareness of the body. Some scholars have noted that, in Descartes’s view, so long as we retain this confusion we cannot have knowledge of the existence of any mind, not even our own.[28] I want to add that the very same point applies to the existence of the mind’s properties, or “what belongs to the mind,” namely thoughts:

All our ideas of what belongs to the mind [i.e., thoughts] have up till now [before the Meditations] been very confused and mixed up with our ideas of sensible things [i.e., bodies] [plane confusas & cum rerum sensibilium ideis permixtas]. (2R, 2:94/AT 7:130–131*)

Before we go through the Meditations, all our ideas of thoughts are “very confused and mixed up with” our ideas of bodies. From (4) and (5) above it follows that:

(6) People in general don’t have certain knowledge that their minds exist or that thoughts exist.

Descartes draws exactly this inference in the very next sentence of the last passage quoted above:

This is the first and most important reason for our inability to understand with sufficient clarity the customary assertions about the soul and God. (2R, 2:94/AT 7:130–131*)

In Descartes’s time, “customary assertions about the soul and God” would include “God exists,” but more importantly for our purposes they would surely include “The human soul exists”—or in the first person, “I exist.” You can’t have self-knowledge with a confused self-conception.

The reason people in general lack self-knowledge is not because they have failed to introspect. As we noted earlier,

(7) People in general have engaged in introspection.

From (6) and (7), it follows that, contrary to Minimal Transparency,

(8) Ordinary introspection doesn’t even provide certain knowledge that a thought exists. [from 6,7]

Again, Rationalism explains why not: it’s because ordinary introspection doesn’t present anything clearly and distinctly, not even that a thought exists.

Further evidence that Descartes rejects Minimal Transparency comes from a proper understanding of his use of doubt.

5.2. The Need for Radical Doubt

My next argument has two main premises:

(i) Minimal Transparency entails that it’s easy to acquire certain knowledge that a thought exists: it doesn’t require the use of any special method, much less the method of radical doubt.

(ii) But Descartes holds that it isn’t easy to acquire certain knowledge that a thought exists: it does require the use of a special method, specifically the method of radical doubt.

So it follows, again, that Descartes rejects Minimal Transparency.

Regarding the first premise, notice that what every version of Transparency has in common is the notion that self-knowledge is easy, in the sense defined above. In the literature, Transparency-through-Introspection is offered as an alternative to Transparency. But although it differs from Transparency-through-Having, I see it as another version of Transparency, because it, too, implies that self-knowledge is easy. If Transparency-through-Introspection is correct, then self-knowledge is not inevitable, as Transparency-through-Having says it is, but it’s still easy in the sense that you don’t need radical doubt for self-knowledge. You just need to look inward.

Now, strictly speaking, what Transparency-through-Introspection entails is not that self-knowledge is easy, but that it’s as easy or as hard as introspection is. One might hold that it is hard to look inward and therefore hard to acquire self-knowledge. But when commentators attribute this view to Descartes, there is no hint that looking inward should be any more difficult than looking to the left or the right. Moreover, Descartes himself seems to hold that introspection is easy. Burman reports Descartes as saying that “the soul … has the power to reflect on its thoughts as often as it likes, and to be conscious of its thought in this way” (16 April 1648, 3:335/AT 5:148–149*). Introspection (reflection) is a “way” of being conscious of our thoughts, and we can introspect at will, whenever we like.[29] More to the point, Descartes’s meditator illustrates that introspection is easy in the precise sense of “easy” that concerns us here: in order to introspect, you do not need the method of radical doubt. Introspection is familiar; radical doubt is rare. The meditator is engaged in introspection right from the outset, observing her beliefs and proceeding to examine them, well before she succumbs to radical doubt. So, introspection doesn’t require radical doubt. Minimal Transparency says that introspection is sufficient for certain knowledge that a thought exists, so—as per premise (i)—Minimal Transparency entails that radical doubt isn’t needed for acquiring certain knowledge that a thought exists.

But—as per premise (ii)—Descartes maintains that radical doubt is needed for that purpose. My evidence for this claim begins with a startling confession by Descartes which seems to have gone unnoticed: he admits, in correspondence, that his Discourse on the Method fails to deliver radical doubt, and that—-for that very reason—it fails to deliver self-knowledge. The Discourse contains “I am thinking, therefore I am,” so it looks like it does the job. But remarkably, it doesn’t. The cogito argument is sound but it doesn’t yield certain knowledge without radical doubt. In three letters, Descartes confesses to three different friends that

(a) the Discourse is inadequate as a means of acquiring certain knowledge of the existence of metaphysical things (the self and God),

and that this is precisely because

(b) the Discourse does not provide grounds for radical doubt—that is, doubt about all sensory perceptions and about the existence of all physical things.

The metaphysical things he mentions are God (in all three letters) and the self (in one letter), but I will explain that the very same lesson applies to properties of the self—thoughts—as well.

In one letter, to Vatier, Descartes concedes that (a) the argument which the Discourse offers for “the existence of God” is “obscure” rather than “certain and evident,” and that “the principal reason for this obscurity” is that (b) the Discourse does not broach the kind of skeptical argument that is “necessary to withdraw the mind from the senses” and cast doubt on “all material things” (22 February 1638, 3:85–86/AT 1:560*, emphasis in original). In another letter, to Mersenne, he makes the same pair of points (27 February 1637, 3:53/AT 1:350).

And in a third letter, to Silhon, he goes further. He says that the Discourse is ill-equipped to deliver certain knowledge not only of the existence of God, but also of the existence of the self:

I agree, as you observe, that [a] there is a great defect [grand defaut] in the [Discourse], and that I have not expounded, in a manner that everyone can easily grasp, the argument by which I claim to prove that there is nothing at all more evident and certain than the existence of God and of the human soul [l’existence de Dieu & de l’ame humaine]. But I did not dare to try to do so, since [b] I would have had to explain at length the strongest arguments of the skeptics to show that there is no material thing of whose existence one can be certain [i.e., to induce radical doubt] – and by this same means [par même moyen] I would have accustomed the reader to detach his thoughts from things that are perceived by the senses, and then I would have shown that a man who thus doubts everything material cannot for all that have any doubt about his own existence [sa propre existence]. (May 1637, 3:55/AT 1:353)

There is (a) “a great defect” in the way his Discourse seeks to demonstrate “the existence of God and of the human soul”—the self, one’s “own existence,” in the first person: I exist—because (b) that work does not induce radical doubt. While Descartes elsewhere acknowledges the difficulty of establishing one’s nature as something distinct from the body (1:9/AT 7:13), his concern here is with the challenge of establishing one’s very existence. The Discourse does not rise to the challenge. The Meditations does.

I will say more about the Meditations in the next section, but for now it will be helpful to register the following points. In both the Discourse and the Meditations, the cogito appears at the climax of their respective presentations of the method of doubt. The difference is that in the Discourse, where Descartes first entertains skepticism, he merely flirts with it. He mentions just two skeptical considerations against the senses—the Misperception Argument (that “our senses sometimes deceive us”) and the Dreaming Argument (that we are sometimes mislead by “the illusions of dreams”) (1:127/AT 6:32)—and then asserts the cogito. The Misperception Argument calls into question some bodies (e.g., ones that are subject to familiar illusions). The Dreaming Argument calls into question any body in one’s apparent surroundings. But neither of them calls into question all bodies, including one’s own, as required for radical doubt.[30] The Meditations, by contrast, deploys a far more extensive battery of skeptical weapons culminating in the Deceiver Scenario, the supposition that there is an evil demon who is deceiving the meditator about the existence of all bodies, including her own. This supposition does give rise to radical doubt. As we’ll see, the meditator’s formulation of the cogito explicitly articulates this doubt, and it works precisely because of it. Indeed, this strategy for self-knowledge, missing in the Discourse, is just what Descartes is gesturing to in the letter above.

It’s important to see that, in this letter, radical doubt is cast in two different roles. In the first sentence, radical doubt plays what I call

the contrastive role: doubt about the existence of all physical things is contrasted with certainty about the existence of metaphysical things (God and the self).

This contrast reveals that metaphysical things are “more evident and certain” than physical things that are simultaneously in doubt. For all this says, you could achieve certainty of metaphysical things without radical doubt; you just wouldn’t have the contrast.

But in the second sentence, radical doubt is portrayed, more fundamentally, as the “means” of liberating one’s mind from the senses and leading one beyond “any doubt”—to certainty—of one’s “own existence.” Here, radical doubt plays what I call

the preparatory role: doubt about the existence of all physical things is required to prepare the mind to achieve certainty about the existence of metaphysical things (beginning with the self).

Doubt is not merely contrasted with certainty—in metaphysics, it is the very means by which certainty is achieved. Descartes sounds this lesson repeatedly. He speaks of “doubt as a means of acquiring a clearer knowledge of the truth [dubitationem tanquam medium ad clariorem veritatis cognitionem]” (To Buitendijck 1643, 3:229/AT 4:63). He says his “method of universal doubt” is “useful to prepare the mind in order to establish the truth …” (5R, 2:270/AT 203–220). Remarking on how he used the method of doubt to “rid [himself] of all the rest of [his] opinions,” he says,

In doing this, I was not copying the skeptics, who doubt only for the sake of doubting and pretend to be always undecided; on the contrary, my whole aim was to reach certainty – to cast aside the loose earth and sand so as to come upon rock or clay. (Discourse, 1:125/AT 6:29)

Moreover, certainty arises not from just any doubt, but from all-embracing radical doubt. His conclusions are

deduced step by step, not from principles which are obscure and unknown, but, in the first place, from total doubt about all things. (To Buitendijck 1643, 3:229/AT 4:63)

From this universal doubt [...] I propose to derive the knowledge of God, of yourself, and of everything in the universe. (Search, 2:409/AT 10:514)

Now the best way of achieving a firm knowledge of reality is first to accustom ourselves to doubting all things, especially corporeal things. (2R, 2:94/AT 7:130)

Indeed, it’s the only way:

[…] there are many preconceived opinions that keep us from knowledge of the truth. It seems that the only way of freeing ourselves from these opinions is to make the effort, once in the course of our life, to doubt everything which we find to contain even the smallest suspicion of uncertainty. (Pr. i.1, 1:193/AT 8A:5; cf. 3R, 2:121/AT 7:172)

In metaphysics, doubt, paradoxically, is the only route to certainty. Not just any doubt. Radical doubt. And this point applies specifically to first principles:

I know of no other way of making sound judgments about the notions which can be taken for principles, except that we must prepare our mind to divest itself of all the views with which it is preoccupied, and to reject as doubtful everything that might be doubtful. (To Mersenne, 15 November 1638, 3:129/AT 2:435)

Descartes is talking about his work in metaphysics. The cogito is “the first principle” of his metaphysics (1:127/AT 6:32; cf. 1:195/AT 8A:7; 1:184/AT 9B:10). So what he’s saying, in effect, is that radical doubt is needed to “prepare our mind” for the cogito.

The implication is striking. Contrary to what is widely assumed, Descartes does not believe that you can acquire self-knowledge simply by reasoning, “I am thinking, therefore I am.” If that were all it took, the Discourse would suffice, but to the contrary, Descartes says, that work is defective.

The difficulty is not in drawing the inference, which is utterly trivial. If you already know by introspection that you are thinking, you can very easily know by inference that you exist. But it’s not easy. Certain knowledge that I exist is difficult to acquire. So, certain knowledge that I am thinking must be difficult, too.

The difficulty, as Descartes sees it, is not in ascribing the property, thinking, to its bearer, I. That sort of difficulty is alleged not by Descartes but by his critics. For example, the 19th century aphorist Georg Lichtenburg (1990) famously quips that Descartes is not entitled to affirm I am thinking, but only There is thinking, or in other words, A thought exists.[31] Those who share this worry sometimes invoke Hume for reporting that when he looks within, all he finds are sensations, perceptions, and other thoughts, but no Cartesian ego lying behind them.[32] But rightly or wrongly, Descartes has no room for this concern. For him, it is manifestly self-evident that there cannot be a thought without a thinker (Pr. i.10, 1:196/AT 8A:8; 2R, 2:100/AT 7:140). Each actual thought, in his view, is just a mind’s way of being at a given moment (just as each actual shape is just a body’s way of being at a given moment). So, what appears to me through introspection are not just my thoughts but also my mind, or my self qua mind, such that my thoughts are presented to me as mine. In other words, when introspection reveals to me that a thought exists, it also reveals to me that I am thinking. If it were easy to acquire knowledge that a thought exists, then, it would be easy to acquire certain knowledge that I am thinking. But, certain knowledge that I am thinking is difficult to acquire. So, certain knowledge that a thought exists must be difficult, too.

Acquiring such knowledge calls for something other than ordinary introspection, namely radical doubt—more doubt than the Discourse delivers. Minimal Transparency entails just the opposite: that it’s easy to gain certain knowledge that a thought exists; you don’t need radical doubt; you just need to introspect. So once again, Descartes rejects even Minimal Transparency.

5.3. Distinctness Through Doubt

In Section 5.1, we saw that ordinary introspection does not provide self-knowledge because even when it is clear, it is not distinct. In Section 5.2, we saw that radical doubt is necessary for self-knowledge. These two points are connected: radical doubt is necessary for self-knowledge because it is what makes introspection distinct.

Let us return to the case Descartes uses for illustration, the clear but confused introspection of pain. In order to make introspection clear and distinct, it must be “sharply separated” from all obscure perceptions bodies. But how can that be done? If you step on a nail, the resulting pain will inherently feel like it’s in your foot. There is nothing you can do to feel such pain as pain simpliciter, without feeling it as if it were in your body. So how could you ever come to perceive pain (or any other sensation) distinctly?

The answer is: by doubting that you have a body. In addition to the inevitable experiential confusion of feeling pain as if it were in our bodies, we normally succumb to the cognitive confusion of judging pain to be (or regarding it as) something in our bodies. “We generally regard [putentur/creu] [pain] not as being in the mind alone, or in our perception, but as being in the hand or foot or in some other part of our body” (Pr. i.67, 1:217/AT 8A:32–33†).

We cannot stop feeling pain as if it were in the foot, but we can stop judging it to be: “But the fact that we feel a pain as it were in our foot does not make it certain that the pain exists outside our mind, in the foot” (i.67, 1:217/AT 8A:33). We can doubt that pain is where it feels like it is (in the body), and it is through such doubt that we make our introspection of pain distinct. At work here is the point that judging and doubting—giving assent and withholding assent—are incompatible attitudes. You cannot give assent to a perception at the same time as you withhold assent from that perception. Thus:

If you assent to perception A while withholding assent from a perception B, you thereby “sharply separate” or distinguish perception A from perception B. 

Applying this point to the case of perceiving sensations, Descartes writes,

In order to distinguish what is clear in this connection from what is obscure, we must be very careful to note that pain and colour and so on are perceived clearly and distinctly when they are regarded merely as sensations or thoughts. (Pr. i.68, 1:217/AT 8A:33†)

Whether or not you perceive something clearly and distinctly depends partly on what you regard it as, what you judge it to be. When you have the confused perception described above, you regard pain (very clearly) as a sensation while you also regard it (obscurely) as being in your foot. The way to make your perception of your pain distinct is to stop regarding it as something in your foot and regard it merely as something in your mind, as a sensation or thought:

Sensations … may be clearly and distinctly perceived provided we take great care in our judgements [judicemus] concerning them to include no more than what is strictly contained in our perception – no more than that of which we have inner consciousness [intimè conscii]. (Pr. i.66*†, 1:216/AT 8A:32)

What you perceive very clearly, through inner consciousness, is nothing other than the sensation in your mind. To make that perception clear and distinct, you must withdraw assent from—that is, you must doubt—the obscure perception with which it is normally fused, so that you no longer judge that a pain in my foot exists, and judge merely that a pain exists. Since this remaining judgement is now based on introspection that is not only clear but also distinct—thoroughly clear—it now constitutes certain knowledge that a pain exists.

Again, coming to perceive a sensation clearly and distinctly in this way generally won’t change how the sensation feels. If you step on a nail, then for as long as you experience the resulting pain it will continue to feel like it’s in your foot, even if you no longer judge it to be your foot. This is just one instance of the general point that, for the most part, our sensory experiences are stubbornly impervious to cognitive revisions. An oar in water still looks bent when you learn that it’s straight; a tower in the distance still looks round when you learn that it has corners; the sun still looks medium-sized when you learn that it’s humungous. Likewise, pain and other sensations still feel like they’re in your body when you doubt that they really are. Nevertheless, it is precisely through that act of doubt that you separate your introspection of your sensations from obscure perceptions of your body, thereby rendering your introspection distinct. In metaphysics, beginning with self-knowledge, the road to distinctness is paved with doubt.

Let us now consider how this doubt works in the meditator’s version of the cogito. At the beginning of Meditation Two, the meditator suffers radical doubt while entertaining the scenario of the evil deceiver. This doubt makes her introspection clear and distinct, such that she can use a cogito argument to acquire self-knowledge. Her version of the argument substitutes the generic premise, I am thinking, with a specific premise that indicates what she is thinking at the pivotal moment, namely,

I have convinced myself [mihi persuasi] that there is absolutely nothing in the world, no sky, no earth, no minds, no bodies. (2:16–17/AT 7:25)

She then notes,

If I convinced myself of something then I certainly existed. (2:16–17/AT 7:25)

This second point could be read as a second premise, in which case the argument is an instance of modus ponens. Or it could be read as expressing the entailment from the first (and only) premise to the conclusion that she exists. Either way, she draws the obvious inference that she exists.[33] Notice that when the meditator says she has “convinced” herself that there is no external world, this does not mean she believes as much. What she means, rather, is that by deliberately supposing herself to be globally deceived, she has brought herself to doubt that there is an external world (7R, 2:319/AT 7:474). Thus, the premise could be paraphrased:

I am doubting the external world.

Looking back at this moment from Meditation Four, the meditator says,

during these past few days I have been asking whether anything in the world exists, and I have realized that from the very fact of my raising this question it follows quite evidently that I exist. (2:41/AT 7:58)

The argument recalled here could be paraphrased: I am questioning (doubting, withholding assent from) the existence of the external world, therefore (“it follows that”) I exist. So the meditator’s cogito argument is, in effect, I am doubting the external world, therefore I exist. And again, the fact that she is doubting the external world is precisely what makes her introspection clear and distinct, such that she can acquire certain knowledge of a thought-claim from which she can infer that she exists.[34]

Self-knowledge, so acquired, does not involve or presuppose certain knowledge of the claim (established much later, in Meditation Six) that the mind is really distinct and separable from the body. Doubting the existence of one’s body is necessary for self-knowledge, but that means withholding assent from all claims about the body; it does not mean assenting to the claim that the mind is separable from the body. Descartes marks this as the difference between “abstraction” and “exclusion.[35] When the meditator acquires self-knowledge through radical doubt, her judgement about her self is one that abstracts from (does not affirm) her body, for she is suspending judgment about all claims concerning bodies. When she completes the real distinction argument in Meditation Six, her judgement about her self is one that excludes her body, for she now comes to judge (allegedly with certainty) that she is distinct and separable from her body.

6. Conclusion

My concern in this paper has been an interpretive one, to challenge the prevailing view that Descartes’s is commitment to Transparency, and to reveal the approach to self-knowledge that he actually holds. Though I cannot defend his view here, let me close by remarking on why we should find it intriguing from our contemporary point of view.

Start with Transparency-through-Having. Like Descartes, many of us today would reject this view, but probably not for the same reason. To see the difference, consider again this line of thought, which has often been mistakenly attributed to Descartes:

1. If S has a thought, then S is conscious of that thought.

2. If S is conscious of a thought, then S has certain knowledge of that thought.

Therefore, Transparency-through-Having: If S has a thought, then S has certain knowledge of that thought.

In our post-Freudian era, we commonly take it for granted that our minds are populated with all manner of beliefs, desires, fears, and other thoughts of which we are not conscious. Indeed, the unconscious mind has become a staple of our folk psychology. We therefore find it natural or even obvious to deny the conclusion above by denying premise 1: we don’t have certain knowledge of all our thoughts, we’re inclined to say, because we aren’t conscious of all our thoughts.

Part of what makes Descartes’s approach so fascinating is that he accepts premise 1 and yet he still denies the conclusion. He does so, of course, by denying premise 2: we don’t have certain knowledge of all our thoughts, he maintains, because consciousness is not sufficient for certain knowledge. If we examine the matter, we too might reject premise 2 (along with premise 1), but I don’t think such examination comes naturally. Given that we reject premise 1, it might seem otiose to examine and potentially reject premise 2. And so long as we don’t examine it, we may very well be assuming it. It may seem to us that we are, as it were, “too close” to our conscious thoughts to be ignorant about them; and so, as Timothy Williamson observes, we have a strong tendency to regard the conscious part of our minds as a “cognitive home” where knowledge is given—easily, without method, for free.[36] So even as we insist (against Descartes) that we aren’t conscious of all our thoughts, we might nevertheless assume (against Descartes) that we have certain knowledge of the limited range of thoughts that we are conscious of. If that pair of views is correct, consciousness fails to deliver self-knowledge only because of its limited range. For Descartes, by contrast, consciousness fails to deliver self-knowledge because of its limited quality. Consciousness ranges over all of our thoughts, in his view, but it never yields certain knowledge of any of them, because it never has the epistemic credentials of clear and distinct intellection. Whether he is right or not, it’s worth noting that—as a way of refuting Transparency-through-Having—his denial of premise 2 is more interesting than our denial of premise 1. Our way of rejecting Transparency-through-Having boils down to saying that we can be ignorant of unconscious thoughts. That’s like saying that we can be ignorant of unperceived bodies. It should come as no surprise.

Unlike consciousness, Descartes holds that introspection reaches only a limited range of thoughts at any given time, but he also holds that, ordinarily, it isn’t clear and distinct. In his view, then, introspection is somewhat analogous to sense-perception which reaches only a limited range of bodies at any given time and also isn’t clear and distinct. But there is a crucial difference, because, unlike sense-perception, introspection is a form of intellection and we can make it clear and distinct. It is difficult to do so, however, and the way to overcome this difficulty is by employing the method of radical doubt, best exemplified in the Meditations. Since people generally haven’t undergone such doubt, they don’t have self-knowledge: not of I exist, nor of specific claims concerning thoughts, nor of the generic claim, I am thinking, nor even of the minimal claim, A thought exists. And since self-knowledge is foundational in metaphysics, it follows, as Descartes says, that people generally don’t have certain knowledge of “anything” in metaphysics, including God and bodies.

Insofar as Transparency is an epistemological doctrine (about certain knowledge, not just consciousness), then, Descartes is not the high priest of Transparency. He may be one of its most radical opponents.

Freud famously compared the mind to an iceberg: the vast bulk of it is the unconscious part submerged underwater while the conscious, introspectible part is only the visible tip. While conceding ignorance about what’s happening beneath the murky waters, we tend to take it for granted that when it comes to the conscious, introspectible tip of the iceberg, we are standing, epistemically, on solid ground. Appropriately enough, Freud’s iceberg metaphor has the effect of a Freudian slip, revealing more than its author intended. For as Descartes might have pointed out, the tip of an iceberg is slippery ice.


I am deeply indebted to Colin Chamberlain for providing extensive comments on multiple drafts and to Michael Della Rocca for invaluable advice. Very special thanks also to Robert Pasnau, Marleen Rozemond, and Alison Simmons, who offered generous feedback on more than one occasion. I am also grateful for insightful input from Omri Boehm, John Bengson, Taylor Carman, John Carriero, David Chalmers, Philip Choi, Jonny Cottrell, David Cunning, David Friedell, Daniel Garber, Sean Greenberg, John Morrison, Christia Mercer, Carol Rovane, Lisa Kerr, Patricia Kitcher, Anat Schechtman, Tad Schmaltz, and Katja Vogt; from audiences at Amherst College, CUNY Graduate Center, and the Harvard History of Philosophy Workshop; from my former students in seminars at Barnard College and Columbia University; and from the anonymous editor and two referees of this journal.


  • Alanen, Lilli (1981). On the So-Called ‘Naive Interpretation’ Of ‘Cogito, Ergo Sum’. Acta Philosophica Fennica, 32, 9–29.
  • Alanen, Lilli (1992). Thought-Talk: Descartes and Sellars on Intentionality. American Philosophical Quarterly, 29(1), 19–34.
  • Alanen, Lilli (2003). Descartes’s Concept of Mind. Harvard University Press.
  • Almog, Joseph (2005). What Am I? Descartes and the Mind-Body Problem. Oxford University Press.
  • Aquila, Richard E. (1988). The Cartesian and a Certain ‘Poetic’ Notion of Consciousness. Journal of the History of Ideas, 49(4), 543–562.
  • Bar-On, Dorit (2004). Speaking My Mind: Expression and Self-Knowledge. Oxford University Press.
  • Barrier, Thibault (2017). La capture de l’esprit: attention et admiration chez Descartes et Spinoza. Les Études philosophiques, 171(1), 43–58.
  • Barth, Christian (2011a). Bewusstsein bei Descartes. Archiv für Geschichte der Philosophie, 93(2), 162–194.
  • Barth, Christian (2011b). Leibnizian Conscientia and Its Cartesian Roots. Studia Leibnitiana, 43(2), 216–236.
  • Barth, Christian (2012). Intentionalität und Bewusstsein in der frühen Neuzeit: Die Philosophie des Geistes von René Descartes und Gottfried Wilhelm Leibniz [Intentionality and Consciousness in Early Modern Philosophy: René Descartes’ and Gotfried Wilhelm Leibniz’ Philosophy of Mind] (Habilitation thesis). Humboldt-Universität zu Berlin.
  • Barth, Christian (2013). [Review of the book The Early Modern Subject: Self-Consciousness and Personal Identity from Descartes to Hume by Udo Thiel]. Philosophy in Review, 33(1), 85–88.
  • Barth, Christian (2016). Consciousness in Early Modern Philosophy: Remarks on Udo Thiel’s Account. Kant-Studien, 107(3), 515–525.
  • Bennett, Jonathan (2001). Learning from Six Philosophers (Vol. 1). Oxford University Press.
  • Berker, Selim (2008). Luminosity Regained. Philosophers' Imprint. 8(2), 1–22.
  • Beyssade, Jean-Marie (1979). La philosophie première de Descartes : le temps et la cohérence de la métaphysique. Flammarion.
  • Boehm, Omri (2014). Freedom and the Cogito. British Journal for the History of Philosophy 22(4), 704–724.
  • Boghossian, Paul (2008). Content and Justification: Philosophical Papers. Oxford University Press.
  • Broughton, Janet (1984). Skepticism and the Cartesian Circle. Canadian Journal of Philosophy, 14(), 593–615.
  • Broughton, Janet (2002). Descartes’s Method of Doubt. Princeton University Press.
  • Broughton, Janet (2008). Self-Knowledge. In Janet Broughton and John Carriero (Eds.), A Companion to Descartes (179–195). Basil Blackwell.
  • Brown, Deborah (2007). Augustine and Descartes on the Function of Attention in Perceptual Awareness. In Sara Heinämaa, Vili Lähteenmäki, and Pauliina Remes (Eds.), Consciousness: From Perception to Reflection in the History of Philosophy (153–175). Springer.
  • Burge, Tyler (1988). Individualism and Self-Knowledge. The Journal of Philosophy, 85(11), 649–663.
  • Campbell, John (2012). Cogito Ergo Sum II: Lichtenberg and the Cogito. Proceedings of the Aristotelian Society112(3), 361–378.
  • Carriero, John (1987). The First Meditation. Pacific Philosophical Quarterly68(3/4), 222–248.
  • Carriero, John (1999). Painting and Dreaming in the First Meditation. Acta Philosophica Fennica64, 13–46.
  • Carriero, John (2008a). Between Two Worlds: A Reading of Descartes’s Meditations. Princeton University Press.
  • Carriero, John (2008b). The Cartesian Circle and the Foundations of Knowledge. In Janet Broughton and John Carriero (Eds.), A Companion to Descartes (302–318). Blackwell.
  • Carroll, Lewis (1895). What the Tortoise Said to Achilles. Mind, 4(14), 278–280.
  • Carruthers, Peter (2011). The Opacity of Mind: An Integrative Theory of Self-Knowledge. Oxford University Press.
  • Cassam, Quasim (2008). Contemporary Reactions to Descartes’s Philosophy of Mind. In Janet Broughton and John Carriero (Eds.), A Companion to Descartes (482–495). Basil Blackwell.
  • Cassam, Quasim (2014). Self-Knowledge for Humans. Oxford University Press.
  • Clarke, Desmond (2003). Descartes's Theory of Mind. Oxford University Press.
  • Cottingham, John (trans. and ed.) (1976). Descartes’ Conversation with Burman. Clarendon Press.
  • Cottingham, John (1978). Descartes on ‘Thought’. Philosophical Quarterly, 28(112), 208–214.
  • Cottingham, John (1986). Descartes. Basil Blackwell.
  • Cottingham, John (2006). The Mind-Body Relation. In S. Gaukroger (Ed.), The Blackwell Companion to Descartes’ Meditations (179–192). Basil Blackwell.
  • Cunning, David (2007). Descartes on the Dubitability of the Existence of Self. Philosophy and Phenomenological Research, 74(1), 111–131.
  • Cunning, David (2010). Argument and Persuasion in Descartes’ Meditations. Oxford University Press.
  • Curley, Edwin (1978). Descartes Against the Skeptics. Harvard University Press.
  • Curley, Edwin (2006). The Cogito and the Foundations of Knowledge. In Stephen Gaukroger (Ed.), Blackwell Guide to Descartes’ Meditations (30–47). Wiley-Blackwell.
  • Descartes, René (1984–1991). The Philosophical Writings of Descartes (3 Vols., John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, Trans.). Cambridge University Press.
  • Descartes, René (1996). Oeuvres de Descartes (new ed., 11 Vols.). Charles Adam and Paul Tannery (Eds.). Vrin. Cited as AT.
  • Della Rocca, Michael (2005). Descartes, the Cartesian Circle, and Epistemology without God. Philosophy and Phenomenological Research, 70, 1–33.
  • DeRose, Keith (1992). Descartes, Epistemic Principles, Epistemic Circularity, and ScientiaPacific Philosophical Quarterly, 73(3), 220–238.
  • DeRose, Keith (2002). Review of Timothy Williamson’s Knowledge and its Limits. The British Journal for the Philosophy of Science, 53(4), 573–577.
  • Donagan, Alan (1989). Spinoza. Chicago University Press. 
  • Dubouclez, Olivier (2017). Attention et simultanéité intellectuelle chez Descartes, Clauberg et Spinoza. Les Études philosophiques, 171, 27–42.
  • Fernandez, Jordi (2013). Transparent Minds: A Study of Self-Knowledge. Oxford University Press.
  • Frankfurt, Harry (1970). Demons, Dreamers, and Madmen: The Defense of Reason in Descartes’ Meditations. Princeton University Press.
  • Gertler, Brie (2011). Self-Knowledge. Routledge.
  • Gertler, Brie (2015). Self-Knowledge. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2015 ed.). Retrieved from https://plato.stanford.edu/entries/self-knowledge/
  • Gewirth, Alan (1943). Clearness and Distinctness in Descartes. Philosophy, 18(69), 17–36.
  • Hatfield, Gary (2009). Attention in Early Scientific Psychology. In Perception and Cognition: Essays in the Philosophy of Psychology (388–408). Clarendon Press.
  • Hatfield, Gary (2011). Transparency of Mind: The Contributions of Descartes, Leibniz, and Berkeley to the Genesis of the Modern Subject. In Hubertus Busche (Ed.), Departure for Modern Europe: A Handbook of Early Modern Philosophy (1400-1700) (361–375). Felix Meiner Verlag.
  • Hatfield, Gary (2017). L’attention chez Descartes: aspect mental et aspect physiologique. Les Études Philosophiques, 120(1), 7–25.
  • Heil, John (2013). Philosophy of Mind: A Contemporary Introduction (3rd ed.). Routledge.
  • Hennig, Boris (2004). Was bedeutet ‚Conscientia‘ bei Descartes? (Doctoral dissertation). Universität Leipzig.
  • Hennig, Boris (2006). Conscientia bei Descartes. Zeitschrift für Philosophische Forschung, 60(1), 21–36.
  • Hennig, Boris (2007). Cartesian conscientiaBritish Journal for the History of Philosophy, 15(3), 455–484.
  • Hume, David (2000). A Treatise of Human Nature. David Fate Norton and Mary J. Norton. (Eds.). Oxford University Press.
  • Jolley, Nicholas (2010). Scientia and Self-knowledge in Descartes. In Tom Sorrell, G. A. J. Rogers, and Jill Kraye (Eds.), Scientia in Early Modern Philosophy: Seventeenth-Century Thinkers on Demonstrative Knowledge from First Principles (83–88). Springer.
  • Kenny, Anthony (1966). Cartesian Privacy. In G. Pitcher (Ed.), Wittgenstein: The Philosophical Investigations (352–370). Palgrave Macmillan.
  • Kenny, Anthony (1968). Descartes: A Study of His Philosophy. Random House.
  • Kim, Jaegwon (2010). Philosophy of Mind (3rd ed.). Westview Press.
  • Lähteenmäki, Vili (2007). Orders of Consciousness and Forms of Reflexivity in Descartes. In Sara Heinämaa, Vili Lähteenmäki, and Pauliina Remes (Eds.), Consciousness: From Perception to Reflection in the History of Philosophy (177–201). Springer.
  • Leijenhorst, Cees (2007). Cajetan and Suarez on Agent Sense: Metaphysics and Epistemology in Late Aristotelian Thought. In Henrik Lagerlund (Ed.), Forming the Mind. Essays on the Internal Senses and the Mind/Body Problem from Avicenna to the Medical Enlightenment (237–262). Springer.
  • Lichtenberg, Georg Christoph (1990). Aphorisms (R. J. Hollingdale, Trans.). Penguin.
  • Macdonald, Cynthia (2009). Introspection. In Brian McLauglin, Angsar Beckermann, and Sven Walter (Eds.), The Oxford Handbook of Philosophy of Mind (741–766). Oxford University Press.
  • Markie, Peter (1992). The Cogito and Its Importance. In John Cottingham (Ed.), The Cambridge Companion to Descartes (140–173). Cambridge University Press.
  • Matthews, Gareth B. (1977). Consciousness and Life. Philosophy, 52, 13–26.
  • McCann, Edwin (1986). Cartesian Selves and Lockean Substances. The Monist69(3), 458–482.
  • McRae, Robert (1972a). Descartes’s Definition of Thought. In Ronald Joseph Butler (Ed.), Cartesian Studies (55–70). Basil Blackwell.
  • McRae, Robert (1972b). Innate Ideas. In Ronald Joseph Butler (Ed.), Cartesian Studies (32–54). Basil Blackwell.
  • Morris, Katherine (2000). Bêtes-machines. In Stephen Gaukroger, John Schuster, and John Sutton (Eds.), Descartes’ Natural Philosophy (401–419). Routledge.
  • Nelson, Alan (1997). Descartes’s Ontology of Thought. Topoi, 16(2), 163–178.
  • Newman, Lex (2004). Rocking the Foundations of Cartesian Knowledge: Critical Notice of Janet Broughton, “Descartes’s Method of Doubt”. The Philosophical Review, 113(1), 101–125.
  • Newman, Lex (2016). Descartes’ Epistemology. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Winter 2016 ed.). Retrieved from https://plato.stanford.edu/entries/descartes-epistemology/
  • Newman, Lex and Alan Nelson (1999). Circumventing Cartesian Circles. Noûs, 33(3), 370–404.
  • Nolan, Lawrence (1997a). The Ontological Status of Cartesian Natures. Pacific Philosophical Quarterly, 78, 169–194.
  • Nolan, Lawrence (1997b). Reductionism and Nominalism in Descartes’s Theory of Attributes. Topoi, 10(2), 129–140.
  • Nuccetelli, Susanna (2003). Introduction. In Susana Nuccetelli (Ed.), New Essays on Semantic Externalism and Self-Knowledge (1–21). The MIT Press.
  • Para du Phanjas, L’abbé François (1779). Théorie des êtres insensibles, ou Cours complet de métaphysique, sacrée et profane, mise à la portée de tout le monde. L. Cellot & A. Jombert.
  • Paul, Elliot S. (2005). Sum Res Cogitans. In Nathan D. Smith and Jason P. Taylor (Eds.), Descartes and Cartesianism (41–54). Cambridge Scholars Press.
  • Paul, Elliot S. (in press). Clarity First: Rethinking Descartes’s Epistemology. Oxford University Press.
  • Paul, Sarah K. (2014). The Transparency of Mind. Philosophy Compass, 9(5), 295–303.
  • Peacocke, Christopher (2012a). Descartes Defended. Aristotelian Society Supplementary Volume, 86(1), 109–125.
  • Peacocke, Christopher (2012b). First Person Illusions: Are They Descartes’, or Kant’s? Philosophical Perspectives26(1), 247–275.
  • Pessin, Andrew (2009). Mental Transparency, Direct Sensation, and the Unity of the Cartesian Mind. In J. Miller (ed.), Topics in Early Modern Philosophy of Mind, Studies in the History of Philosophy of Mind 9 (1–37). Springer.
  • Radner, Daisie (1988). Thought and Consciousness in Descartes. Journal of the History of Philosophy, 26(3), 439–452.
  • Reynolds, Steven L. (1992). Descartes and First Person Authority. History of Philosophy Quarterly, 9(2), 181–189.
  • Rickless, Samuel (2005). The Cartesian Fallacy Fallacy. Nous, 39(2), 309–336.
  • Rodis-Lewis, Geneiviève (1950). Le problème de l’inconscient et le cartésianisme. Presses Universitaires de France.
  • Rodis-Lewis, Geneiviève (1963). Nicolas Malebranche. Presses Universitaires de France.
  • Rorty, Richard (1980). Philosophy and the Mirror of Nature. Princeton University Press.
  • Rozemond, Marleen (1995). Descartes's Case for Dualism. Journal of the History of Philosophy33(1), 29–63.
  • Rozemond, Marleen (1998). Descartes's Dualism. Harvard University Press.
  • Rozemond, Marleen (2006). The Nature of the Mind. In Stephen Gaukroger (Ed.), The Blackwell Guide to Descartes’ Meditations (48–66). Blackwell Publishing.
  • Rozemond, Marleen (2010). Critical Notice of Janet Broughton, Descartes’s Method of Doubt. Canadian Journal of Philosophy, 34(4), 591–613.
  • Russell, Bertrand (1949). A History of Western Philosophy. Simon and Schuster.
  • Ryle, Gilbert (1949). The Concept of Mind. The University of Chicago Press.
  • Schechtman, Anat (2014). Descartes’s Argument for the Existence of the Idea of an Infinite Being. Journal of the History of Philosophy, 52(3), 487–517.
  • Schiffer, Stephen (1976). Descartes on His Essence. Philosophical Review85(1), 21–43.
  • Schmaltz, Tad (1997). Descartes on Innate Ideas, Sensation, and Scholasticism: The Response to Regius. In M. A. Stewar (Ed.), Studies in Seventeenth-Century European Philosophy (33–73). Oxford University Press.
  • Schwitzgebel, Eric (2014). Introspection. In Edward N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy (Spring 2014 ed.). Retrieved from https://plato.stanford.edu/entries/introspection/
  • Shapiro, Lisa (2008). ‘Turn My Will in Completely the Opposite Direction’: Radical Doubt and Descartes’ Account of Free Will’. In (Eds.) Paul Hoffman, David Owen, and Gideon Yaffe, Festschrift for Vere Chappell (21–39). Broadview.
  • Shapiro, Lisa (2012). Cartesian Selves. In Karen Detlefsen (Ed.)Descartes’s Meditations: A Critical Guide (226–242). Cambridge University Press.
  • Shoemaker, Sydney (1990). First-Person Access. Philosophical Perspectives4, 187–214.
  • Sievert, Donald (1975). Descartes’s Self-Doubt. The Philosophical Review, 84(1), 51–69.
  • Simmons, Alison (2003). Descartes on the Cognitive Structure of Sensory Experience. Philosophy and Phenomenological Research, 67(3), 549–579.
  • Simmons, Alison (2012). Cartesian Consciousness Reconsidered. Philosopher’s Imprint, 12(2), 1–21.
  • Smithies, Declan (2012). Mentalism and Epistemic Transparency. Australasian Journal of Philosophy, 90(4), 723-741.
  • Sorrell, Tom (2010). Scientia and the Sciences in Descartes. In Tom Sorrell, G. A. J. Rogers, and Jill Kraye (Eds.), Scientia in Early Modern Philosophy: Seventeenth-Century Thinkers on Demonstrative Knowledge from First Principles (71–82). Springer.
  • Strandberg, Hugo (2015). Self-Knowledge and Self-Deception. Palgrave-MacMillan.
  • Thiel, Udo (1994). Hume’s Notions of Consciousness and Reflection in Context. British Journal for the History of Philosophy, 2(2), 75–115.
  • Thiel, Udo (2011). The Early Modern Subject: Self-Consciousness and Personal Identity from Descartes to Hume. Oxford University Press.
  • van Cleve, James (1983). Conceivability and the Cartesian Argument for Dualism. Pacific Philosophical Quarterly64(1), 35–45.
  • Vinci, Thomas C. (1998). Cartesian Truth. Oxford University Press.
  • Williams, Bernard (1978). Descartes: The Project of Pure Enquiry. Penguin.
  • Williamson, Timothy (1996). Cognitive Homelessness. Journal of Philosophy, 93(11), 554–573.
  • Williamson, Timothy (2000). Knowledge and its Limits. Oxford University Press.
  • Wilson, Catherine (2003). Descartes’s Meditations: An Introduction. Cambridge University Press.
  • Wilson, Margaret D. (1976). Descartes: The Epistemological Argument for Mind-Body Distinctness. Noûs10(1), 3–15.
  • Wilson, Margaret D. (1978). Descartes. Routledge & Kegan Paul.
  • Wright, Crispin (1998). Self-Knowledge: The Wittgensteinian Legacy. In Crispin Wright, Barry Smith, and Cynthia Macdonald (Eds.), Knowing Our Own Minds (13–46). Oxford University Press.
  • Yablo, Stephen (1990). The Real Distinction Between Mind and Body. Canadian Journal of Philosophy 20(Suppl. 1), 149–201.


    1. In citing Descartes, I generally refer (by volume and page number) to The Philosophical Writings of Descartes, 3 vols. (edited by Cottingham, Stoothoff, Murdoch, and in the case of the 3rd volume, Kenny), followed by Adam and Tannery’s (AT) edition of the original. I use (*) to indicate when I have altered the translation or provided my own. I use () to indicate when I have added italics or boldface.return to text

    2. Descartes holds that the content of self-knowledge—like the content of any mental representation—can be specified in either of two ways. We can specify it propositionally: e.g., “I have certain knowledge that I am thinking about ducks.” Or we can specify it objectually: “I have certain knowledge of my thought.” In Descartes’s view, the difference between these two formulations is merely a verbal one: we may choose to speak either way depending on the purpose at hand (AT 3:417–418; cf. AT 3:395). (This point is widely noted; see, e.g., Gewirth 1943: Footnote 5; Nelson 1997: Footnote 23; Simmons 2012: Footnote 17.) While the objectual formulation is often convenient, we need the propositional formulation to specify which proposition(s) are known in a given case of self-knowledge.return to text

    3. Descartes uses the statements “I exist” (existo/j’existe) and “I am” (sum/je suis) interchangeably. When a proposition is meant to be the content of a thought, I put it in italics. So, for example, I exist is the content of the thought one would express by saying “I exist,” with the indexical “I” referring not to me but to the speaker or thinker in question. Of course, I also use italics in conventional ways, for emphasis, etc.return to text

    4. As is often noted, the premise entails the conclusion only if it affirms the first-person possession of the thought; the minimal thought-claim that a thought exists would not do the trick, which may be one reason Descartes never asserts it on its own.return to text

    5. See Alanen (1981) and Markie (1992).return to text

    6. It is an instance of the general principle, Necessarily, if x has a property, then x exists. See, Pr. i.7, 10–11, 1:195–196/AT 8A:7–8; 2R, 2:100/AT 7:140.return to text

    7. Frankfurt reconstructs the cogito of the Meditations as a reductio argument for the incorrigibility of I exist. He rightly concedes that, on his reading, the cogito “does not consist in a proof of sum” (1970: 106; compare pages 93, 101, and 105). One might argue that this is a problem for his reading. Citing Frankfurt as inspiration, Broughton (2002, 2008) and Curley (1978, 2006) offer similar readings; see Footnote 18 below.return to text

    8. For different readings of the relevant passage, see E. Paul (2005) and M. Wilson (1976; 1978: 74).return to text

    9. For different readings of the real distinction argument, see Alanen (2003), Almog (2001), Clarke (2003), Rozemond (1995; 1998), Schiffer (1976), van Cleve (1983), and Yablo (1990).return to text

    10. Another claim I will not discuss here is, I persist through time. For insightful discussions of how Descartes handles that claim, see McCann (1986) and Shapiro (2012).return to text

    11. This commitment may seem to conflict with Descartes’s positing of innate ideas, stored memories, hidden sensory processes, and more. Simmons carefully resolves these conflicts (2012). For more on innate ideas, in particular, see Schmaltz (1997).return to text

    12. Transparency-through-Having is widely attributed to Descartes both by historians of philosophy and contemporary philosophers. Among historians, an early, influential proponent of this (mis)reading is Anthony Kenny, who writes,

      Descartes makes it true by definition that if I think, I know that I think. It is here that the indubitability of the premise of ‘cogito ergo sum’ is to be found. […] [I]t is not just the occurrence of thought that cannot be doubted, but the occurrence of the particular thought in question. (1968: 49; cf. Kenny 1966)

      Compare Alanen (1981: 11), Bennett (2001: 65–66, 107–108), Cottingham (1986: 41), Donagan (1989), Markie (1992: 145), Matthews (1977: 13–26), Williams (1978: 80), and C. Wilson (2003: 69). Adopting a phrase from Ryle (see below), Margaret Wilson says that “Descartes holds all our mental acts are self-luminous” (1978: 161), meaning that “[t]here is nothing in my mind of which I don’t have certain and indubitable knowledge” and that “there is never any internal or intrinsic feature of my own mental states of which I am ignorant” (1978: 153–154). She adds, however, that “(mere) luminousness corresponds to implicit knowledge” (1978: 161), whereas explicit knowledge of one’s thoughts requires introspection (more on this below in Footnote 14).

      In contemporary epistemology and philosophy of mind, it is almost de rigeur to define one’s views in relation to the supposed views of Descartes (see Cassam 2008). This practice gained special prominence with anti-Cartesian polemic of Gilbert Ryle, who says,

      Philosophers, chiefly since Descartes,” have held that [t]he states and operations of a mind are states and operations of which it is necessarily aware, in some sense of ‘aware’, and this awareness is incapable of being delusive. […] If I think, hope, remember, will, regret, hear a noise, or feel a pain, I must, ipso facto, know that I do so. (1949: 152)

      Ryle identifies the “awareness” in question as consciousness, which is he says is supposed to make thought “self-luminous” (1949: 153). Like Margaret Wilson, Timothy Williamson echoes Ryle’s language and says that, for Descartes, thoughts are “luminous” in the sense that “nothing is hidden from us” within our own minds (1996: 554). In a recent textbook, John Heil teaches that “Descartes can be taken to embrace […] Transparency: if you are in a particular state of mind, you know you are in that state” (2013: 19–20, original italics, my boldface). Compare Boghossian (2008: 140), Rorty (1980: 97), Shoemaker (1990: 187), and Strandberg (2015: 59). Some theorists say Transparency-through-Having is “Cartesian” without claiming to speak for the historical Descartes: e.g., Carruthers (2011: 13–14), Kim (2010: 160–161), S. Paul (2014: 295).

      While Descartes is concerned with metaphysically certain self-knowledge, contemporary theorists often seem to be concerned with ordinary self-knowledge instead. I address this complication at the end of Section 3 below.return to text

    13. See Alanen (2003: 54–56, 99–100), Broughton (2008: 192), Hatfield (2011), McRae (1972), Radner (1988), Rozemond (2006), Simmons (2012).return to text

    14. Among historians of philosophy, Gary Hatfield says that, for Descartes, “if we concentrate on [our] thoughts, we would have indubitable knowledge of them” (2011: 365). Donald Sievert says that Descartes “‘observes’ (introspects) his own occurrent mental acts and […] does not doubt, and cannot imagine doubting, the existence of what he observes” (1975: 58). Compare McRae (1972a; 1972b) and Reynolds (1992). (More on Reynolds in Footnote 23 below.)

      Margaret Wilson is a more complex case, since her interpretation includes both versions of Transparency, for two levels of knowledge: implicit and explicit. We noted above (Footnote 12) that she ascribes to Descartes a qualified version of Transparency-through-Having, according to which having a thought guarantees consciousness and, with it, implicit knowledge of that thought. She also ascribes to him a qualified version of Transparency-through-Introspection, according to which introspecting a thought guarantees explicit knowledge of that thought. She says that, in Descartes’s view, “to form explicit knowledge of ‘what is in us’ we must reflect on our thoughts and ideas. To ‘reflect on them’ is evidently the same as to attend to them, or turn the ‘eye of the mind’ to them.” Further, she says, directing such “reflective (or attentive) consciousness” towards one’s thoughts is not only necessary but also sufficient—it’s “all that’s needed”—for acquiring explicit knowledge of one’s thoughts (M. Wilson 1978: 161–162). Descartes does distinguish between implicit and explicit knowledge (AT 5:147); but as Radner (1988) has argued, it is questionable whether Wilson interprets that distinction correctly.

      Another complication in Wilson’s reading is that she charges Descartes with inconsistency. She herself amasses textual evidence against her Transparency reading (M. Wilson 1978: 150–165). But she needs to ascribe Transparency to Descartes as part of her “naïve interpretation” of the cogito, so she portrays him as contradicting himself. On her reading, what Descartes says against Transparency is in “conflict – or at any rate in tension – with [his] doctrine of the epistemological transparency of thought or mind” (1978: 50). Unlike Wilson, Edwin Curley does not deploy Transparency in his reading of the cogito, but like Wilson, he suggests that Descartes both accepts and denies Transparency, and cannot “be accused of a foolish consistency” (1978: 181). I agree with Anat Schechtman (2014: 493 Footnote 17) that the inconsistency is a problem for the Transparency reading of Descartes, not for Descartes.

      Among contemporary philosophers, Gilbert Ryle is like Wilson in that he attributes both forms of Transparency to Descartes. On his reading, then, the Cartesian “mind has a twofold Privileged Access to its own doings” (1950: 149)—firstly through consciousness and secondly through introspection. But unlike Wilson, Ryle does not distinguish between two levels of self-knowledge (implicit and explicit) resulting from consciousness and introspection; and so, as Curley (1978: 172) observes, the first seems to render the second otiose. Peter Carruthers identifies both versions of Transparency as “Cartesian,” without claiming to speak for the historical Descartes. He formulates Transparency-through-Introspection as follows: “Perhaps one only has to direct one’s attention in the right sort of way [i.e., inward] to acquire authoritative knowledge of [one’s thoughts]” (2011: 14). Crispin Wright speaks of the “Cartesian” view that “inner observation” is a form of “immediate awareness that is “infallible and all-seeing” for “phenomenal avowals” and “very, very reliable” for “basic attitudinal avowals” (1998: 22–23). According to Brie Gertler, Descartes argues that “introspective knowledge is peculiarly direct and therefore especially secure” (2011: 87) and that “so long as you are carefully attending to your own thoughts, nothing […] can render misleading your evidence that you are thinking” (2015).

      As with the previous form of Transparency (see Footnote 12 above), Descartes is concerned with metaphysically certain self-knowledge, while contemporary theorists often seem to be concerned with ordinary self-knowledge instead. See the end of Section 2 below.return to text

    15. Simmons’s topic is psychological transparency while mine is epistemological transparency; but my understanding of Cartesian consciousness is indebted to her excellent paper, and to the work of other scholars whom she and I both cite. Of particular note is the pioneering work of Geneviève Rodis-Lewis (1950; 1963).return to text

    16. In claiming that, for Descartes, consciousness is a same-order, reflexive feature of each thought itself, I concur with Alanen (1992: 24–25; 2003: 99–102), Aquila (1988: 546–547), Beyssade (1979: 244–249), Hennig (2004; 2006; 2007: 482), Radner (1988), Pessin (2009: 4–5), Simmons (2012), and Vinci (1998: 39ff.). However, some commentators propose that, for Descartes, basic consciousness is a higher-order perception (i.e., reflection): e.g., Cottingham (1978: 211–214), Morris (2000: 403–404), and Thiel (1994: 90–92; 2011). For a critique of Thiel, see Barth (2013; 2016). There is also a hybrid reading, according to which basic consciousness of perceptions is a same-order phenomenon, while consciousness of acts of will is a higher-order phenomenon; see Barth (2011a; 2011b) and Lähteenmäki (2007). Though I cannot settle this debate here, we should note that by assuming the same order-view of basic consciousness I am making the Transparency-through-Introspection reading more defensible. For if Descartes were to hold the alternative view, that basic consciousness is higher-order reflection, then there would be no difference between Transparency-through-Having and Transparency-through-Introspection, and the counterexamples we’ve already seen against former would also count against the latter.return to text

    17. Simmons (2012) makes this point.return to text

    18. Janet Broughton (2002: Chapter 7; 2008) is a prominent example. As she acknowledges, her alternative to Transparency—which treats I exist and I am thinking as “conditions of doubt”—goes back to Curley (1978; cf. Curley 2006) and Frankfurt (1970). (Recall Footnote 7 above.) For critiques of this alternative, see Newman (2004) and Rozemond (2010).return to text

    19. Indubitability: e.g., “The nature of my mind is such that I cannot but assent to these things, at least so long as I clearly perceive them” (M5, 2:45/AT 7:65). Infallibility: e.g., “Everything which I clearly and distinctly perceive is of necessity true” (M5, 2:48/AT 7:70).return to text

    20. Samuel Rickless (2005) helpfully gathers texts where Descartes accords some degree of clarity and distinctness to sense-perception as well as imagination. Compare Cunning (2010), Nelson (1997), Newman (2016), Simmons (2003). I concur with Rickless that, for Descartes, clear and distinct intellection is necessary and sufficient for certain knowledge (cognitio).return to text

    21. Descartes also explains, however, that clear and distinct intellection is not sufficient for the highest epistemic achievement, scientia, which he thinks one cannot have without certain knowledge of God (2R, 2:101/AT 7:141*). On Descartes’s distinction between cognitio and scientia, see Carriero (2008b), Cottingham (1986), Della Rocca (2005), DeRose (1992), Jolley (2010), Newman and Nelson (1999), and Sorrell (2010).return to text

    22. For more on Descartes's Rationalism, see Newman (2005).return to text

    23. Steven Reynolds may seem to be an exception. He rightly argues that, for Descartes, “certainty about one’s own mental states” comes through clear and distinct perception, not thanks to “first person authority”—his term for Transparency-through-Having (Reynolds 1992: 181). But when he speculates about what is required for clearly and distinctly perceiving one’s own thoughts, he says, “presumably by attending to them in the appropriate way” (1992: 186), with no mention of the need for doubt—which is to revert to Transparency-through-Introspection.return to text

    24. This assumption is encouraged by the standard English translation (CSM), which erases Descartes’s distinction between “consciousness” and “knowledge” (conscientia and cognitio) by rendering both terms with a single English word: “awareness.” Cottingham rendered ‘conscius’ as ‘conscious’ in his early translation of Descartes’s Conversation with Burman (1976), but he and his co-editors changed it to ‘aware’ for CSM, published in 1984 and 1985 (see, e.g., CSM 3:235/AT 5:159). In a work he published around the same time as CSM, Cottingham seems to say that Descartes endorses Transparency-through-Having, and does so across the board: “Whatever mental act the meditator engages upon, there will always be a narrowly intellectual element involved – an act of reflective awareness. And such reflection will be indubitable […]” (1986: 41). (Cottingham identifies consciousness with reflection.) But in later work, he says that Descartes is committed to such “transparency” for intellectual thoughts and decisions, whereas the contents of sensory thoughts are “opaque” to us (2006: 181–182).return to text

    25. This assumption is usually implicit in the literature, but it is sometimes explicit, for example when Broughton asserts that “Descartes plainly uses ‘natural light’ [the faculty of intellectual perception] to mean ‘faculty of distinct perception’” (1984: 607 Footnote 16).return to text

    26. I attempt to offer such an account elsewhere (E. Paul in press).return to text

    27. This is how Descartes describes the effect of such confusion in a letter we’ll examine below (To [Silhon], March or April 1648, 3:331/AT 5:136–137).return to text

    28. David Cunning (2007; 2010) is excellent on this point.return to text

    29. Likewise, Descartes writes in the Passions of “that special state of reflection and attention which our will can always impose upon our understanding when we judge the matter before us to be worth serious consideration” (ii.76, 1:355/AT 11:385). But some qualifications are in order. Infants are not yet capable of reflection (3:254/AT 5:192–193; 3:356–357/AT 5:220). Even for adults, reflection can be difficult in certain conditions such as sleep (2:248/AT 7:358), illness (3:189–190/AT 3:424), or emotional agitation (1:339/AT 11:350). Our attention can be seized against our will, especially in a state of wonder or astonishment at something novel or extraordinary (1:353–356/AT 11:381–386). And it’s difficult to maintain focus on any one thing for long, particularly if that thing is not present to the senses or imagination (Pr. i.73, 2:220/AT 8A:37; M3, 2:32/AT 7:47). For more on Descartes’s treatment of attention, see Barrier (2017), Brown (2007), Dubouclez (2017), Hatfield (2009; 2017), and Leijenhorst (2017).return to text

    30. See Rozemond (1996: 27–34) and B. Williams (1978: 54) for more on why the Dreaming Argument doesn’t yield radical doubt. The most extensive case for the opposing view is due to Carriero (1987: 230–239; 1999). Carriero is focused specifically on the version of the Dreaming Argument that appears in the Meditations, where it is uniquely supplemented with the painting analogy, which is crucial to Carriero’s reading. So, even if we grant to Carriero that the Dreaming Argument of the Meditations yields radical doubt, we could still acknowledge, as Descartes indicates above, that the version of the Discourse does not.return to text

    31. Lichtenburg’s dig was popularized by Russell (1946: 567). In a brilliant exchange, Peacocke (2012a) defends Descartes, while Campbell (2012) deepens the critique. See also Peacocke 2012b.return to text

    32. Treatise, to text

    33. Descartes does not use the word “therefore” here, but that’s beside the point. The meditator is certain that the premise is true, and she is certain that if the premise is true, then she exists. It is a basic feature of the rationality built into the nature of the human mind that, in general, when you are certain that p and you are certain that if p then q, you automatically draw the inference and become certain of q. Failing to infer q would be obtuse in the manner of Lewis Carroll’s Tortoise (Carroll 1895), and Descartes’s meditator is a normal, rational human being, not Carroll’s Tortoise. The fact that the meditator does infer that she exists is plainly confirmed by other passages looking back at this one, where inferential terms—e.g., “therefore” and “it follows that”—are used explicitly. One of those passages, from Meditation Four, is cited above: 2:41/AT 7:58; compare 2R, 2:100/AT 7:140; 7R, 2:323/AT 7:479–480; 5R, 2:244/AT 7:354; and 6R, 2:285/AT 7:422.return to text

    34. Shapiro (2008) and Boehm (2014) also highlight the importance of doubt in the meditator’s cogito, but for them the role doubt plays is to demonstrate the freedom of the will. I elaborate on my reading of the mediator's cogito in Paul (in press).return to text

    35. 2:15–61/AT 7:220–30; 2: 276–7/AT 8A:215–6; 3:236/AT 4:120. For more on this distinction, see Nolan (1997a; 1997b).return to text

    36. Williamson (1996; 2000: Chapter 4) levels his own critique of what he calls “transparency”, though he defines it somewhat differently that I have defined “Transparency,” in part because he is concerned with ordinary rather than certain knowledge. For responses to Williamson, see Berker (2008), DeRose (2002), and Smithies (2012).return to text