A Look at the Cogito: Cogito Ergo Sum (Column 363)

With God’s help

Disclaimer: This post was translated from Hebrew using AI (ChatGPT 5 Thinking), so there may be inaccuracies or nuances lost. If something seems unclear, please refer to the Hebrew original or contact us for clarification.

To my dear granddaughter, Uriya,

who at the age of four uses the term “dilemma”

I’ve been asked many times about Descartes’ cogito argument (see for example here), and the time has come to address it systematically, as far as a column like this allows.[1]

The Background to Descartes’ Cogito Argument

René Descartes (Cartesius, in his Latin name) was a philosopher, scientist, and mathematician, a Frenchman of the 16th–17th centuries. Of all these, his chief renown came from his classical philosophical consideration, the cogito, one of whose Hebrew formulations is: “I think, therefore I exist.”

The background is a waning of the rationalist approach that reigned from antiquity until the dawn of the modern era. According to this approach, reason is the central—and certainly legitimate—tool for knowing the world. In the fifteenth century an empiricist critique of rationalism began to take shape; that is, thinkers began to doubt that reason could serve as a reliable instrument for knowing the world. The fact that we think in a certain way doesn’t mean that the world itself runs that way, for it attests only to the structure of our thought and not necessarily to the way reality itself functions. As Mark Twain said: the world doesn’t owe you anything; it was here before you. Aristotle thought that bodies fall to earth at a speed proportional to their mass (weight), and he didn’t bother to conduct a simple experiment to test it. It was obvious to him that if his logic says so, then that is exactly how reality operates (well, he didn’t yet know Mark Twain). Empiricism held that conclusions of reason cannot be a basis for claims about the world. Insights and claims about the world must be solely the result of observation. Following this philosophical revolution, empiricism supplanted rationalism, and modern science was born. The most fundamental principle that distinguishes it from ancient science is reliance on observations (at the same time, the separation—now so self-evident—between science and philosophy took place).

Descartes, who lived precisely during the period when this shift occurred, tried to halt that process. He was an ardent rationalist, and therefore desperately sought a solid rational foundation for philosophy and science. It is important to understand that he believed in observations and was certainly a man of science, but to the same extent he believed in the supremacy of reason and its being a reliable and effective instrument for knowing the world. In his view, reliance on observation must itself be grounded in reason. Therefore, the most basic foundation of our thinking and our science, he thought, should be reason. Observation can come after it and be based upon it.

Here I’ll just note that the common notion that modern science is an expression of pure empiricism is mistaken. Already the early empiricists (especially David Hume) understood that quite a few of the assumptions on which science rests are not drawn from experience but from reason. Science is indeed based on observations, but not on them alone. Reliance on observations alone would not yield even a single law of nature. Observations provide particular facts (we can know only the specific fact we observed). Laws of nature are general claims that concern infinitely many facts, and as such they cannot be the product of observations alone. The process always involves a generalization from particular facts to a general law, and in these generalizations analytic, a priori modes of reasoning—i.e., thought procedures—are involved. Rationalism thus returned through the back door.

As a rationalist, Descartes was also deeply troubled by the lack of certainty in philosophy and science; that is, by the fact that almost any claim (and certainly factual claims) can be accepted or rejected without contradiction. This means that our knowledge is not certain. It was clear to him—something empiricists tend to ignore to this day—that observation too is not a certain source of facts. The senses can deceive us, and we have no guarantee that sensory data correctly reflect reality itself. This, in itself, is a working assumption of a rationalist character (meaning it has no empirical basis). Beyond that, as I already mentioned, the move from observational facts to general laws of nature involves thought processes. The foundation of any philosophical, theological, and scientific system seemed to Descartes shaky. Descartes lived and worked before the flowering of British empiricism (Locke, Hume, and Berkeley), which in large measure was born as a reaction to his (rationalist) doctrine, and in fact he sought to prevent its emergence and dominance.

From this we can understand that in order to find an Archimedean and certain foothold for his philosophy, Descartes could not use observation. He had to look for a claim that stands on its own feet—i.e., one not based on observation—and yet (indeed, because of that) could not be doubted. The trajectory of his reasoning is described in several books he wrote, and in each of them the path is slightly different, and so is the wording of the conclusion. Here I’ll outline the essential lines of the move and focus on two formulations that appear in his writings.

Descartes’ Skeptical Move

Descartes began his philosophical inquiry by casting radical doubt on anything that could be doubted. Contrary to what many think, he was certainly not a skeptic. It was a methodological doubt, whose function was to clear the ground of anything that could be doubted. He adopted a methodological assumption, for discussion’s sake, that any claim that can be doubted is false, with the aim of seeing whether, at the end of this process, something still remains. If something remains, it will be a claim that cannot be doubted and will, of course, not be the product of observation (for he doubted any observational finding precisely as such). On it Descartes hoped to base all other claims—philosophical and observational alike—i.e., science.

At the stage of methodological doubt he was unwilling to accept anything, not even claims that seem self-evident to us. He raised the possibility that there is a demon deceiving him, or perhaps mental illnesses causing him to reach wrong conclusions. At this skeptical stage he was not even willing to accept that what he sees actually exists. He was also unwilling to accept the claim that God exists (which for him was self-evident), and so on. So what could remain after such sweeping doubt? Can we even think of something that would pass such stringent tests?

The Result: The Cogito Principle

At the end of the process, Descartes reached the surprising conclusion that there is indeed such a claim. There is a claim that cannot be eliminated in any way, i.e., it cannot be doubted. His contention is that his very own existence cannot be doubted by him. The reasoning that led him to this is his well-known cogito argument: “I think, therefore I exist.” This is the wording in his Discourse on the Method[2] (p. 48). In Latin the formulation is: Cogiti Ergo Sum, and so this argument is called “the cogito.” In another place in his writings[3] a different wording appears: “I doubt, therefore I exist.” To understand the argument and its significance, let us spell out the inference a bit more.

If I doubt my very existence, that indicates there is someone who is entertaining that doubt (or: someone thinking it). It is not possible to say “I doubt my existence.” If I doubt, then I (the one who entertains the doubt) exist, and therefore the doubt dissolves. Descartes is essentially asserting that my ability to doubt is itself proof of my existence. Thus it follows that although I can doubt almost everything, I cannot doubt my own existence.

What, in fact, Did the Cogito Introduce?

As many of his commentators have already explained,[4] this is a partial and imprecise presentation of the argument. Any logical argument bases its conclusion on a premise or premises, and if we reject one of the premises, we are not obliged to accept the conclusion. In our example, if someone were to say “it is not true that I doubt,” or “it is not true that I think,” then he is exempt from the conclusion “I exist.” Our attitude toward the conclusion depends on the truth of the premise.

If this argument were based on the premise “I doubt” or “I think,” as one might gather, then there are countless parallel proofs, such as: I walk, therefore I exist. Clearly, from the premise “I walk” the conclusion “I exist” follows no less necessarily than from the premise “I think.” For if I do not exist, how can I walk?! So what is special about the premise Descartes chose? Why did he decide to derive his conclusion specifically from the premise “I think,” and not from the premise “I walk”?

To understand this, we must note that in an ordinary logical argument the truth of the conclusion is not unconditional. It depends on the truth of the premise. Even if the argument is valid (i.e., the conclusion necessarily follows from the premises), that still does not suffice to prove the conclusion with certainty. The reason is that the premises can be doubtful. True, if I walk then clearly I exist, but why assume that I walk? Is that not subject to doubt? I remind you of Descartes’ skeptical assumption that even our most basic sensory data are inadmissible. Knowing that I walk is the result of observation, and as such it is not a sufficient basis for a certain argument that grounds all our thinking in reason. To arrive at a certain conclusion, we must use certain premises, or alternatively, devise an argument without premises. That sounds very odd, but that is precisely what Descartes sought in order to attain the coveted certainty.

Here lies the sting of the cogito argument. Descartes’ principal innovation was not the logical derivation of the conclusion “I exist” from the premise “I think,” but the insight that “I think” itself (the premise of the argument) is a necessary claim, as opposed to the premise “I walk.” If I think that I do not think, that too is a thought, and therefore I am again thinking. In other words, “I think” is a necessary proposition that cannot be denied. Its denial is itself proof of its existence. This, of course, cannot be said of the proposition “I walk.” If I think that I am not walking, or if I doubt it, that does not lead to any contradiction. Therefore there is no necessity in the claim “I walk” (in philosophical parlance, it is a contingent claim). For this reason, Descartes’ cogito derives the conclusion “I exist” specifically from the premise “I think.” An argument based on “I walk” will not necessarily persuade us that we exist, since we can always deny the premise of the argument. By contrast, one cannot deny the premise “I think,” and accordingly we are compelled to adopt the conclusion “I exist.”

In sum, Descartes’ main novelty is not deriving the conclusion “I exist” from the premise “I think,” but the insight that “I think” itself is a necessary claim. The derivation of “I exist” from this premise is simpler and contains little innovation.[5] Indeed, in Descartes’ principal work, the Meditations[6] (p. 42), the cogito is formulated somewhat differently. There he actually proves the necessity of the claim “I think,” and only in passing notes that from this premise the conclusion “I exist” follows. As I have shown here, this is in fact the more precise logical presentation of the cogito argument.

A Note on the Difference Between the Two Formulations

Let us now consider the other formulation of the cogito: “I doubt, therefore I exist.” The premise “I doubt” is not necessary at all; its negation does not lead to a contradiction. And yet the claim “I exist” is necessary, since if I doubt my existence it is still clear that I exist. That is, the formulation that derives “I exist” from the premise “I doubt my existence” is one in which what is certain is the conclusion (“I exist”) rather than the premise. The premise “I doubt” is not necessary at all, and thus this argument appears similar to one that derives existence from the premise “I walk.”

I think the second formulation of the argument is based on the logic of a dilemma (in the Talmudic formulation: mima nafshakh): If I doubt my existence—then I exist (since I am the one who doubts); and if I do not doubt my existence—then again I exist. Therefore, whichever way you look at it, I exist. Thus, in this formulation the structure of the argument differs from that of the better-known formulation that starts from the premise “I think.”

At the end of the process, Descartes concluded that “I exist” is a certain claim, and he saw it as the beginning of the path to all other conclusions and insights, indeed to the rest of his scientific and philosophical system. From here he proceeded to prove the existence of God, our modes of thinking and observation, and of course trust in the senses. In this way he tried to halt the empiricist movement that was based on distrust in the ability to reach claims about the world (to know the world) by means of reason, and that placed observation at the center.

Note: What, Exactly, Exists?

There is another point here that is very easy to miss, and because of it I dealt with the cogito in the prologue to my book Sciences of Freedom. Did Descartes, by his argument, prove the existence of our body? From the structure of the argument it is clear that what is proven here is the existence of the spirit, that is, the thinking part of us, not the body. What this argument proves is that the entity that thinks itself (=the intellect and consciousness)[7] exists. The “I” whose existence we have proven is our spirit, not the body. Even after the cogito, it is still not necessary that my body actually exists.

In other words: if we were to assume that our intellect exists but we have no body, we would not arrive at any contradiction. This is a fully coherent picture, and the cogito principle does not rule it out. If so, the proposition “my body exists” is not necessary. The proposition whose necessity the cogito proves is “my spirit exists.” Descartes himself sensed this clearly in several places in his writings.

In the Meditations, pp. 42–43, there is an even sharper formulation of the cogito:

…I must decide and think that this statement—“I exist, I am present”—is necessarily true whenever I utter it or conceive it in my mind.

We see here that the sentence “I exist” is considered by Descartes necessarily true only whenever he expresses it or conceives it in his mind. Here one sees even more strongly the dependence between thinking and the existence of the thinking entity specifically (and not of the body). Therefore, if we are to be careful with the conclusion of this argument, we should formulate it thus: existence is necessarily true only at those moments when it is being thought. When I engage the question of whether I exist, then there, at that moment, I indeed exist. But in principle it is possible that at other moments I do not. The fact that I reach this conclusion whenever I ask the question does not mean that I always exist, but that I exist precisely at the moments when I ask the question.[8]

Many have inferred from this conclusions about the importance of thinking (only when I think do I truly exist), but these are witticisms (though in my opinion there is much truth in it, but it has no connection whatsoever to the cogito). What can perhaps be inferred is that, contrary to the common conception, the existence of the spirit is clearer and more fundamental than the existence of the body. The spirit apprehends the body, and therefore recognition of its existence precedes recognition of the body’s existence. One can doubt the existence of the body, but one cannot doubt the existence of the spirit. It is important to understand that the accepted view is that it is obvious there is matter in the world, and the only debate (between materialists and dualists) can be over the existence of the spirit. Matter is self-evident, while the spirit is debatable. But the cogito reverses matters and grounds the existence of the material body on the existence of the spirit that apprehends it. Only after the existence of the spirit has been proven—and only then—can we trust the observations it brings to our knowledge, and in particular the existence of the body.[9]

If we widen this surprising conclusion, we can determine more generally that we have no way of knowing whether the chair beside us really exists. What is clear is that the image of the chair present in our cognition (the result of the data that reach us via sight and touch) exists. The conclusion that somewhere out there there is also a “real” chair whose image we are seeing is a speculation whose reliability is not entirely clear. What is immediately present before us is the cognitive image of the chair, not the chair itself.[10] Therefore, one may say that the existence of this image is more certain than the existence of the chair, for a conclusion can never be more certain than the premises that lead to it. Needless to say, the same applies to every material object around us. In particular, the existence of the brain and the physical processes occurring in it is a conclusion drawn from the image formed in the human cognition observing it.

Solipsism (the view that what exists is only what is in our cognition) has always been considered a ridiculous philosophical quibble, whereas materialism is a very respectable and prevalent philosophical and scientific position. But in light of what we have seen here, the philosophical and logical situation is the reverse. It is very odd to think there are materialists who are prepared to deny the existence of a knowing spirit yet accept the content of its cognitions.

The Continuation of the Cartesian Project

We saw that from the certain anchor of “I think,” the cogito argument derives the first conclusion: “I (the intellect that thinks itself) exist.” This is only the start of his rationalist path. From here Descartes continues and builds his entire philosophy on this premise. He derives from it philosophy, theology (the existence of God), and even science—and in particular, the existence of the body. After I reach the conclusion that my spirit and my cognition exist, I am prepared to adopt also the content of my cognition, namely, the material world, and in particular my body, as well as God.

It must be said that the next part of his analysis is much less strong, and it seems that his proofs for the existence of God and of the body and the material world are much weaker than the cogito argument, which deals with the basic substrate of all this. One may conclude that this argument is not very helpful to us, for we cannot in fact continue and build on it all our empirical and philosophical insights. Still, there is an important philosophical significance to this argument, for it seemingly shows the possibility of drawing rationalist conclusions about the world—that is, the possibility of arriving, by reason alone, at cognitions about the world. But as we shall now see, even the cogito argument is not free of problems, and it likely does not truly deliver what Descartes hoped to extract from it.

First Critique: The ‘Ontological’ Character of the Cogito

The cogito is an “ontological argument” (in Kant’s terminology), namely an argument that yields a conclusion containing a claim about the world purely from conceptual analysis. It is an argument without premises, whose conclusion says something about the world (that we exist). It is important to understand that Descartes was well aware that this is the nature of the argument; indeed, he sought precisely an argument of that type. Some have accused him because of the argument’s ontological character, claiming that an argument without premises that yields a conclusion about the world is impossible (this is what Kant and many others claimed about Anselm’s ontological argument for the existence of God; see the first conversation in my book The First Being). But this critique begs the question, for this is precisely the dispute between empiricists and rationalists. Rationalists, like Descartes, claim that reason can indeed supply us with information about the world; empiricists claim it cannot. If so, the claim that it is a rationalist argument cannot serve as an argument in this debate. From the rationalists’ standpoint there is no culpability here (they admit the facts and deny the charge).

Steinitz, in one of his books (I don’t recall where at the moment), makes an interesting claim in this context. Many arguments can derive negative facts from logical contradiction. For example, I can prove that in Berkeley College there is no square dome in the shape of a circle (this is Quine’s example). Alternatively, I can prove that there is no wall that stops all shells and also a shell that penetrates all walls. Here are claims about the world whose basis is purely logical reasoning and conceptual analysis. Why, then, should there not be “ontological” arguments that prove positive claims about the world? If reason can say something negative about the world (prove non-existence), there is no reason it cannot say something positive about it (that I exist, or that God exists). Regarding the Berkeley dome, one might say that “a dome whose shape is a circular triangle” is an empty expression (meaningless), and therefore it is not true that we learned something about the world from it. It is like an argument proving that there is no pedgaliskhok in the world (what is that?). By contrast, the argument about the shell and the wall does appear to be a claim that says something about the world. So too the claim that there will never be a basket with three oranges that later—without any orange leaving or entering it—contains five. This claim also has a clear meaning, and it asserts something about the world by virtue of an a priori determination. If so, Steinitz argues, all of us use reason to derive from it conclusions about what happens or exists in the world. Therefore, in his opinion, ontological arguments are indeed possible.

But even if he is right in principle, at least in the case of the cogito I think Descartes’ project fails, as we shall now see.

Critique of the Second Formulation

Consider the following claim: “The current king of France is bald.” To check whether it is true or not, we must scan the set of bald people and the set of non-bald people and see in which of the two we find the king of France. When we do so, we discover that he is in neither of the sets. The conclusion is that the aforesaid claim is not true, but also the opposite claim, “The current king of France is not bald,” is not true. How can this miracle occur? It seems to contradict the law of the excluded middle (either X is true or it is not true, and there is no third option). But of course the explanation is simple: in today’s France there is a presidential regime, not a monarchy. There is no current king of France, and therefore it is no wonder we don’t find him in either set. We learn to our surprise that dilemma arguments are not as crushing as they appear at first glance.

Now let us consider the second formulation of the cogito, which is built as a dilemma:

If I doubt my existence—then I exist (for I am the one who doubts).

If I do not doubt my existence—then I exist (for my existence is not in doubt).

Conclusion (mima nafshakh): I exist.

You can now easily see the fallacy in the argument. To that end, let us look at the cogito from the perspective of someone who thinks that I do not exist. Descartes wants to persuade him by means of the cogito that he is mistaken. From his perspective, I am not in the set of those who doubt their existence nor in the set of those who do not doubt their existence, simply because there is no such person. Therefore, the dilemma argument presented to him will fail. Such an argument assumes that necessarily either I doubt my existence or I do not, and there is no third option. And therefore, if both doubting leads to my existence and not doubting leads to my existence, then necessarily (whichever way) I exist. But we have now learned that it is not true that one of those two claims must hold (either I doubt or I do not doubt). There is a third possibility: that I do not exist. Then I am neither in the group of those who doubt their existence nor in the group of those who do not doubt their existence.[11]

Critique of the First Formulation

The first formulation is not directly built on mima nafshakh. We saw that it assumes the premise “I think” and infers from it the conclusion “I exist.” As we observed above, the premise “I think” is necessary, for even if I do not think, I think something (that I do not think). Therefore, in any case it is clear that the claim “I think” is true. From there one can infer that I exist.

But now you will see that it can be translated into the structure of a dilemma:

If I think—I think.

If I do not think—I think (that I do not think).

Conclusion: I think.

And from there: I exist.

But now it is easy to see the logical leap here. Note that the second premise is not “I think that I do not think,” but simply that I do not think. From this there is no necessity to infer that I think. It may be that I do not think because I do not exist, and therefore I am not thinking even that thought (that I do not think).

In other words:

If I think that I think—I exist.

If I think that I do not think—I exist.

Conclusion: I exist.

This is the structure of a dilemma, but here it is even easier to see the previous fallacy. There is a third possibility: that I do not exist at all, and therefore I am neither among those who think they think nor among those who think they do not think.

The Meaning of the Cogito

These caveats show that the argument is not of an “ontological” character; that is, it does not arrive at its factual conclusion (“I exist”) merely from conceptual analysis and without premises. You are implicitly assuming a premise—that only these two possibilities exist—and only then can you indeed infer from them the conclusion that I exist. But the premise that only these two possibilities exist essentially assumes that I think (and therefore either I think I think or I think I do not think), and of course this implicitly assumes that I exist. If so, it is no wonder that one can infer from here the conclusion that I exist. This is begging the question.

You may ask: whence do I draw this premise (that I think or that I exist)? Apparently from observation. If so, there is an observational premise that the cogito relies on as well, and this of course topples Descartes’ rationalist attempt. The import is that the conclusion is indeed true, but contrary to Descartes’ claim it cannot be proven by conceptual–ontological means. It is based on observation. I experience my existence and infer from this that I exist. This is an ordinary empiricist claim, grounded in my observation of myself.

In columns 157–158 I presented the claim of Ron Aharoni, in his book The Cat That Isn’t There, that all the problems and issues of philosophy stem from a single fallacy: the conflation of object and subject. The cogito is a clear example of this, since I play two roles there: I observe myself and think about my thought. My thought thinks itself. In this situation I am both the observer (the subject) and the observed (the object).[12] Alternatively, I am the one conducting the cogito reasoning (I am the one pondering my existence) and I am also the subject of the argument (the one whose existence is proven). Aharoni argues in his book that one must not identify these two entities, even though they both refer to the same person, and from this he reduces all philosophy to factual claims. He says that identifying the “I” as subject and as object is the mother of philosophical sin, indeed the mother of philosophy. In those columns I explained why I do not accept his claim and showed the possibility of philosophy. But in the case of the cogito he is indeed right: as I have shown, this is an observational claim in which I observe myself, and this is precisely Descartes’ fallacy—identifying these two roles of the self. Therefore, the conclusion of the argument belongs to science and not to philosophy.

Some will say that even arguments of the type of the wall and the shell—i.e., arguments that prove a negative claim about the world by virtue of logical contradiction—bear the fingerprints of observation. Observation says that the world operates in a way subject to logic. Hence, if there is a claim that contains a logical contradiction, it probably will not be realized in the world. Even here, beyond logic, some kind of observation is involved, or at least a factual assumption. This I do not accept. In my view, logic is not the result of observation of the world but something true by its own force, but I cannot go into that here.

To sharpen this further, consider the empiricists. They trust their senses as the firmest basis for insights about the world. But there is an assumption here that the images in their cognition indeed match what happens in the world, i.e., that their senses faithfully present the world to them. If so, the observational approach itself is based on a rationalist (intellectual) assumption. This is an example of an assumption that seems self-evident to us and that we use as a basis for insights about the world, even though it has no observational source. There are more such assumptions at the foundation of scientific thinking. This means that true certainty regarding any fact about the world is impossible, since at the base there are always uncertain assumptions. Yet from our point of view, this is reliable information about the world and it is reasonable to accept it.

In this sense, Descartes is right that the claim about my existence seems self-evident. It is very implausible to doubt it, at least because there is an implicit observation here (an inward look, inward). This does not mean that the conclusion is certain, but it is very plausible. On the other hand, as we have seen, it is the result of observation and not of conceptual analysis alone, as Descartes thought.

The lesson is that even if an argument does not give us certainty, it does not mean it is worthless. It can be a good argument at the philosophical level, even if there is no certainty in it. We can learn from the cogito that the existence of the spirit is indeed more secure than the existence of matter. The conclusions that there are material objects in the world (including my own body) are the result of adopting trust in the existence of the knowing spirit and in the reliability of its cognitive instruments. In this sense, I think Descartes was nevertheless right. His need for certainty is what tripped him up, and here we come, of course, to my theses about the synthetic a priori (that is, there are things that are true by their own force, i.e., without observation, but not with certainty).[13]

[1] There is a very detailed discussion of this in Hebrew in A. Z. Brown’s book, The Question of Being, Magnes, Jerusalem 1977, chapter four (hereafter: Brown). He deals there with the various critiques and interpretations of the argument. A concise discussion can also be found in the prologue to my book Sciences of Freedom.

[2] René Descartes, Discourse on the Method, trans. Yosef Or, Magnes, Jerusalem 1974.

[3] The Search for Truth by the Natural Light, cited in Brown, p. 126.

[4] See Brown throughout chapter four.

[5] Although various objections have been raised even to this, I will not go into them here.

[6] René Descartes, Meditations on First Philosophy, trans. Yosef Or, Magnes, Jerusalem 1977.

[7] In this sense, Descartes does not assume mere thinking but reflexive thinking, namely self-awareness as well. One cannot, in this way, prove the existence of a being that thinks but is not aware that it thinks.

[8] Physicists call this effect a stroboscopic effect. Imagine a particle in a dark room moving in a circle at very high speed. There is a flashlight illuminating one particular point on the path. The flashlight turns on and off at a rate controlled by us. If the flashlight’s rate differs from the particle’s rotation rate, sometimes we will notice the particle (if it happens to be at that point when the spot is illuminated) and sometimes not. Our conclusion will be that it is moving. What happens if we set the flashlight’s rate to be exactly the particle’s rotation rate (or an integer multiple thereof)? In that case we will see the particle at the same point each time. Assuming very high rates, we will get the illusion that the particle stands at the same point all the time. That is how we should relate to the conclusion that we exist. In fact it is true only at the moments when we think it, but at all other times there is no reason to infer it. It only seems to us like continuous existence because of the stroboscopic effect.

[9] This line of thought is very similar to Schopenhauer’s argument. Kant distinguished between the world as it is in itself (the noumenon) and the world as it appears to us (the phenomenon). He claimed we have no way to know things as they are in themselves, only their appearances before us (or before our cognition). We have no way to know the chair in itself, only its image that appears in our cognition. Schopenhauer pointed out one exception to Kant’s distinction: our spirit. When I look inward I encounter myself directly—that is, Michael Avraham as he is in himself—and not merely his phenomenon (the image that appears to the eyes of others who look at him).

[10] In this context, see the article by Rafi Malach, “From Photons to Fantasies,” here:

 http://www.weizmann.ac.il/neurobiology/labs/malach/pdf/photonsToFantasies.pdf. Malach argues there that vision does not photograph reality but rather constructs a model of it that is useful to us in certain respects. This claim greatly reinforces the superiority of the image of reality over the reliability of our conclusions about reality itself.

[11] This fallacy is reminiscent of the popular expression “go prove you don’t have a sister.” For example, someone asks me whether that prostitute is my sister or whether my sister is a different prostitute. To this I respond that my sister is not such a thing at all—or that I have no sister. Here too, two options are presented from which a conclusion (that my sister is such-and-such) is proven by way of mima nafshakh, while ignoring a third possibility: that I have no sister at all.

[12] All philosophy and logic begin with such loopiness: a person thinks about his thought, that is, uses reason to understand how reason operates. In a certain sense, indeed all philosophy is a loop with its tail in its mouth. Therefore Aharoni argues that all these processes are observational (I observe myself) and denies the existence of a non-scientific domain (philosophy). His claim is that if we separate the observing subject from the observed object, philosophy collapses: part of it vanishes and another part becomes a collection of empirical claims. In this context it is interesting to consider a lesson I taught on a Talmudic sugya (Shabbat 5a, “Two powers in one person”) that discusses whether one person can play two different roles and the law will be as if two different people are involved. The Talmud there leaves the question unresolved, and in a sense this is exactly my dispute with Aharoni.

[13] This differs from Kant, who tied synthetic a priori propositions to the human being and to the scientific image of the world within his cognition, and not to the world itself.