Is the World Really Real? (Column 502)

I’ve just finished reading the book Is the World Really Real? by Gai Finkelstein and Yitzhak Ben-Israel. The book was published in what seems at first glance to be an excellent series, “The Hedgehog and the Fox,” edited by Shmuel Rosner. Yitzhak Ben-Israel is a retired IDF major general who has won two Israel Defense Prizes (technology and security), head of Israel’s Space Program, and a lecturer and researcher at the Cohn Institute for the History and Philosophy of Science and Ideas at Tel Aviv University. The book is based on a course Ben-Israel teaches at the university. Gai Finkelstein is a novelist who teaches writing and completed an M.A. at that institute. I’ll add that Ben-Israel and I have a mutual friend who occasionally sends me his articles, so I’ve encountered his writing more than once. I’ve found more than a few times that I disagree with him—mainly with the conclusions and the philosophical implications he draws from science and facts—so I wasn’t surprised to discover that this is also the case with this book.

The book addresses the question of whether the external world really exists, and its novelty is the claim that modern science—chiefly relativity and quantum theory—sheds new light on this ancient philosophical issue. The book essentially argues that science has decided it. When I approached the book I held the opposite starting point: science cannot, in principle, settle this question. Anyone who reads my series of columns on Zeitlin and Kant (494 – 496) will understand why.

Those columns were written before I knew the book, so it isn’t mentioned there; but right after I finished them I received the book—that was the first “miracle” on the way to this column. As I read it, I thought it appropriate to write a companion column to that series engaging this book. The second “miracle” accompanying this column’s birth was Gai Finkelstein’s chance visit to the site. She posted a brief comment explaining that she had happened upon the series and thought it would be interesting to connect it with the book. By then I was already in the thick of reading, as I replied there. This strengthened my desire to write the column, and here it is.

I’ll note that the first part is a critique of the book that will interest less those who haven’t read it (it relates to my series on Kant). In the second part I’ll enter my main topic: whether modern science indeed dispels doubt regarding the existence of an external world.

The Structure of the Book

The book is divided into four chapters: 1) A philosophical introduction on Kant and the question of whether the external world (to us) exists. 2) Relativity: from Newton (18th century) to Einstein (20th century). 3) Quantum theory (20th century). 4) Back to the question: a proof that the world exists.

I must say I was impressed by the two middle chapters, which present conceptually very difficult scientific material and make it remarkably accessible to the general public with almost no prior knowledge assumed (occasionally terms like “complex numbers” are mentioned without explanation, but in my impression this doesn’t really hinder reading). They describe thought and real experiments in relativity and quantum theory that have transformed our entire way of thinking about science and the world. These aren’t superficial, top-down descriptions as is customary in the genre, but rather genuine presentations of the science itself and its implications, including fine, nontrivial nuances. For this the authors certainly deserve high praise.

Philosophy and Science

But the book’s goal and framework concern a philosophical question, not science. Science is only a means.[1] The book’s question is whether the world exists. As noted, their claim is that modern science enables us to move toward a decision on this question, although it has always been considered a philosophical question inaccessible to scientific treatment.

It’s interesting that a similar situation exists regarding the question of free will, which I discussed in my book The Sciences of Freedom. There too, a consensus has been forming among researchers that science has resolved this ancient philosophical question, and because this seemed to me prima facie untenable, I decided to dig in and examine whether that is indeed the case. It isn’t. In that book I showed that this ambition fails, largely because of philosophical and conceptual arguments many stumble over. To a great extent, the same is true here. It’s very hard for me to accept that the question whether the external world exists—which by definition concerns the world as it is and not as it is perceived by us—is even accessible to scientific treatment, since science deals with the world as we perceive it. Therefore the book’s ambition surprised me greatly, and I entered it to examine the matter, I hope without bias. My conclusion here too is that we are still very far from it.

Does the World Exist?

As stated, the first chapter offers a philosophical introduction and presents the question of whether the world exists. Simply put, it’s the ancient philosophical question: is everything we know nothing but an illusion existing only within us, or does our cognition apprehend an external world that exists independently of us? On the face of it, this is hair-splitting that no one—except a few eccentric philosophers—is really troubled by. It’s obvious to all of us that the world exists, even if we have no good arguments for this stance. Just as every axiom I hold is obvious to me even if I cannot prove it or perhaps even argue for it. Some things strike us as self-evident and require no proof or justification.

Some raise arguments from various illusions that we all agree can occur within us—some natural (Fata Morgana) and some artificial (the direct artificial induction of sensation, mood, consciousness, or experience in the brain). But even here, this doesn’t significantly challenge our intuitive view, since the very fact that we distinguish truth from illusion indicates that illusions, which indeed exist, are exceptional cases. There’s no reason to think that if we have illusions then everything is an illusion. This can be illustrated via Russell’s celestial teapot. If someone tells me there is a small teapot orbiting Jupiter, I will have no counter-evidence that he’s wrong. On the other hand, I have no reason to assume he’s right. So I won’t consider this a situation of balanced doubt (there is or isn’t such a teapot), but will continue to assume there is no such teapot. Even if examples are shown where some of my assumptions turned out false (this has happened to some people—though not to me, of course), I won’t abandon this stance. Errors exist, but absent evidence, my assumption is that my intuitions are correct. Of course, if evidence appears I must be honest enough to admit error and switch paradigms.

Skepticism is a very healthy approach at the methodological level. It’s important to challenge basic assumptions and subject them to rational and empirical testing, and if I’m wrong, to replace them. That’s exactly how we discover errors and learn and improve. But one mustn’t confuse methodological skepticism (à la Descartes) with substantive skepticism (as in Pyrrhonism and its successors). The claim that everything we think is wrong—or even doubtful—is a claim, not a question, and in my view a highly problematic claim. In short, healthy doubt ends with a question mark, not an exclamation point.

Kant’s Phenomenon and Noumenon

Kant’s question of phenomenon and noumenon, which we encountered in those columns, seemingly deals with the same issue. But on further consideration, it’s not quite accurate. Kant discusses whether what we apprehend (the contents of our cognition) describes something in the world itself, or whether it’s all the product of our imagination—existing only within us. It’s quite clear this isn’t entirely subjective experience, because all humans experience similar things; but Kant wonders whether it might be inter-subjective (existing within each human subject, but only there and not “outside”).

This, of course, is not the same as the previous question. The first concerns ontology (what there is; the domain of metaphysics), whereas the second focuses on epistemology (what we know and how; the theory of knowledge). In yeshiva jargon I’d say the first is about the cheftza (the object) and the second about the gavra (the subject). There is, however, a one-way connection between them. If the world doesn’t exist, then clearly everything said is about us and not the world. But if the world exists, that still says nothing about the second question. It may be that the world exists, but our cognitions are generated within us and don’t reflect anything out there—and it may be that they do reflect what happens in the world as it is (the noumenon). True, the epistemic question presupposes ontic assumptions: if I assume my cognitions reflect the world, I can claim something about what is in the world (what my cognition describes exists there). But that’s only an indirect ontic consequence—and we’ll soon see that even that isn’t entirely right.

Kant argued that our cognitions (the phenomenal realm) are indeed a reflection of the world as it is (the noumenal realm), and implicitly presupposed its existence. But in his view the reflection isn’t transparent. It’s shaped in our image; therefore some of its features are products of our cognitive and conceptual structures rather than of the world itself. By analogy, if we look at the world through green cellophane, it will all appear green to us. According to Kant, the images we see reflect what’s happening in the world itself and not only in us, but the green color (their mode of appearing to us) is the result of the perceptual tools we use, not only of the world as it is.

Kant assumes the world exists and, as far as I know, doesn’t even raise doubts about this nor does he argue in favor of it. The picture he presents concerns cognition and claims that cognition reflects what happens in the world; but it is still not correct to say that now I can tell what there is in the world as it is. The world affects what happens within me, but the image formed in my cognition is not the world itself. Epistemology does not simply project onto ontology.

First Note on the Book

Here I come to my first remark on the book (this is remark 2 in my reply to Gai Finkelstein on the site). They present the Kantian doctrine as if its foundation and focus are ontological (what there is)—as if Kant came to solve whether the world exists and what’s in it. But to my understanding this is wrong. His assumption is that the world exists. Kant focuses on epistemology—on how and what we know about the world and the relation between our cognition and the world as it is.

In this context I’ll note one critique of Kant (Lev Shestov’s second difficulty that I brought in column 495): whence does he infer that there is any external world at all, if our cognitive tools are the only medium by which we can apprehend it? The cognitive tools give us only the phenomenal, so how do we know there is a noumenal realm that generates it?! This is further indication that the ontological problem isn’t addressed in Kant’s doctrine. He came to solve Hume’s epistemological question (how we may trust the laws of science), not the ontological question (whether the world exists and what is in it).

Second Note on the Book

Ben-Israel and Finkelstein repeat the claim that, according to Kant, cognition filters out contents that don’t fit its structure; therefore the laws of nature that arise within us fit the facts we apprehend. This is not a match between the world as it is and our cognition, but between what we apprehend of the world and our cognition—simply because what doesn’t fit cognition won’t be apprehended and thus won’t cause difficulties.

In column 495 I showed that this is a common mistake among Kant’s interpreters (including Shestov), and at most—even if they are right—it would be Kant’s own mistake. As I showed there, it’s an implausible picture and it doesn’t resolve the difficulties. There is no reason to assume that if there are bodies accelerating with acceleration different from what Newton’s second law (F = m*a) dictates, we simply won’t see them. Why shouldn’t we see a body upon which a force of 2 newtons acts, whose mass is 2 kg, yet its acceleration is 3 m/s²? After all, that same body, if a force of 6 newtons acts on it and it accelerates at 3 (which matches Newton’s laws), will be perfectly visible. Would a curtain then fall over our eyes when the acting force changes, such that we can’t see the same body accelerating at the same acceleration? In both cases it’s the same given body with the same specific mass accelerating at the same acceleration (only the force differs between the situations). Would something suddenly keep our eyes from seeing it?

This “mystical” explanation doesn’t help us, since it’s no more logical or plausible than the riddle it is meant to solve (how and whether we apprehend the world correctly). It has no explanatory advantage. It’s like explaining the perfect match between the image in our cognition and the world by saying there’s a demon who always paints the right picture for us. A wonderful explanation—but it doesn’t help, because we will now ask how the demon works and how you know such a demon exists at all.

As I explained there, in my view the Kantian picture doesn’t speak of filtering out ill-fitting contents, but of tinting reality with colors taken from our cognitive tools. I explained that in the world as it is there are no colors, and neither are there sounds, accelerations, smells, energies, fields, and the rest of our cognitive terms. Therefore, when a tree falls in a forest with no ear and eardrum present, it clearly doesn’t emit a sound; at most it generates what we call an acoustic wave. When such a wave strikes an eardrum it generates in the ear’s owner the sensation of sound. But the sound exists only within us and not in the world itself. Of course, the assumption is that there is also something in the world itself: the acoustic wave that produced the sound is a raw fact existing in the world. Any description of it (certainly the sound it generates in us, but also concepts like wavelength, speed, amplitude, and so forth) is made in terms belonging to our conceptual world and using our modes of thought. The same holds for light and any other perception in our cognition.

Think of creatures who don’t apprehend the world using the categories of space and time. They obviously won’t speak in terms of acceleration, and their physics wouldn’t include Newton’s second law. Creatures not equipped with eyes and an optical system wouldn’t speak of colors. Creatures who don’t think causally might not include cause and effect in their scientific doctrine. Such creatures would describe the world in a language and conceptual scheme entirely different from ours, yet they would describe the same world—no less correctly than we do. They would simply use a different language. The translation of raw facts into their language would differ from the translation into ours. But there is no right or wrong here, and none of these translations is what exists in the world; rather each is a reflection of what exists and its casting into the patterns of the perceiving creature.

My claim is that if we adopt this picture of the relation between facts in the world as it is and their appearance in our cognition—namely if these are processes of translation and tinting (systematic ones) and not filtering (as presented in the book)—then the description of the world in our cognition (the phenomenal) precisely matches what happens in the world as it is. We don’t have incorrect perceptions arising from our cognition and its limitations. There are, of course, known limitations—for example, that we cannot see beyond our visual and auditory spectrum. But these limitations are specific; they don’t stem from the very difference between the world as it is and the phenomenal. My claim is that these pictures are formulated in our language and conceptual system (tinted by the colors of our cognition), but they are a correct and fitting description of the world itself. They are entirely correct and not subjective at all in their content. This is our way of describing reality. Using Hebrew or English shouldn’t change the facts described in it (beyond nuances, of course).

I explained the ramifications of this modification to the Kantian picture in that column (thus Shestov’s and other difficulties are answered). But as I explained there, this assumes that the solution to Hume’s problems lies in our having a non-sensory cognitive faculty (ideatic intuition, in Husserl’s terms). Therefore I’m not sure that what I’m describing is really what was in Kant’s thought, or at least in his awareness. But I’m fairly certain that it’s the correct description. My concern here, as there, isn’t interpreting Kant’s doctrine but answering the ontological and epistemological questions at issue here. Add to this the principle of charity (see column 440), by which I interpret the Kantian picture in the way most optimal for him, and you’ll get this picture even if we assume Kant wasn’t consciously aiming at it (and I’m not at all sure of that).

In any case, my second remark on the book is that it assumes Kant speaks of filtering out ill-fitting facts from the world itself; not so: he (or at least the correct formulation of his doctrine) speaks of tinting or translating them into the language of our cognition and thought. This was remark 1 in my reply to Gai.

A Note on Intuition

By the way, the picture I described in answer to the second question is a good example of our intuition’s failure. If you ask a person on the street whether colors and sounds exist in the external world itself, he will confidently answer “yes.” Most people would treat such a questioner as a peculiar skeptic. When a tree falls in a forest, in their view it certainly emits a sound even if no one is there to hear it. Yet notice that here we’ve seen that this is a mistake. Even without Kant, it’s now completely clear that colors, sounds, smells, and other cognitive concepts exist only within our cognition and not in the world itself.

Consider that we recognize seven colors, although in the electromagnetic spectrum there is no seven-fold division (there is a continuum of wavelengths). The division into colors is the result of sensors in our brain. We have seven of these, and so they divide the visible spectrum into seven parts, each perceived by us as a different color. Another creature might see ten or a hundred colors, or might not perceive images via colors or sight at all (but through sound, smell, or another kind of cognition entirely unfamiliar to us). This is an indication that colors are not in the world but only in us—contrary to every sane person’s intuition. And still, as I explained, regarding the first question (does the world exist), I think there’s no reason to reject our intuition.

Interim Summary

So far, critiques of the book. We saw that Kant assumed the existence of an external world (the noumenal), but offered no arguments for it. One could hang this on the principle of causality: if within our cognition there is a world presented in varied, multi-sensory images (tastes, sounds, tactile sensations, smells, and sights), then apparently something generates all this, namely the world as it is. If so, the principle of causality seemingly teaches us that an external world does exist.

But Kant and Hume taught us that the principle of causality itself is synthetic a priori; that is, it is part of the nature implanted and structured in us, which we impose upon reality (in my view it describes reality itself and we apprehend it with the “eyes of the mind,” but that is precisely the debate here). Therefore it’s implausible to use this phenomenal principle to infer the existence of noumena beyond phenomena. We are confined within our modes of perception, and any claim about what lies beyond them cannot rely on those very modes.

But this is a substantive problem. It seems any rational or cognitive argument we bring will be vulnerable to the same attack. The argument will be based on our ways of thinking and knowing; assuming these originate in our cognition, such tools cannot teach us anything about the external world. If we adopt the opposite assumption, then of course it’s possible—but then we’ve begged the question: having assumed there is an external world, it’s no wonder we can construct an argument that it indeed exists.

Thus, for example, on p. 273 the book presents a Kantian argument (unknown to me) for the existence of an external world. He claims that if there is no external world then all the things we see are arbitrary and dependent on us. If so, asks Kant, why are there laws of nature? Why can’t we decide what the world’s nature will be? Why be dejected about what happened? One needs only decide that this is not what happened, and that’s it. He also asks what the difference between wakefulness and dreaming is, if in both cases these are subjective phenomena. These are such feeble arguments—some even downright mistaken—that I find it hard to believe Kant actually made them. First, even if things depend on us, it doesn’t follow that we can decide whatever we want. We ourselves have a nature that may impose itself upon us. The assumption that if things aren’t objective then they’re arbitrary is baseless. More generally, the laws of phenomena may be unrelated to an external world, and still they are our phenomena and we cannot change them.[2]

Bottom line: philosophy doesn’t seem to advance us here, and after exhausting the philosophical level, the question remains open. But perhaps science can move us further? I’ll already say here that this seems impossible for the same reason. Science deals exclusively with phenomena; as such it cannot, logically, reveal anything about what lies beyond them (the noumena). This was the reason for my skepticism regarding the book’s stated aim. Let’s now touch on what the book says about this question.

Does the World Exist: Kant and Quantum Theory

In the book’s fourth and final chapter, the authors return from science to philosophy and argue that the picture presented by modern science can resolve the ontological question of whether the external world exists. Let us now read this chapter critically, and I’ll try to show that it does not (this was remark 3 in my reply to Gai Finkelstein).

The chapter opens with the claim that after Kant it’s clear there is a gap between what we see and the world; that in some sense our cognition determines the world. The authors write that this claim resembles the common interpretation in quantum theory, according to which cognition determines reality.

But we must return to the meaning of Kant’s claim and the novelty quantum theory might add here. Regardless of quantum theory, from current scientific knowledge it’s clear that the images in our cognition are formulated in our language. We saw that colors and sounds exist only within our cognition and not in the world itself. The assumption is that in the world there are trees and falls and even an acoustic wave, but the sound exists only within our cognition. This is a relatively new scientific finding, but it’s not related to quantum theory. When we understand that the brain has seven visual receptors that divide the spectrum into seven colors, we understand that “color” is a cognitive concept not found in the world. This has nothing to do with quantum theory, and in fact today this is not only a philosophical claim (as in Kant’s time) but almost a scientific fact.[3]

What, then, is quantum theory’s novelty here? (See my series 322 – 327.) Quantum theory teaches us that the act of measurement in physics affects the measured entity, i.e., the measurement’s outcomes. What’s new here? The claim is that here there isn’t only “tinting” of the result with our cognitive color, but the result itself changes due to measurement. If we measure position, the particle loses the datum of velocity (it has no defined velocity), and vice versa. This isn’t merely a limitation of our cognition imposed on reality, as we saw in Kant and generally (a “tinting” process), but an effect of measurement upon reality. In the book’s interpretation of Kant—according to which it is filtering rather than tinting—perhaps it’s more plausible to connect this to quantum theory, since in both cases there are results which, even if they exist in the world, we won’t be able to observe (under a subjective interpretation of quantum theory; under the common interpretation they don’t exist at all, even in the world itself). But I showed that the filtering thesis is problematic, and the tinting thesis presents a Kantian picture unrelated to quantum theory.

Beyond this, as the authors rightly write (contrary to common errors), the effect on the measurement result isn’t due to human cognition’s involvement but to the mere placement of a measuring device. The device interacts with the measured system (the particle) and thus changes its properties. If so, this isn’t at all a feature of our cognition and its relation to reality, but of the meaning of the physical act of measurement itself (even if no one ever receives its results).[4] Therefore I see no connection between this and the Kantian question.

From the picture I’ve presented it follows that it’s very plausible that all quantum phenomena belong to the phenomenal realm. The measurement’s effects on outcomes are entirely phenomenal; therefore they say nothing about the noumenal. If someone assumes there is no external world and that everything is our invention, quantum theory won’t change that. On the contrary, following Kant it’s quite clear that the very distinction between the measurer and the facts is entirely a division within cognition. Kant taught us that the “facts” in question are phenomenal facts, not facts in the world as it is. More generally we can recall Shestov’s claim (presented in that column) that even the very division between noumenon and phenomenon is itself within the phenomenal realm, since it is a product of our thought/cognition.

Later in the chapter the authors note that within the Kantian framework it’s easier to accept quantum oddities (like Schrödinger’s cat). The cat is simultaneously alive and dead, but its life and death are claims about the cat’s phenomenon (what we know or say about it), not about the cat itself. I think they are completely right about this; but does this say anything about the question of the external world’s existence (the noumenal)? In my view, no.

True, under the common interpretation, quantum oddity is not only in cognition but in the world itself (this is the usual interpretation of the results of the Aspect experiments and Bell’s inequality, nicely described in the book, which I cannot discuss here). Yet it remains clear that we are dealing with the world as it is apprehended by us. Quantum theory, like physics and science generally, deals with phenomena, not noumena. Hence even statements about “things themselves” in quantum theory are statements about something phenomenal, not the external world as it is. As noted, the very division between phenomenon and noumenon is also made within the phenomenal. Philosophy—Kant’s and philosophy in general—just like science, is carried out by our mind and cognition and, as such, belongs to the phenomenal.

The authors claim there is similarity between Kant’s doctrine and common interpretations of quantum theory: both assume a world exists and we are dealing with knowing it and creating images of it dependent on cognitive tools (with the differences I pointed out earlier in this section). That’s true, but it’s similarity in assumptions. They bring no evidence that these assumptions are correct (below I will critique the one argument they offer on this). For now we have no argument for the external world’s existence—just a kind of ad hominem: not only Kant thought so; quantum interpreters think so too. It may still be that all of them are wrong. Moreover, as noted above, these quantum interpretations don’t really speak about the world as it is, but about facts within the phenomenal realm.

Later in the chapter they arrive at a two-stage argument that proves the existence of an external world. This is the book’s crescendo toward which the entire arc was meant to lead. But as I’ve already remarked, the chapter in general, and this argument in particular, are weak and very disappointing. Let us now discuss the two stages.

Stage A: Order in Cognition and in Things-in-Themselves

On p. 274 the authors begin from the Kantian premise that the order and lawfulness we find are part of the phenomenal world, not something in the world as it is. Now they claim that for this order to exist, the sensory data reaching us must undergo filtering (I remind you that for them it is filtering, not tinting/translation). If this order also exists in things-in-themselves, then there is astounding coincidence responsible for the match between order in our cognition and order in reality itself.

By itself this means nothing, of course. On the contrary, perhaps it’s an argument against an external world. But the very claim is problematic. The order in cognition can certainly be a product of order in reality itself. Here they fall into the fallacy of filtering instead of tinting. On my view—that it’s tinting—there is nothing surprising: order in things dictates order in cognition; only that the cognitive order and lawfulness are formulated in the language of cognition (translated/tinted into our cognitive language).[5]

Stage B: An Argument from Randomness in Quantum Theory

Earlier in the chapter they note that unlike Newtonian physics, which is deterministic, quantum theory teaches us that microscopic physics is not deterministic. Even in the macro realm, quantum random effects are merely smeared and canceled out—coherence breaks—but there is no true determinism there either.[6] They rightly explain that this is essential randomness, not merely uncertainty and vagueness in our information about reality. When you throw a die, there is uncertainty regarding which face will come up; but that is only our lack of knowledge. If we knew the initial velocity and all environmental conditions and the die’s shape and mass, we could compute the final result uniquely. There it’s not real randomness, but a lacuna in our knowledge of it (epistemic randomness). By contrast, in quantum theory (according to the common interpretation—not all agree) there is genuine randomness, i.e., indeterminacy in reality itself (ontic randomness). In my series 322 – 327 I called this “indeterminacy” (ontic) versus “uncertainty” (epistemic).

Now they ask (p. 275) whence quantum randomness. If the source of all lawfulness and phenomena is in cognition and there is no external world generating it all, then we should expect everything to be deterministic. According to Kant, our cognition is causal-deterministic; it imposes itself on reality (filters out whatever doesn’t fit). At this point they ignore the possibility that there is no randomness in quantum theory, only nonlocality (an alternative interpretation of Bell–Aspect experiments). But beyond that, they claim that randomness poses a challenge to the Kantian picture. If cognition operates causally, it should impose causality on all the facts we apprehend. So whence random facts? Apparently they don’t arise within us but come from the world itself. This is the book’s argument for an external world. After the entire arc, here appears the “Q.E.D.”

Critique

There are several problematic, certainly non-necessary, assumptions here. First, they again apply the principle of causality problematically. Their claim is that if we have random phenomena within us, then necessarily something causes them. But as we saw, the principle of causality itself belongs to the phenomenal realm; whence do they know it can be applied to the noumenal? One might say these are phenomena within the phenomenal realm and as such should have a cause—but if so, then the proposed cause should also belong to the phenomenal. It’s not right to hang it on the noumenal. In other words, the natural conclusion is that our cognition has a non-deterministic random component, and that is what generates random phenomena.

Beyond that—and this is really the upshot of the previous point—perhaps the very assumption of causality is wrong (even within the phenomenal). In that case, random phenomena in our cognition could stem from a random component within us. True, Kant didn’t think so; but that’s only because that component was discovered only in the 20th century. Until then people thought we are built in a causal-deterministic way, so Kant, who wrote in that period, naturally assumed this. Now it has been revealed that there is a random, non-deterministic component within us—i.e., Kant and his contemporaries erred. What does this say about the external world’s existence? At most it’s a normal scientific advance, which in Kant’s terms means exposing more and more parts of the phenomenal realm and our cognition. Kant also thought time is objective (because he presupposed Newtonian mechanics); since Einstein, it turned out he was wrong. But that’s a scientific error by a person who didn’t know later scientific discoveries. Does this mean there must be a philosophical error in his picture? In short, all our lawfulness and science could indeed be a product of cognition, as Kant posited; only, contrary to what he thought, cognition doesn’t consist solely of deterministic components. In this, Kant may have erred. But his first assumption can remain in place even if the second turned out wrong.

If this is the argument we marched toward throughout the book, it is indeed disappointing. The groundwork contained exciting material—concept-shattering accounts of quantum and relativity—but we end up with a very weak argument that doesn’t even hint at relativity (which in some sense even undermines it; see my note on time in the last paragraph). Why was relativity needed at all? And even on its own terms, the argument is weak, simplistic, and does not hold water.

In short, I highly recommend reading the book for its scientific parts. I don’t know of another Hebrew text that presents the basic ideas of relativity and quantum theory so clearly and accessibly without sacrificing precision. But the philosophical framework is very disappointing. If that was the trigger for presenting the scientific picture, it turned out quite superfluous. As I wrote in the footnote above, I suspect the situation is the reverse.

A Note on Free Will

At the end, the authors briefly address free will. This is no longer part of the main argument, but I’ll comment on it as well.

On p. 278 they open with a remarkable note: if Kant had known Bohr’s theory, he would have understood that randomness can also exist within the phenomenal, and he would have had no need to see free will as a challenge to his doctrine. Even before I enter the discussion itself, I cannot help noticing that this claim flatly contradicts their main argument (described in the last sections). Their argument presupposes that randomness must originate in the world itself. But if they themselves hold (exactly as I objected to their argument) that it’s possible there is a random component in cognition (in the phenomenal) itself, then their argument collapses of its own accord. At most, it shows Kant made a marginal mistake, while at the same time toppling their (already failing) “proof” of an external world.

Kant himself, they say, offers a kind of clumsy solution to this difficulty (I won’t go into it here—it’s not really worth it). But the authors offer him another solution, supposedly better, in light of the picture they support. They claim there is a non-deterministic component in the world itself (as quantum theory, in their view, shows), and this opens a door for free will. There is no need to assume the source is cognitive-subjective; Kant’s embarrassment disappears. So far their claim.

But if the source of randomness is in the world itself and from there it reaches cognition, two problems return:

• We’re begging the question: we assume a world exists and that it’s the source of randomness in the phenomenal. But that is precisely the point at issue. Recall we’re seeking a proof for the existence of an external world.
• If cognition indeed imposes causality upon us and filters facts accordingly, as they interpret Kant, why and how does it allow random phenomena to penetrate it? We’re forced to conclude that the relation between things-in-themselves and cognition is one of tinting, not filtering—exactly as I explained and contrary to their stance. It appears this chapter, which was meant to integrate all the preparatory claims accumulated along the way, actually contains a bundle of contradictions within itself.

Beyond all this, there is a categorical error in their identification. Quantum theory speaks of randomness, but free will is not a random or stochastic mechanism. If our will runs lotteries, then there is no place to hold us responsible for our actions. Randomness is not choice, and therefore cannot account for what is usually meant by “free will.” Quantum theory sets a wave function from which a distribution of possible outcomes is derived; the outcome is generated by a lottery of some kind whose probabilities are dictated by that distribution. But free will is supposed to be free, not determined by a draw. If it’s determined by a draw, it may be non-deterministic, but it isn’t free in the substantive sense.

I have often pointed out this error, which also appears in arguments like van Inwagen’s (see my article here and many more). The libertarian claim, as opposed to the determinist, is that there are three mechanisms, not two: determinism (causality), randomness (chance, without cause and without purpose), and free will (teleology). Free will is action without cause but with purpose and deliberation—unlike a random action, which is a kind of lottery (without cause and without purpose). Therefore we mustn’t equate indeterminism with free choice. Hence hooking quantum theory (random) up to choice or free will is a categorical mistake.

At the end of the book the authors propose to see free will as the product of a quantum computer (as opposed to a Turing machine). Without getting into details, this proposal suffers from the same problem I described in the previous paragraph (randomness is not free will). Beyond that, there’s another problem: they assume an identification between machines and human beings. Computation in a machine is not deliberation or human thinking. By the way, Ben-Israel makes the same identification elsewhere in the book when he refers to missiles as a sort of machine that “knows” or “thinks.” But a missile neither knows nor thinks. A missile is a machine. Likewise, our brain doesn’t think; it performs physical operations (conducting currents in neural networks). Thinking, deliberation, and choice are done by the mind. The mind uses the brain (it gives interpretation and meaning to those currents), but the mind is what does the thinking, not the brain. We think by means of the brain just as we walk by means of our legs. Therefore, even if our hardware were a quantum computer, we would still need to explain the mind—i.e., the thinking that uses this quantum computer, and in it (and only there) choices, understandings, and decisions arise. This is a common error among AI people, but I discussed it in column 175 and won’t return to it here.

[1] It’s hard to avoid the impression that contrary to the authors’ statements, science for them is not merely a means to philosophical discussion. My impression is the reverse: the philosophical question is a means allowing them to deal with modern science and describe it to readers or students. The ratio between the scientific and philosophical parts makes no sense if the goal were philosophy. There is elaborate detail in the scientific part and a very brief final chapter that draws philosophical conclusions. The level of scientific detail is in no way required for their philosophical argument. Relativity, for example, plays no role whatsoever. The arguments in the fourth chapter rely solely on quantum theory. All this suggests the authors actually wished to make science accessible to the general public, and the philosophical question is a trigger—a methodological device enabling discussion of science and the claim that it is relevant to non-scientific questions as well. Incidentally, the fourth chapter, which was supposed to be the book’s focal point—bringing us back from scientific details to the big ontological picture and presenting arguments proving that the world does indeed exist—is extremely terse (and, as noted, hardly needs any scientific fact). Its arguments are very general and rather weak, and their connection to the science previously described is tenuous. This is the main chapter—and it’s the weakest. I’ll get to it in the second part of the column.

[2] True, we ourselves are also objects in the world, and our psychological constitution is also in some sense external world. But if that is the intent, then we’ve returned to the cogito. Clearly Kant’s discussion isn’t about our own objective existence. On the contrary, some will say that if it’s clear that we ourselves have objective existence, that is a proof that there is something that exists. From there it’s easier to proceed and argue that there’s no reason to doubt the existence of other things in the world as it is. Incidentally, if one seeks a philosopher who actually deals with the world’s existence, it’s not Kant but Descartes (and of course the various idealists, like Berkeley and others).

[3] I write “almost” because one could claim that even though our brain has seven color receptors, somehow they reflect a division that also exists in the world itself. But this is a baseless claim; there’s no reason to assume it. By Occam’s razor, it’s enough that our brain’s sensing is divided thus to explain the phenomenon of colors; so why—and whence—assume that the world itself is similarly divided?!

[4] In my book The Sciences of Freedom I pointed to the double-slit experiment in which a detector is placed by one slit, but the detector’s measured results are sent to a computer that immediately destroys them. That is, a case where the detector’s result never reaches human cognition. It turns out the quantum phenomena—i.e., the effect of placing the detector on the measurement result—occur there as well. This proves it isn’t an effect related specifically to human cognition. This of course raises a nontrivial question: what is a measurement at all, and how does it differ from any other physical interaction? I won’t enter into this (mainly because I don’t know).

[5] Incidentally, even if it is filtering, according to their view, this neatly explains why cognitive lawfulness fits order in “things themselves,” because only the filtered picture of things themselves—the phenomenon—appears in cognition. So what’s the problem?!

[6] Incidentally, in describing quantum randomness the authors use the tunneling effect, in which particles pass through “walls” (potential barriers). They attribute it to the uncertainty principle (they explain that it penetrates the wall because position fixes a small uncertainty in place, hence a high momentum that passes through the wall). To my understanding, this is not scientifically correct. On the contrary, the particle is not confined in the well between the walls, and therefore it is located also beyond the wall. This means that precisely the uncertainty in position increases.