Atheism as “Scientism”: On Science and Philosophy (Column 586)

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.

Following the debate I held a few weeks ago with Aviv Franko about the rationality of belief in God, I found myself reflecting on the nature of the discussion. To the best of my judgment I showed there that his arguments don’t hold water, and most of them didn’t address my arguments on the relevant plane at all (see below). But what especially provoked my thinking was the character of his arguments. These focused on observation and facts while consistently denying philosophical–a priori arguments. One could say that the debate wasn’t about God’s existence at all but about rationality and rationalism. From experience I’ve gained in such discussions I anticipated this, and therefore devoted my entire opening remarks to a methodological preface (I hardly mentioned the argument itself), since from experience I know where atheists tend to take these discussions.

In the Christian world of theology (some would say apologetics), this phenomenon is called “scientism” (scientism). It’s an approach that sees empirical science as the be-all and end-all, and greatly disparages philosophical and a priori arguments. This debate helped me formulate and sharpen the connection between atheism and scientism/empiricism, and in this column I wish to discuss it. I’ll also address the arguments that came up in the debate, but it’s important to clarify that my aim is not to examine them in and of themselves, but mainly to use them as a medium through which to examine the relationship between atheism and empiricism. Therefore, before I get to the arguments themselves, I’ll first offer a brief introduction to rationalism, empiricism, and science.

Rationalism, Empiricism, and Modern Science

I’ve often dealt with the dispute between rationalism and empiricism (see, for example, columns 363 – 364, and more). Until the dawn of the modern age, Western thought was based mainly on reasoning and less on observation. There was trust in our mind’s ability to understand the world by force of logical considerations. Thinkers and scientists assumed that what seems reasonable to us indeed obtains in reality. In time, this approach was called “rationalism.” It’s important to understand that rationalism is not a synonym for rationality. Rationality is a kind of sensible approach, whereas rationalism is a philosophical view that advocates relying on thought to know the world. Non-rationalist approaches also claim the crown of rationality; in fact, empiricists argue that rationality requires abandoning rationalism. Reason itself, they say, tells us that it’s wrong to rely on reason alone when seeking to know the world.

In the late Middle Ages and early modern period, the empiricist stance gradually replaced rationalism. This is the view that the only way to know the world is through observation. According to empiricists, the intellect operates in ways dictated by its structure, and therefore there is no reason to assume that what seems logical to us accurately describes the world itself. To understand how the world works we must observe it. In this context I like to quote Mark Twain: “The world doesn’t owe you anything; it was here first.” From the sixteenth century onward, the rationalism that had dominated since ancient Greece began to be pushed aside, and empiricism took its place.

This philosophical revolution was the basis for the emergence and development of modern science in the modern era. Thus science separated from philosophy and became a distinct discipline. What characterizes modern science is mainly its severance from a priori speculations and adherence to observation—and it truly achieved most impressive results. Ancient science, which relied mainly on thought and less on observation, was stuck with notions and claims not necessarily connected to facts and to the world (think of Aristotle’s four elements—earth, water, fire, and air—versus Mendeleev’s modern periodic table of chemistry). When scientists began to adhere to observation and were ready to change their concepts in light of the facts, science began to advance. In this way we reached, in our day, the moon, the decoding of the genome, quantum theory, atomic bombs, understanding the human brain, nuclear reactors, and much more. Unsurprisingly, these achievements repeatedly strengthened science’s status, and many today think that science is the only tool for understanding the world. Unsurprisingly, empiricism entrenched its status as the most conspicuously rational expression. Rationalism became a non-rational method.

The Line Isn’t Sharp

I’ll add another important point that I also emphasized in the debate. Modern science is indeed based on empiricism and adherence to observation, but to the same extent it rests on a sizeable set of a priori assumptions that have no observational basis. Chief among them is our very trust in our senses, which is the most basic foundation of empiricism—this too is an a priori assumption. In principle, our senses could mislead us (Descartes’ evil demon), and the assumption that they do not is not grounded in observation (an empirical corroboration that our senses don’t mislead us would of course be circular). In addition, there are causality, induction, Occam’s razor, the denial of action at a distance (physicists’ assumption that two bodies cannot act upon one another from a distance; there must be something that mediates that action and transmits the effect from one to the other), and many other assumptions. The notion that science is based on observation alone is naïve and reflects a misunderstanding of science’s essence and its mode of operation.

David Hume, one of the fathers of British empiricism, presented a very skeptical empiricism, because he understood well that in science we use quite a few assumptions that have no observational basis. As an empiricist, he tended to see them as methodological assumptions that guide us rather than claims about the world (for according to his empiricism, claims about the world can be based only on observation). But many scientists (and humble me among them—see e.g. column 363) see all these as factual claims about the world. The justification I’ve offered more than once (see especially in Shtei Agalot and Emet Ve-Lo Yatsiv) is intuition, which in my view is a non-sensory contemplative faculty (and not pure ratiocination). In any case, a simple fact is that there are many such assumptions at the foundation of empirical science. This greatly blurs the distinction between rationalism and empiricism.

One must understand that this fact by itself ought to shake the certitude of empiricism. If indeed we cannot accept any factual claim about the world that is not the product of observation, then science itself is not admissible as a system of claims about the world, for it, too, relies on principles such as causality, induction, and the like. And if we are prepared to accept scientific claims (laws of nature) as claims about the world, then implicitly we also recognize thought as a legitimate tool for knowing the world. Rationalism ceases to be invalid in principle. The conclusion to be drawn from this complex picture is that while it is indeed very important to rely on facts and to beware of our intuitions—avoiding taking them for granted or as certain (extreme rationalism is a recipe for error)—it is equally clear that observation is not a tool that suffices by itself for knowing the world. Pure empiricism (reliance on observations alone) is an illusion that doesn’t hold water. Their combination is the winning formula.

Science and Falsifiability: Updated Empiricism

If every scientific theory is based also on reasoning and not only on observation, the question arises: what distinguishes it from a purely philosophical claim? In what sense is science indeed a more reliable tool for knowing reality than Aristotelian philosophy (which saw thought and logic as the chief instrument)? Where does science have something beyond philosophical conjectures? Is there any difference left between empiricism and rationalism? This means that if we don’t want to return to ancient Aristotelian conceptions, we must redefine science.

Here we must invoke Popper’s criterion of falsifiability (see columns 403 – 405). Karl Popper argued that a theory is scientific if it can be subjected to a test of refutation, i.e., if one can propose an experiment whose failure would refute the theory and whose success would corroborate it. One must grasp the importance of this criterion. We saw that a scientific theory is based, beyond facts, also on generalizations and a priori rational assumptions, and in that sense it resembles a philosophical thesis. But there is still a difference between them: a scientific theory is willing to subject its claims to an empirical test of refutation. I won’t get into the subtleties and disputes around Popper’s criterion here, as that’s less important for our purposes and I’ve addressed them elsewhere.

What matters for us is that Popper’s criterion preserves empiricism as a distinct position even after Hume’s insights. We saw that there are rationalist components in science, and therefore to be an empiricist in the old sense is an untenable position. One who relies solely on observation won’t be able to rely on anything. But after Popper, there is a more moderate formulation of empiricism that can replace its classic definition: claims need not be based exclusively on empirical observation, but they must be empirically falsifiable.

Extreme views like logical positivism hold that claims that don’t stand up to tests of refutation don’t claim anything at all. According to Carnap, for example, these are pseudo-claims. More sober approaches understand that such claims do say something about the world, but they see no value in dealing with them because we have no (empirical) way to examine them. They are empty of relevant content (which for the positivists is only empirical content). One must understand that claims that don’t subject themselves to tests of refutation in effect make no concrete claim about the world: nothing we could detect in some situation we encounter—or even in a situation we could create artificially in a lab experiment.

Updated empiricism no longer claims that every claim must be based solely on observations without a priori, logical components, but it does require every claim to subject itself to a test of refutation. In other words, it demands that a claim have practical implications. The extremists (logical positivists) will say that without this it’s meaningless. Others will say that without this it’s not reasonable and not admissible. This is the contemporary version of empiricism (though of course many remain stuck with its older versions).

The Status of Belief in God: Rationalism vs. Updated Empiricism

I opened my remarks in the debate by saying that belief in God is, in my view, a factual claim. Moreover, it’s also based on observational facts. I further argued that it’s entirely rational—but definitely not scientific. I see no way to subject belief in God to tests of refutation (see the claims about His providence and involvement in the world—claims far easier to refute, and yet no one is truly willing to give them up or even to try examining them empirically). Note that this is a decidedly non-empiricist approach. I’m asserting a factual claim and am willing to accept it even though it does not meet Popper’s criterion of scientificity. And after all that I also claim it’s rational. This is a decidedly rationalist position, and it seems to me that this was at the heart of our disagreement in the debate—not belief in God itself, which served merely as an example or a special case of rationalism.

What, then, grounds the rationality of claims according to my rationalist approach? If it isn’t necessary to rely on observations alone (without a rationalist component), then what is? I claim that reasoning—or cognitive reasoning (intuition)—is itself a sufficient basis for factual claims about the world. Not only do I reject classic empiricism, which sees observation as the be-all and end-all, I even deny updated empiricism that accepts only factual claims that are falsifiable. In my eyes, belief in God as a factual claim is rational because it can rest on a cogent, a priori philosophical argument.

At the beginning of the debate I explained that my self-assigned goal was modest: not to prove the existence of God, but to show that belief in God is rational—that there is a reasonable and admissible rational way to arrive at it. In short, my goal was to show that rationalism is rational. For my purposes here, the atheist can remain in his position even after my argument (though in my opinion that would be most illogical and implausible), since my purpose here is only to show that my stance is rational (even if someone thinks it’s not true). I’ll add that I also think atheism is implausible and illogical (and perhaps not even rational), but that’s a matter for another debate. That’s why in the debate we didn’t get into the various objections to the physico-theological argument I presented, since my aim wasn’t to show that it’s correct and convincing but only that it’s rational—that a rational and reasonable person can (and in my view should, though that wasn’t the topic) accept it.

Causality

The final stage of the preface is the place of the principle of causality on this map (the question also arose in the Q&A here). The physico-theological argument I presented rests on the principle of causality, and the main dispute revolved around it. But in the subtext, the dispute was really about whether this is an a priori or an a posteriori principle—whether it’s the product of thought or of observation. This only illustrates the more basic dispute: whether it is rational at all to accept principles that are not based on observation, solely by force of reasoning or intuition.

To understand why this is in fact the core of the discussion, we must return to David Hume. Causality is one of his examples of a fundamental scientific principle that has no observational basis. We have no way to ascertain that event A is the cause of event B. For example: when I kick a ball, it flies. Does that mean the kick caused the ball’s flight? Perhaps—but the observation by itself doesn’t say that. What we saw is that there was a kick and immediately after that the ball flew. Moreover, we see that this is always the case, i.e., whenever one kicks a ball it flies. From observation one can derive that there is a correlation between the kick and the ball’s flight and that this happens always (this is the logical component of the causal relation). One can also derive that the flight occurs after the kick (the temporal component of the causal relation, which in my view isn’t necessary, as I also argued in the debate). But one cannot derive that the kick produced the ball’s flight—that there is between these two events a relation of production (this is the physical component of causality). I discussed causality’s three components at length in the series 459 – 466, so I won’t go into it again here.

What matters for us is that one of the most fundamental principles in science—the principle of causality (which states that everything has a cause)—is not the product of observation. Whence, then, do we derive it? From intuition, or from our reason. When we look at the world and discern cause–effect relations between events, the determination of those relations is not the product of observation but the foundation upon which observation stands. Causality is a principle we bring with us to science, not something we learned from science. This is a central philosophical insight for understanding the debate, since much of it revolved around this.

On Atheism and Scientism

From the picture I’ve described thus far, it clearly follows that an empiricist stance is very problematic. If you cling to science and place your trust in it, then you must accept reason as a source of insights about the world—and hence you cannot reject rationalism. Note that I’m not speaking only of classic empiricism (which is simply a misunderstanding), but also of updated (Popperian) empiricism.

Therefore I am a rationalist, despite (and perhaps because) I have great trust in science and its findings. I am not an empiricist, for despite that trust I don’t see science and observation as the be-all and end-all. For me, even claims that cannot stand up to tests of refutation can be true. Common sense and rational thinking are also a solid basis for factual claims—certainly when those claims are supported by observations (even if not only by them, as in science). We have quite a few insights about the world that aren’t based on observation (like the principle of causality, for instance), and we all accept them. Ignoring this is the approach dubbed “scientism,” which in my view is a form of denial.

For a long time I’ve felt that atheists tend toward scientism. They are essentially empiricists—i.e., they see observation as the be-all and end-all. The extremists among them adhere to classic empiricism, rejecting out of hand any claim not based solely on observations. The moderates espouse updated empiricism, realizing that observations alone can yield hardly any interesting claim about reality, yet they are prepared to accept only claims that are falsifiable. I must say that most of those I’ve met don’t even distinguish between these two types of empiricism. I, as noted, oppose both views.

Belief in God is merely an example of a more fundamental dispute between rationalism and empiricism. For me, belief in God fits neither of those two categories: it does not arise from observations alone (but, as we saw, neither does science), and it is not falsifiable—but it is certainly admissible and rational because there are good (rationalist) arguments in its favor. The typical atheist rejects this, for in his view a rationalist claim is invalid in principle. That’s why in my opening I emphasized that there are two toolboxes one can use in discussing belief in God: the scientific–observational (empiricist) and the philosophical–a priori (rationalist). I ruled out using the former (because belief in God is not a scientific claim), but insisted that using the latter is eminently rational.

In light of what I’ve said here, one can put it this way: the dispute over whether belief in God is rational is actually based on tacit agreement that belief in God is rationalist. The more fundamental dispute is whether rationalism is rational or not.

On Scientists and Scientism

Let me add parenthetically that scientists—even the most prominent and intelligent among them—are usually not trained philosophers. Many of them recognize this and do not engage in philosophy, but some think that if they are good scientists they can also make clever philosophical claims. I once recounted here (see columns 40 and 331) an incident I experienced firsthand. Many years ago I took part in a conference at Bar-Ilan University, and in one of the talks an elderly American physicist spoke on some philosophical topic (I no longer remember the talk’s subject or even the conference). At the end, Itamar Pitovsky z”l (who had studied the sciences but was a philosophy researcher at the Hebrew University), who was in the audience, stood up and said to him sarcastically that philosophy, too, is a profession. That’s how I feel about figures like Dawkins, Einstein, Hawking, and others. Scientists, some of them very gifted, can say quite a lot of nonsense about philosophy, and the scientific aura gives them the confidence and status that lead them to very assertive pronouncements in a field they don’t understand.

The problem among “scientists-in-the-pejorative-sense” is that many of them are laymen in the scientific field and even more so in philosophy, which leads them to follow those scientists who do opine on philosophical matters and to adopt their philosophical stances (especially their empiricism). This leads them not only to use scientific findings but also to attach to them very problematic and imprecise philosophical interpretations due to lack of philosophical skill. Incidentally, this is equally true of believers who use statements by believing scientists in the same superficial ways.

In any event, for our purposes I claim that many atheists belong to the category of empiricist scientism. In the name of science and empiricism they deny philosophy and a priori arguments and therefore unjustifiably dismiss good arguments. From here on I will try to show the empiricism peeking out of Aviv’s remarks throughout our debate and the fallacies that follow from it. Note that my aim is not to examine the arguments in themselves, but only to use them to expose the empiricist fallacies hidden behind them. I will not revisit the flaws in empiricism or why it is unreasonable, since I’ve covered that above. So as you read the rest of my remarks, keep all we’ve seen thus far in the background.

I will now walk through the debate in the order of its stages and try to show this through Aviv’s arguments.

The Debate: Aviv’s Opening

I think the principle appears in all its glory already in Aviv’s opening. He claims there that he doesn’t believe in a god or gods because he has never encountered one. This is in itself a conspicuously empiricist statement.

Then comes the dragon example. Aviv claims there is a difference between the claim that there’s a computer under the table and the claim that there’s a dragon under the table, because dragons don’t exist (after all, we’ve never encountered them—remember?!). After my clarification, it seemed at first that there was no real disagreement between us. I agreed with him that someone’s testimony that there is a dragon under the table would not convince me, unlike testimony that there’s a computer there. That’s entirely obvious. But if there are evidences or indications that dragons exist, I won’t deny them just because I’ve never encountered a dragon, and consequently I’ll be willing to accept that there is a dragon under the table. Therefore there’s no real difference between the claim that there’s a dragon and the claim that there’s a computer under the table, as long as they’re based on arguments. This very distinction—and especially the sweeping, mistaken form it took in Aviv’s phrasing—reflects, in my view, empiricism. It seems he didn’t consider the possibility that indications for the existence of dragons might be brought even though we’ve never seen such things. Such evidences, by their nature, wouldn’t be empirical but philosophical (that’s how one treats dragons: they’re invisible), not falsifiable—and so, at least in his initial formulation, he rejects them. From this you can understand that even after we agreed, it seems to me our agreement wasn’t complete. He meant that if there were empirical evidence for the existence of dragons, he would accept it. But I argued that philosophical evidence is admissible as well. Again—empiricism.

Consider, for example, a case in which before us is a table that floats in the air. We have no explanation for this wonder, and someone proposes that there is a transparent object (a dragon) beneath it that holds it up. Is this like the dragon question? Seemingly yes. Aviv would apparently not accept that claim. But assuming there is no other explanation, this recourse is compelled. Here the proposer isn’t only asserting that there’s a dragon under the table—a claim I would also not accept—but is proving it from difficulty in the facts. He has an argument, not just a claim. Aviv would presumably reject this explanation by claiming we’ve never seen transparent objects (which shouldn’t bother us, of course, since transparent objects can’t be seen. But for the sake of the discussion let’s assume here that we also haven’t sensed them in any other way). Is that reasonable to you? Certainly not. Is it a scientific proof? No. Empirical? Also no. Is it falsifiable? Again, no. And still, this seems like very strong evidence (again, assuming there is no other explanation). At the very least, one must concede that this is a thoroughly rational consideration.

Next in his opening comes the courtroom analogy. Aviv casts himself as a juror who must be persuaded that the defendant committed the offense attributed to him (i.e., that God exists). The starting point is that He does not exist, and evidence must be brought to change that. Of course that’s true in my view as well (a claim to the existence of something requires argument or justification), but the question is how the juror can be convinced. It seems to me that underlying the courtroom analogy is the assumption that a court requires empirical evidence. A philosophical argument will not suffice to convict a defendant. Therefore this analogy itself, in my view, reflects latent empiricism. Aviv argues that, for this reason, the same is true regarding God’s existence: this is the only thing that could work.

To this I must say two things: (1) Even in court I’m not sure this is the case (this takes us to the issue of circumstantial evidence; see columns 224 – 226, and more). (2) We are not in a courtroom but in a philosophical discussion. Therefore even if a court requires empirical evidence, in a philosophical discussion this need not be the case. Later on he even says so explicitly, when he argues that to prove something exists one must show that it appears in reality in some fashion. And again, I assume that by “appears” he means the possibility of observing it—not a philosophical argument that indicates its existence. The courtroom analogy continues for him, since in court the starting point is that the defendant is innocent unless proven guilty. So too for us: God doesn’t exist unless it’s proven that He does. And again, I agree with that; the important question is what “prove” means here. Only observation? In my view, certainly not.

His opening continues with an argument that again expresses his empiricism. He suggests imagining two possible worlds (a hint at modal semantics; see column 580): a world with God and a world without, and claims that to assert that God exists we must show that in a world with God something happens that differs from a world without God. He adds that even if we don’t refer to science, this must be something observable that appears in reality. For our purposes it makes little difference, since there is no principled difference between science and observation. But for Aviv, the entire realm of admissible claims belongs to the empiricist category; therefore, even when he speaks of something outside science, this too, for him, is within science. He adds that in his view I won’t bring such evidence, because I said I am speaking of logical arguments. Behold: he expects direct observations—i.e., he assumes empiricism (indeed, this seems like naïve classic empiricism). He ignores what we saw above: even a scientific theory isn’t based on direct observations, but processes empirical findings with tools of thought—i.e., it has a rationalist component.

In my opening I explained that my evidence would not be via direct observations (since God is not seen), but there would be indirect observations with philosophical processing of them—precisely as is done in science (and as in the example of the transparent dragon beneath the table). The empiricism that Aviv expresses here is naïve and classic—i.e., reliance on direct observation. As we saw above, no scientific theory meets that criterion, and certainly one should not expect it of belief in God. If one insists on adopting empiricism (which in my view lacks basis and logic), one ought at least to choose updated empiricism.

Toward the end of his opening is another remark in which he explains why a logical argument doesn’t satisfy him. “Logic is something we invented… reality is the arbiter,” he says. Again we see the characteristic empiricist disparagement of philosophy, viewing whatever is non-empirical as a human invention or subjective construction. This is, of course, mistaken, for as I told him in my reply, logic is the most objective and stable thing we have—far beyond science and observation. If we’re already speaking of possible worlds and human inventions, as he himself put it, I can imagine a world where science is different, but there is no world—even imaginary—where logic is different (I discussed this in columns 549–550, and in my article here on Occam’s razor). So I could accept (even if not agree) claims that science is something we invented (i.e., that the laws of science are merely our way of arranging the array of phenomena—there are such views in the philosophy of science), but to claim that logic is an invention is simply a misunderstanding.

Later he mentions that we agreed in the preliminary discussion to rule out Spinoza’s pantheism. But he rejected it because the pantheist claim has no empirical implication, whereas I rejected it because it says nothing (I intend to devote an upcoming column to pantheism). To say that “God is nature” is merely to call nature by another name. It still depicts a picture in which only nature exists and nothing besides it—which, for me, is atheism in disguise. Again, Aviv as is his wont returns to empiricism, even where a simple, a priori conceptual argument could have brought him to the same goal, more easily and with greater precision.

Later we had a dispute over whether the fact that belief in God obligates us in various ways is relevant to the question of whether He exists—or to the strength of the arguments required for His existence. He said yes; I said no. He said that in his view belief is a tool for decision-making (what to do or not to do). But in my opinion that claim is absurd. You can say the issue doesn’t interest you if it has no practical implications, but you can’t say that if it has no practical implications for you then it’s not true (that’s a flawed logical positivism)—or alternatively, that if it has practical implications you won’t accept good arguments for God’s existence. Even the claim that may sound more reasonable—that the strength of the evidence should be proportionate to the obligations the conclusion imposes on me—is, in my view, incorrect. At most one can say that this evidential strength brings me to the conclusion that He exists, but doesn’t suffice to obligate me to obey the commandments given in His name. That I’m prepared to accept (it’s a kind of “palginan” that appears also in halakhah: there is evidence that a husband died sufficient to permit a woman to remarry but not sufficient to allow his heirs to inherit him—to descend to the estate). But if there is good evidence for His existence, then the conclusion is that He likely exists. At most you could say you are not sufficiently convinced to keep His commands. I brought this up because here, too, there is a practical aspect (though not an empiricist one) that Aviv smuggles into the theoretical discussion, and here too I disagree. I’ll only note that in the preliminary discussion it was clarified that our topic was a deistic, not a religious, God; therefore the question of whether He obligates or not—even if I accepted its relevance—should not have entered the discussion at all.

To conclude the analysis of the opening, let me recall that we agreed the topic was not “proofs” for belief (since proofs always depend on basic assumptions, and those will always be disputed in a way that’s hard to resolve in such a debate), but a more modest goal: to show that there is a rational path to belief. In his opening Aviv spoke of the need for empirical proof, but in light of this note, it’s important to grasp that this is a much stronger claim than it sounds. Not only is he an empiricist who rejects a priori philosophical arguments, but in his view such arguments are not rational. Essentially he’s claiming that the entire rationalist philosophical tradition is not rational. Later I explained that by his form of engagement even Hume and Kant—and indeed almost all the great philosophers—were “irrational.” If so, I’m in good company. To sharpen the point: had I wanted to prove belief in God, relying on philosophers—however great—would be an ad hominem (or ad populum) flaw; but if I merely want to show that belief is rational, one could say that if we’ve reached the conclusion that I’m no less rational than Leibniz, Descartes, or Kant—then I rest my case.

The Debate: The Physico-Theological Argument

My argument in a nutshell was the physico-theological argument. In brief form I presented it thus:

Premise A: Regarding all the objects in our experience, if they are complex there ought to be a cause for this (complexity doesn’t arise by itself).

Premise B: Our world is very complex.

Conclusion: Our world has a cause.

In another formulation I put it this way: If our world ought to have a cause, we must discuss the nature of that cause. If it is not of the sorts of things in our experience—then we can stop the explanation here and say that this is God and He has no cause that created Him. If it too requires a cause—we repeat the process for the second cause. In the end there are only two possibilities: (a) There is an infinite chain of causes. (b) The chain stops at some link that is not of the sort in our experience and does not need a cause. Since an infinite chain is not an explanation (an infinite regress is a fallacy), this is a proof by contradiction that there must be a first link that stops the chain. I call that link “God.” QED.

In my book I pointed out that this mode of reasoning parallels scientific reasoning entirely, since it starts from observational facts (the world’s complexity) and infers from them a theoretical conclusion regarding abstract entities—exactly like the transparent dragon under the table. The fact that the table is suspended in the air is an observational basis for the conclusion that there is something transparent under it holding it up. But the conclusion I reach in both cases is not scientific, since it cannot be subjected to a test of refutation. That, however, is not a defect regarding it, since we’re dealing with a domain that by its very nature is not given to our experience. We were not supposed to observe God (or the dragon) directly; therefore not observing Him is not a shortcoming.

One can quibble here with all sorts of objections and responses, and I elaborated on them at length in my book HaMetzi HaRishon (in the third dialogue and a bit in the second). I won’t go into that here, since my concern isn’t to prove God’s existence but to examine our discussion of this topic, especially its empiricist aspect. It is important, however, to mention that in my book I expanded this argument and applied it to the laws of nature themselves. One may claim that the laws explain the world’s complexity (evolution, the Big Bang, and so on). But even if the laws explain the world, an explanation is still needed for the laws themselves (who legislated them). In my view, this is the updated physico-theological argument, and it deals with laws rather than phenomena. I showed there that some of the objections to the argument that deals with phenomena do not stand against the argument that deals with laws. I can’t elaborate further here, but later on I will comment again on this point.

Aviv raised the claim that we can’t say our world is complex if we don’t know other worlds with different degrees of complexity. In his view, complexity is relative. In my book I explain the error in this claim. There is a mathematical-scientific measure of complexity defined objectively with no need for comparisons: it’s called “entropy.” Low entropy expresses high complexity, and vice versa. Thus, for example, living things have very low entropy compared to nonliving things (which are also very complex). Therefore, when one looks at our world one sees that it contains highly complex beings (like animals), and my assumption is that it is implausible that they arose by a random, blind process. There must be a guiding hand involved (in thermodynamic terms: the system is not closed).[1]

So we are dealing with a world that is complex to a degree that is implausible without a guiding hand. The principle of causality points us to the existence of such a guiding hand. Note that this is a cause for the complexity, not for existence as such (this is the difference between the physico-theological argument and the cosmological argument, in Kant’s terminology). From this it also follows that the principle of causality applied to the world and its laws is not dependent on science or observation. This is a purely logical (indeed, statistical) consideration. A complex system does not arise by itself in a blind process without a guiding hand. This does not derive from observation but from reason, and therefore it does not necessarily apply only to objects of our observation. The cosmological principle of causality (as opposed to the physico-theological one) says that every object has an efficient cause that brought it into being (and not merely a cause that engineered it, as in our case). Here there would perhaps be room to claim that this isn’t true regarding objects outside our world—though even that is, in my view, very dubious—but the principle that denies the chance-formation of complexity is pure reason.

The Debate: My Rebuttal

Already at the outset of our exchange appears the main problem in Aviv’s arguments. It seems he didn’t understand the concept of proof by contradiction. He repeated again and again that if I get stuck in an infinite regress, I cannot solve it by positing a different kind of entity. That only means I have a problem, and one doesn’t solve one problem with another. But he is mistaken, and doubly so: (a) I’m not in a problem. (b) The “problem” wasn’t solved by introducing another. I’m not in a problem because the argument I presented is nothing but a standard proof by contradiction, in mathematics and beyond. One shows that one option leads to a contradiction (or is highly implausible), whence it follows that the alternative is correct. In Sherlock Holmes’s words (in The Sign of the Four): after we’ve eliminated the impossible, whatever remains—however improbable—must be the truth. For our purposes: if indeed infinite regress is a fallacy, then we are compelled to conclude that the regress stops. QED.

Aviv’s second mistake was that even the way I “solved” my alleged problem is not problematic. The conclusion that at the head of the causal chain there is an entity that does not require a cause is not problematic in any way. To explain what he takes to be the problem in my proposed resolution (God as the link that stops the regress), Aviv accused me of the logical fallacy of “special pleading.” I have no small experience in logic, and to my sins I do not know of such a fallacy. But even if there is such a fallacy, it certainly didn’t appear in what I said there. My saying that God is a special case—an object of a kind that does not require a cause—is not an assumption I’m adding ad hoc to protect something (that might perhaps be seen as a fallacy—and even that only with great effort), but the necessary result of the proof by contradiction. It is the conclusion of a valid logical argument. So where is the fallacy? I did not assume it; I proved it.

In passing I’ll add another remark, though according to my view it’s unnecessary. He is also mistaken in thinking that on my account this is necessarily a special case. In principle, other such objects may exist. God is such necessarily (otherwise one falls into an infinite regress), and the objects in our world are not such (simply from our acquaintance with them). But there could be many additional objects that either are or are not such. Thus, for example, when we see indirect indications for the existence of dark matter (an example he himself later invoked), we infer that there exists in the world a different kind of matter. Is that also a special-pleading fallacy? When there is evidence for another kind of objects, we must infer that there indeed is such another kind. What on earth is fallacious here?

Note that here too empiricism sneaks in through the back door, since regarding dark matter even Aviv will agree it’s a special kind. So what bothers him in my analogous claim about God? That unlike dark matter, the evidence I adduced is not empirical but philosophical. Beyond that, he also implicitly assumes that all objects are of the kind familiar to us from observation, and therefore he thinks that when I take God out of that framework He is necessarily a special case. Again: empiricism. But, as noted, none of this is important, for even if God were on my view indeed a special case, there would still be no fallacy (if such a fallacy exists at all), as I explained above.

Moreover, Aviv does not accept that an infinite regress is a fallacy. He didn’t really manage to explain why, so I’ll just note that in the second dialogue of my book HaMetzi HaRishon I explained this well (briefly: an explanation of the “infinite-regress” type posits an actual, not merely potential, infinity, and is therefore an evasion of explanation, not an explanation—cf. “turtles all the way down”). What he did explain was that an infinite regress requires infinite time; therefore, if we have such an infinity, there is no problem in regress-type explanations. As noted above, this is simply incorrect. Beyond that, it’s also not true that an infinite regress requires infinite time. It could be an infinite series that converges to a finite time (think that the first link in the causal chain takes one second, the second half a second, the third a quarter, and so on. The entire infinite chain will take two seconds in total). Beyond this, as I pointed out to him in the debate, he assumes there must be some time interval between cause and effect. That itself is not a necessary conclusion, and at least in ancient Greek philosophy it was held not to be possible. If there were such a time interval, there would be a state in which the cause occurred while the effect had not yet occurred—but the cause is sufficient for the effect. I won’t get into this sophistry here (it’s the idea of the continuity of the time axis, similar to the convergent-series argument; in calculus one can see that in principle the wicked Greeks are right, and nonetheless there can be a time gap between cause and effect if we see it as a continuous process while observing its two endpoints). But this suffices to show that his remarks are far from plausible.

These numerous logical mistakes point to a problematic mode of thinking; and again I will argue that the focus lies in empiricism. The characteristic disparagement of logic and philosophical arguments is typical of one who always seeks empirical corroboration. Such a person doesn’t understand (or doesn’t accept) that one can justify claims about the world also by logical argument. And in general, lack of logical skill is itself a typical trait of empiricists.

The Debate: The Discussion

The core of our discussion revolved around the principle of causality. My claim was that causality did not arise from observation (following Hume and Kant; see above), and therefore there’s no reason to apply it only to things in our experience. On the contrary: since it’s an a priori rational principle, it will apply in every context—unless proven otherwise (as with God, where we proved it by contradiction). Aviv repeated again and again that one cannot apply causality to things outside our world.

Further on, Aviv repeats the claim “Who says there’s something out there?!” I’ve already explained the error in this claim above. No one “said so,” and I’m not asserting it, either—I’m proving it. The physico-theological proof shows that there is something outside. He treats this as though I’m merely asserting it and therefore wonders why he should accept it (as with the dragon under the table), and I explain to him that I’m not asserting it but proving it (by contradiction). When I prove the existence of a dragon under the table, or when there are indications of such, the dragon is exactly like a computer. But for Aviv, a philosophical argument is not a proof nor even an indication, and this is presumably why he ignores it.

But even his very argument about the scope of the principle of causality involves misunderstanding. First, because one can certainly apply the principle to what is beyond experience, since causality doesn’t arise from observing the world in our experience but is a product of our reason; therefore, its scope is any context our reason deals with, not only objects in our world. Beyond that, I did not apply the principle to objects outside the world. I am seeking an explanation for objects in our world, and even according to him they are subject to that principle. The cause I propose (God) is outside our familiar world, but the principle of causality is here applied to objects in the world itself. True, in the chain of explanation I continued to apply it also to links outside the world, but only by way of two possibilities: either there is a cause, or there is no need for a cause. Once we reach a link that doesn’t need a cause—that’s God. Therefore this claim reflects a misunderstanding of my argument.

Toward the end of the debate, Aviv claimed that even if causality is not based on observation, it still says nothing until we apply it to our world. He explained that it’s like geometry which, by itself, says nothing (like logic). I explained that he’s again mistaken, and again the problem is his empiricism. Mathematicians distinguish between a theory and a model. The mathematical theory is true in itself, irrespective of any reality and without relying on observations (unless you’re an Platonist who speaks of contemplation in the world of ideas; see columns 383 – 385, 435, and more). But a mathematical theory can have different models—i.e., different contexts in which it can be applied. Geometry is not our invention as Aviv thinks (as we saw above regarding logic). It is true in itself. It’s a Platonic truth that doesn’t derive from observation (and is therefore not falsifiable). Its applications to the world are factual claims and depend on observation. In that sense I claimed that causality is analogous to a mathematical theory, true in itself (it is a priori). But indeed, it can also be applied to our world. To the same extent, however, it can have other models—i.e., other contexts in which that a priori theory applies. On the contrary: since it is a rational truth, it ought to apply in every context whatsoever, not just our world. Causality is indeed applied to the world, but it is not derived from it and is not part of science (contrary to Aviv’s claim). In mathematical terms: causality is a theory, and the world is a model that satisfies it. To say that geometry and causality are part of science, as he claimed, is a typical misunderstanding of empiricists who are unwilling to accept claims that are not facts about the world and who don’t recognize other truths. (This also connects to my columns on philosophy; see the end of this column.)

Moreover, Aviv also claimed we must distinguish between finding a cause that produces something ex nihilo and a cause that produces it ex existentia (in kabbalistic terms, the distinction between creation and formation). I replied that I see no logic in this distinction. In our world, if he were to see an object suddenly appear, he would assume that something produced it—not that it simply appeared. That is, in our world even Aviv does not distinguish between these two kinds of causality. So if we accept that causality holds also in contexts outside the world, again there is no justification to distinguish between these two subtypes. And certainly one cannot say that the assumption that there ought to be a cause for the world’s existence—and particularly for its complexity—is irrational. I remind you that I did not come to prove that God exists but only to show that one can reach that conclusion by a rational route.

In any case, we again see the empiricist clinging to experience. Aviv insists repeatedly on seeing causality as a principle learned from experience, whereas it’s clear that this is not so. The great philosophers like Hume and Kant already pointed this out. It is that same scientistic disparagement of philosophy that sees empirical science as the be-all and end-all. As we saw above, Hume and Kant taught us that the truth is the opposite: even the use of causality in science is not based on observation but on an a priori reason; therefore there’s no reason to confine it precisely to our experience and to science.

When I asked him on what he bases causality—what, in his view, justifies applying it to things in our experience—he repeated that “it works.” What is that if not experience? I couldn’t get him to explain the difference. Note that there is circularity in this justification. One cannot ground causality and induction on themselves. Causality “works” only because we see the causal relation as including a component of production. If we didn’t see it that way, it would also “work”—we would simply speak of correlation rather than causality (this is precisely what Hume argued). It’s akin to trying to prove the principle of induction on the basis of induction (because it worked in the past). Hume already pointed this out and explained that this is circular: you want to prove the principle that what worked in the past will work in the future, and you do so on the basis of that very principle. But, as noted, atheists tend to ignore logical and philosophical arguments.

Further on he claims that the laws of physics as they are suffice to explain the world, and therefore there’s no reason to invoke the existence of something else. Hence he claims that no explanation is required. When I asked what the cause of the laws of physics is (see above that the updated physico-theological argument refers to the laws and not the phenomena), he said it’s a fact and as such needs no justification. I must say I found it very odd to hear this from someone speaking in the name of rationality. As I said in my summary, I am offering an explanation for the complexity and for the laws; he claims no explanation is needed because it’s a fact. To say that his alternative—no explanation—is the more rational one, as opposed to the one offering an explanation, is one of the oddest things I’ve heard. Note that the dispute is not over whether we know the explanation. His claim here is that there is no explanation and none is needed—not that we don’t know it. Is that a rational stance?!

Some of the time he claimed what we’ve just seen—that no explanation is needed, and I addressed that. But at other times he claimed that maybe there is another explanation besides the one I’m proposing (God). Here again is a mistake. I didn’t propose any specific explanation. I simply argued that there is an explanation—i.e., an efficient cause that produced the laws. I refer to that cause as “God.” I didn’t say anything else about it beyond that it is the explanation of the laws. So what does it mean that there is “another explanation”? Does he mean it’s not Shakespeare but his cousin who was also named Shakespeare?! He claims that this cause shouldn’t be called “God” but “Reuven.” Fine—call it Reuven. The question is whether there is a cause, not what the cause is called. The empiricist is unwilling to accept non-observational explanations, and therefore prefers the stance that there is no explanation. This is a clear expression of the irrationality of empiricism.

Toward the end, Aviv conflated non-necessity of a claim with its truth. I explained to him that indeed, in my view nothing is necessary; but that doesn’t mean that no claim is true. This connects to another mistake of those with an analytic bent (in my books Shtei Agalot and Emet Ve-Lo Yatsiv I explained the relation between this and positivism/empiricism). Empiricists who place full trust in observation tend to identify truth with necessity and necessity with empiricity. But that’s a categorical error. As I explained to him, he is mixing up modal logic with fuzzy logic. In modal logic we speak of different possible worlds; in fuzzy logic we speak of degrees of certainty of claims. It is true that for a non-necessary claim there is a possible world in which it does not obtain, but that doesn’t mean it is not true (in our world). It is true—but not necessary.

I’ll end with Paley’s watch. The Reverend Paley argued that if we were walking down the road and saw a watch on the ground, we would surely assume there was a watchmaker who made it. So the world—which is far more complex than a watch—surely attests to someone who created it. Aviv here presented the common atheistic critique according to which one can say this only because a watch is an artificial object. Regarding a plant on the roadside, Aviv told me, you wouldn’t ask the same question. From this he concluded that the world—which is of course a natural object—is irrelevant to the question of who made it (here he assumed that there is no cause, not that there is a different explanation. As noted, this shifted for him, apparently unwittingly). But of course this is a mistake, whose root is again his empiricism. I explained to him that I certainly ask the same question about a plant and a rock on the roadside, since their complexity too is so great (their entropy so small) that it’s implausible they arose just like that. In the updated formulation of the physico-theological argument, the laws that produced them themselves demand an explanation. What’s the difference, then? Regarding natural objects, I have an explanation: God. He created the world, its laws, and all that is in it (the world’s complexity points to this; that is the physico-theological argument). By contrast, when I see an artificial object, it’s clear to me it wasn’t created directly by God; therefore I wonder about it and infer that there was a watchmaker who made it. That is: there is no real difference between a watch and a tree or a rock. In both cases we have complexity that demands explanation. In the natural case the explanation is God, and therefore the question doesn’t arise; in the artificial case a question arises and a specific explanation is required for that case. But atheists don’t seek explanations for reality, since those would necessarily be found outside it. To see this as a rational stance is, to me, most puzzling.

The Debate: Concluding Remarks

In his summation, Aviv repeated that I am positing the existence of something outside (material) reality, and that is the main point he disagrees with. For some reason he chose to ignore what I explained more than once in our discussion: that this isn’t an assumption but the result of a logical argument (a proof). I showed that there is a “dragon” by force of the physico-theological argument; I didn’t just assume or assert it. Therefore his discussion of the null hypothesis—i.e., the starting point we remain at if no persuasive arguments are brought—is irrelevant.

In my own summary I argued that the physico-theological argument I presented is an eminently rational mode of thought. It assumes the principle of causality, as every reasonable person (and almost every great philosopher) does. And it is certainly more rational than a picture in which there is no explanation (not merely that we don’t know it) for our complex world. Aviv also argued that perhaps there is an explanation other than mine—but here he misunderstands my argument, for I didn’t assume anything besides the existence of some explanation. I didn’t speak of a specific explanation.

Given that my aim wasn’t to prove that God exists but only to show that there is a rational path to belief, I think I fully shouldered the burden of proof. Aviv’s resistance to the conclusion that there is a God is irrelevant to the discussion, since his task was to show that the path I proposed to that conclusion is not rational.

As I’ve shown here, the roots of almost all the flaws in his arguments lie in the empiricist worldview he assumes. Usually it’s naïve, classic empiricism; sometimes it seems more self-aware (updated empiricism). But as I’ve explained, that too is very implausible.

Closing Lines: Atheism, Scientism, and Philosophy

To conclude, I’ll reiterate that my aim in this column was not to discuss whether there is or isn’t a God, but to use the arguments that arose in the discussion to point to a correlation between atheism and empiricism. From my (not inconsiderable) experience, the roots of almost all philosophical disagreements between believers and atheists revolve around empiricism. I think this debate illustrated that well; in that sense it was entirely typical. As I wrote above, we didn’t argue about God but about empiricism and rationality.

At this point it’s important for me to direct you to my series of columns on philosophy (155 – 160), where I showed that empiricists in effect deny the existence of philosophy (or reduce it to science, thereby emptying it of its unique content and methodology). For them there is no such domain of discussion at all. There I also explained where and why they err and showed that philosophy is a field that deals with reality but not with empirical tools (much like the argument for belief in God I presented here). It is based on rationalism—and indeed that is the definition of philosophy (not merely one current within it). Here I tried to show that contemporary atheism is based on empiricism, and that teaches us that denial of God is essentially a denial of philosophy, of thought, and of rationalism (and in my opinion also of rationality—but that’s not the topic). That is the essence of the scientism I described above, which is the principal atheistic flaw—at least of contemporary atheism.

[1] In the aforementioned book I explained why I’m not making here a scientific claim by virtue of the second law of thermodynamics. I’m only using the concept of entropy to clarify the notion of complexity.