Commanding Knowledge

Commanding Knowledge

Elliot Salinger

Print Friendly and PDF

The term “philosophical” may colloquially designate those questions that are, or at least seem, irresolvable. This usage betrays the popular view that philosophy is a field incapable of making progress. But philosophy has in fact progressed stunningly in the past century, starting with the work of Gottlob Frege (1848-1925), Bertrand Russell (1872-1970), and Ludwig Wittgenstein (1889-1951). Philosophers inspired by their legacy of rigor and precision—adherents of the “analytic school”—have shed much light on obscure philosophical problems.

In this article I’ll discuss a dispute from the Middle Ages that has recently received new attention from analytic philosophers: whether there could be a commandment from God to believe that God exists. In his recent paper “Commanding Belief,” Tyron Goldschmidt convincingly reconstructs and defends Hasdai Crescas’s argument against Rambam that there could not be such a commandment.[1] Goldschmidt’s “translation” of Crescas’s argument into contemporary analytic jargon has enabled philosophers to consider Crescas’s argument afresh.

I will argue here against Crescas and in defense of Rambam. But I will not argue that Crescas’s argument fails. Rather, I’ll first argue that Rambam enumerated a commandment not merely to believe, but to know that God exists. Then I’ll argue that Crescas’s argument, as formulated by Goldschmidt, poses no problem for that view.


Belief and Knowledge

Let’s first get clear on the distinction between belief and knowledge. By “belief” I just mean the attitude of taking some proposition to be true. For instance, I believe that I am seated on a chair, and the Muscovite sect believed that men ought to part their hair. Beliefs can be false, such as the Muscovites’ belief about men’s style. Knowledge comes in several varieties, but here we’ll only discuss propositional knowledge. This kind of knowledge is a species of belief. For the limited purposes of this article we may define knowledge as justified true belief. This means that there are two criteria that render my belief in some proposition to be knowledge. The first criterion is that proposition be true, and the second that I be epistemically justified, or warranted in believing that proposition. Some philosophers, probably including the historical Rambam, have held a stricter view of what counts as knowledge. These philosophers have thought that for me to count as knowing some true proposition, I must be capable of demonstrating the truth of that proposition, where that demonstration explains why that proposition is true. On this view, what it would take for me to know, for instance, the Pythagorean Theorem, is that I have knowledge of the axioms of geometry and be capable of demonstrating how the Pythagorean Theorem follows logically from those axioms.


A Commandment to Believe or to Know?

Now back to Rambam. Rambam enumerates the commandment of Anokhi Hashem E-lohekha in Sefer ha-Mitzvot, and codifies it in Hilchot Yesodei Ha-Torah ch. 1. These two texts have traditionally been read as contradictory since Sefer ha-Mitzvot uses the verb “leha’amin” (to believe) and Mishneh Torah uses the verb leida” (to know)  to articulate the substance of the commandment.[2] However, it seems that the two sources are in agreement. Rambam wrote Sefer ha-Mitzvot not in Hebrew but Judeo-Arabic, and many scholars believe that the Arabic word “itikad” traditionally translated as “to believe” in context really means “to know.” If so, there would be no contradiction between the Sefer ha-Mitzvot and Mishneh Torah to resolve.[3] Furthermore, it seems that Rambam conceived of this commanded knowledge as propositional: knowledge that (to be distinguished from knowledge of in its strict sense of acquaintance or knowledge how). So the commandment is not to know God in some intimate, mystical sense, but just to know a fact: that there exists a Prime Existent (“she-yesh sham matzui rishon”).  

What might a commandment to know that God exists amount to? Recalling our earlier distinction between knowledge and belief, we may say that a commandment to know that God exists just amounts to a commandment to believe with epistemic justification the true proposition that God exists. Or, on the stricter view of what counts as knowledge, I must be capable of offering a demonstration from first principles that God exists, where that demonstration explains why God exists.

Before we continue, a brief note. Analytic philosophers read the great philosophers of the past less to identify what they “really” believed than to find inspiration for good arguments. Here I’ll do the same. So we now take leave of the world of historical scholarship for the world of pure philosophy. This means that in my remarks below I do not claim to represent the historical Rambam’s epistemology with full accuracy.


Why God Could Not Command Belief in God

We’ll now explore Goldschmidt’s reconstruction of Crescas’s argument. Goldschmidt first distinguishes between conformance and compliance. (I’ll use a Greek letter in my presentation, following scholarly convention in the field of analytic philosophy of pretending to be a mathematician.) I conform with a command to ϕ, where ϕ is some action, if and only if I ϕ. I comply with a command to ϕ if and only if I ϕ, where my reason for ϕ-ing relates to its status as a command.[4] For instance, suppose a sloppy teenager’s mother asks him to clean his room. If he cleans his room, no matter his reason (perhaps to find a lost pen), he conforms with his mother directive. If he cleans his room because his mother so commanded, he complies with his mother’s directive. As Goldschmidt clarifies, in the case of a divine command to ϕ I comply with that command if and only if I ϕ because (I believe) God commanded me to ϕ.[5] Exploiting this distinction, Goldschmidt summarizes his reconstruction of Crescas’s view thus:

Complying with God’s command to believe in God means conforming with the command because we believe He commanded it–which means conforming with the command prior to believing in God. But conforming with a command because God commanded it means having a prior belief that God commanded it, which means having a prior belief in God – which means believing in God prior to conforming with the command. So conforming with the command is prior to the belief, and the belief is prior to conforming with the command – which is impossible. Therefore, we cannot comply with God’s command to believe in God.[6]


Goldschmidt is careful to note that the above only shows that it would be impossible to comply with such a command, not that God could not issue it. But if we assume that God only issues commands with which we can comply, then we have an argument for the stronger conclusion.

A slight problem: the argument so formulated isn’t quite right. It’s not true that having a belief that God commanded something means having a prior belief in God. At least, this is not a necessary truth. One can imagine a person who had the belief that God commanded something but not the belief that God exists. Such a person could comply with a commandment to believe that God exists, but would be irrational for holding beliefs that logically contradict one another. But the argument can be easily fixed up if we make the stronger but still plausible assumption that God only issues command with which we are capable of rationally complying.

I’ll now begin to argue that a commandment from God to know that God exists would not, unlike a commandment from God to believe that God exists, fall into vicious circularity. I’ll argue that we are capable of rationally complying with a commandment from God to know that God exists.


Why God Could Command Knowledge of God

Let’s first cash out the difference between conforming with a command and complying with a command. When I ϕ because God commanded me to ϕ, I take God’s having commanded me to ϕ as a justification, or justifying reason to ϕ. Notice that the attitude of taking God’s having commanded me to ϕ as a justifying reason to ϕ also sounds like a belief, and one that would be irrational to hold unless I also believed that God exists. Also notice that this belief about my having a justifying reason will motivate whether or not the belief amounts to knowledge.

We are now in a position to understand how one could rationally comply with a commandment from God to know that God exists. The circularity that perturbed Crescas consisted in the identity of the propositional attitude required to fulfill the command—the belief that God exists— with that required to rationally comply with that command, also the belief that God exists. But since on Rambam’s view the command is to know that God exists, the propositional attitude required to fulfill the command is not identical to that required to rationally comply with that command, still just the belief that God exists.

I’ll now tell a little story to illustrate how one might rationally comply with a command from God to know that God exists. Consider Avi. As a youngster he is taught that there exists a Being we call God, Who created Heaven and Earth, led our ancestors out of Egypt, and gave us the Torah. A rather trusting child, Avi believes everything he is taught without question. As Avi grows up, he comes to re-examine his beliefs. Though he has moments of doubt, when asked Avi affirms that he believes that God exists. And Avi is quite sure that if God exists, then God could command human beings and that those commands would provide humans with justifying reasons for action. Avi wants to come to know that God exists because he has some credence that God commanded him to know that God exists. One day Avi chances across a proof for the existence of God. He studies the proof intently and becomes convinced that its premises are true and that its conclusion follows necessarily from its premises. Avi is now epistemically justified in holding the belief that God exists. Capable of offering a logical demonstration from first principles that God exists, Avi even counts as knowing that God exists on the stricter analysis of knowledge. He has successfully complied rationally with a commandment from God to know that God exists.


Credences and Rational Action

One might entertain the following worry about this account: Wouldn’t it be irrational to act on the basis of a belief that didn’t amount to knowledge? And wouldn’t that render rational compliance impossible even for a command to know that God exists?

This objection, however, rests on a false premise. It can be rational to assume the truth of a proposition while deliberating about what to do even when I don’t have knowledge of that proposition. I don’t even need to believe that it be true, I just need to have a sufficiently high credence that it be true, where the standard for sufficiency depends on other facts about the situation. For instance, say I have a desire for seltzer and there is a cup of liquid in front of me. Suppose further I believe that there is a 25% chance the cup contains seltzer, and a 75% chance it contains water. For some reason I am unable to study the contents of the cup, but I believe it contains nothing I dislike or that would be harmful to me. In this case it would still be rational for me to drink from the cup even though it would be false to say that I believed, much less knew, that the cup contained seltzer.[7]

Could it be rational for a person with only a high credence that God exists to assume its truth in practical deliberation? It would depend on the action one’s deliberation concerned. But one action it would certainly render rational is investigating whether God exists. And if one also had sufficiently high credence in the conditional belief that if God exists, then God commanded me to know that God exists, then it would be rational not only to investigate whether God exists, but to do so in order to comply with God’s command to know that God exists.



In this brief article I have only tried to address one small, technical problem. It goes without saying that important questions remain. How high must be my credence that God exists to make compliance rational? And what sort of justification, on the more lenient view of what counts as knowledge, could turn my belief that God exists into knowledge? Could I gain knowledge of God by acquaintance, or only by some proof? Are there any logical proofs for God’s existence that would, in fact, allow me to know that God exists on the stricter view of what counts as knowledge? And suppose God could command us to know that God exists. Is this really why we should want to know God? Perhaps God only demands compliance for ritual law, whereas for laws whose purpose can be grasped by human reason, God prefers us not to comply. God might prefer that we offer assistance to our neighbors because it is morally right and that we come to know that God exists because it is true.


Though I have not addressed these questions directly, it should be clear that they emerge naturally from my account. And it is clear that answering these questions marks the next step in understanding more fully Anochi Hashem E-lokekha. Jewish analytic philosophy must begin with such modest studies, showing which views might stand up to scrutiny and which will not, paving the way for future work.

But though the belief that analytic Jewish philosophy might soon yield many rich, substantive philosophies of Judaism might not yet count as knowledge, I have a high credence in its truth.

[1] Tyron Goldschmidt, “Commanding Belief,” Ratio XXVIII (2 June 2015): 163-174. Crescas’s argument is found in Or Hashem, hatza’ah (pp. 9-10 in the edition edited by Shlomo Fischer [Jerusalem, 1990]).

[2] Hilkhot Yesodei ha-Torah, enumeration of mitzvot; 1:1; 1:6.

[3] See, e.g., James A. Diamond, Maimonides and the Shaping of the Jewish Canon (New York: Cambridge, 2014), 247 n69.

[4] Goldschmidt, “Commanding Belief,” 166-7.

[5] Ibid, 167.

[6] Ibid.

[7] Max Baker-Hytch has offered a similar argument to show how it might be possible to rationally comply with a commandment even to believe that God exists. He argues that one might have a credence that God exists that is simultaneously low enough not to count as belief but high enough to make acting on its basis rational. His paper may be found here:

The Troubling Trend of Photoshopping History

The Troubling Trend of Photoshopping History

Does Lying Make You A Liar? On Truth And Truthfulness in Rabbinic Thinking

Does Lying Make You A Liar? On Truth And Truthfulness in Rabbinic Thinking