by Joseph Shieber
Completely by chance, I happened to come across a discussion of Tyron Goldschmidt’s paper, “A Proof of Exodus: Judah HaLevy and Jonathan Edwards Walk into a Bar”, in Cole Aronson’s review of the 2019 book Jewish Philosophy in an Analytic Age. I was intrigued by Aronson’s celebration of Goldschmidt’s “characteristic verve”, so — with the help of my college’s outstanding Interlibrary Loan — I got hold of the paper.
Just as a matter of literary quality, Aronson undersells Goldschmidt’s paper. Goldschmidt is a delightfully engaging writer. If you’ve dipped into some contemporary academic philosophy and come away with the impression that it’s all turgid and dry, check out Goldschmidt’s essay. It’s a treat.
Now, of course, I’ve got to go ahead and ruin your impression of the delightfulness of academic philosophy by attempting to point out flaws in Goldschmidt’s argument. I can’t help it; it’s in my nature.
Goldschmidt begins by noting that testimony is central to our knowledge. Much of what we know is based on our having learned it from others. If anything, Goldschmidt underappreciates our dependence — the example he uses is historical (our knowledge that Napoleon existed), but he could have easily included countless examples. It’s only because of testimony that I know that Raphael Warnock is a newly-elected U.S. Senator from Georgia, that the top five warmest years on record have been since 2015, or even what my own name is!
Goldschmidt suggests that, by appreciating the centrality of testimony we can appreciate an underrecognized argument for the truth of biblically recorded miracles — in particular, the miracles associated with the Jewish tradition surrounding the Exodus from Egypt.
Building on the work of the philosopher and logician Dovid Gottlieb, Goldschmidt advocates for the Kuzari Principle:
A tradition is likely true, absent evidence to the contrary if … it is (1) accepted by a nation; and describes (2) a national experience of a previous generation of that nation; and (3) the national experience would be expected to create a continuous national memory until the tradition is in place.
Applying the Kuzari Principle to the case that concerns Goldschmidt — that of proving the truth of the miracles that occurred during Exodus — the result is supposed to lead us to accept that it is likely true that the miracles that supposedly occurred during Exodus really did occur. This is what Goldschmidt refers to as the “Kuzari Argument” in favor of belief in the truth of the miracles reported in the Book of Exodus.
Goldschmidt’s argument piqued my interest for two reasons. The first is that my own view on the ways in which testimony can support knowledge bear a superficial similarity to the Kuzari Principle. The second reason is that Goldschmidt’s discussion obscures crucial functions of testimony that go beyond merely providing evidence for belief — for example, the role of testimony in building national identity.
Let’s look at Goldschmidt’s argument from the point of view of epistemology first, from the point of view of whether the Kuzari Principle supports the idea that the national traditions regarding the account of Exodus provide sufficient evidence for believing that the supposed miracles of Exodus really did occur.
My own view, roughly, is that at least some social processes or mechanisms are conducive to reliably producing true beliefs in the people embedded in those processes or mechanisms. Furthermore, I hold that people can *know* those truths by virtue of being embedded in those social processes or mechanisms.
We could characterize the conditions enumerated in the Kuzari Principle as a process — call it the “Kuzari Process”. The Kuzari Process is one that involves (1) a tradition’s being accepted by a nation, (2) that tradition’s describing the national experience of a previous generation of that nation, and (3) that tradition’s justifying a reasonable expectation that the national experience that serves as its origin would create a continuous national memory.
Once we characterize the Kuzari Process in this way, the question, to put things in the terms that are congenial to my way of evaluating the epistemic status of testimony, would be whether the Kuzari Process is one of the social processes or mechanisms that is conducive to reliably producing true beliefs in the people embedded in the Kuzari Process.
In particular, whether Goldschmidt’s argument for the truth of the miracles of Exodus works will depend on whether the Kuzari Process counts as one that reliably conduces to the formation of true beliefs in the people embedded within it, to use my formulation. It is highly questionable, however, that the Kuzari Process counts as such a reliably truth-conducive process.
There are at least three obstacles to any attempt to argue for the reliable truth-conducivity of the Kuzari Process. First, there is the problem of an adequate understanding of “nation”. Second, there is a problem with the notion of national experience. And third, there is a problem with the notion of national memory.
Let’s look at the first difficulty, the challenge in understanding what is meant by “nation” in the formulation of the Argument. Note that we won’t even consider whether it makes sense to speak of “a” Jewish national experience, or whether it makes sense to subsume religious group identities under the rubric of national identity.
Instead, I want to assume for the sake of argument that we can discuss religious memory and tradition as a form of national memory and tradition, and to focus on some conceptual difficulties that remain when appealing to “nation” in this way. In particular, Goldschmidt seems illicitly to be using two different features of “nation” in the argument, features that work at cross-purposes. When Goldschmidt wants to emphasize the reliable truth-conducivity of national memory, he underscores the fact that nations are composed of many individuals, but when he wants to focus on the limited nature of the Argument, he emphasizes the special status of national memory as a form of group belief.
I want to focus on two problems with this strategy. Here’s the first problem. In defending the propriety of relying on the Kuzari Process for justification, Goldschmidt suggests out that it is intuitive that it would be much harder to perpetrate a fraud in public, with many independent observers. It is indeed intuitive that this would be the case – and Goldschmidt cites a number of early Modern thinkers to attest to this intuition. However, it’s far from obvious that this is true.
Recent events would suggest that it is indeed possible to engage in widespread deception of the public, even in the presence of many independent observers. Of course, if the observers are independent they will include dissenters who are in a position to point out the deception. However, Goldschmidt’s argument concerns historical evidence, so one of the most important impacts on whether a deception perpetrated in public is successful will be whether that perpetrator is ultimately successful. The dominant narrative of history – particularly history that can survive millenia – will be written by those who control that narrative.
Suppose, however, that we ignore this worry. For the sake of argument, let’s grant that the presence of independent observers offer a reason for thinking that a national tradition is true. Now here’s the second, central, problem stemming from the definition of “nation”: depending on how we define “nation”, that group might not include any independent observers.
This is particularly pressing given that, as Goldschmidt points out, the phenomenon of group belief allows for the possibility that a group believes something even if the majority of the members of the group don’t believe it.
What this means is that a nation’s beliefs might not coincide with the beliefs held by a majority of the individuals who comprise that nation. So, for example, we might say that the White House believes that Donald Trump won the 2020 US Presidential election even if virtually no individual person working in the White House believes that to be true. Or, to use an example more pertinent to Goldschmidt’s argument, we might say that the nation or people of Israel in 2500 BCE believed the golden calf to be idolatrous even if a majority of the individuals comprising that nation or people were complicit in worshiping the golden calf.
The upshot of this is that Goldschmidt’s argument plays on an ambiguity. The argument is strongest when it allows for the idea that independent, dissenting believers all converge on the same belief. The concept of nation as a group believer, however, is antithetical to this idea: what the nation believes is independent of the dissenters.
This problem – that what constitutes a nation is a slippery idea – is central to the other two difficulties for Goldschmidt’s argument. Let’s tackle the second difficulty: what constitutes national experience?
To see the problem with this, let’s take an historical example not from Exodus, but from the experience of a different enslaved people. In 1850, the population of South Carolina included approximately 275,000 whites and 385,000 enslaved Blacks. What should we say about South Carolina’s “experience” of slavery in 1850?
For much of the 170 years since 1850, the dominant narrative has focused on what white South Carolinians thought about slavery. In fact, even today, surveys of “Southerners’ arguments for slavery” tacitly presuppose that those Southerners are white. But in South Carolina in 1850, enslaved Blacks outnumbered free whites by roughly 3 to 2!
This same issue infects the third difficulty: the problem of how national experience is transformed into national memory. This third problem addresses the issue of how an experience is preserved as memory. Goldschmidt’s suggestion is that, as long as the experience is transformed into a memory, this will be sufficient to insure the truth of the resulting tradition. What this ignores, however, is that not all memories are equally reliable.
To see this, consider another example from United States history: the cult of the “lost cause” of the Confederacy. Until recently in the United States, the dominant narrative of the Confederacy has involved myths of States’ Rights, noble Confederate generals, and the plantation-and-magnolia-tree picturesque images of the (white! rich!) Antebellum South.
Only in the last few years has popular culture caught up to the fact that this cult of the “lost cause” is just that – a cult. It took the concerted effort of propagandists on the losing side of the Civil War to turn the “lost cause” into something that was portrayed as central to the heritage of the very country that the Confederacy was intended to destroy.
Note that this “lost cause” narrative is not — or not merely — reflective of the historical memory of the adherents of the defeated Confederacy. The point of this example is that, for nearly a century, this narrative reflected the popular historical memory of the United States itself! This does not inspire confidence that national traditions provide strong evidence for the truth of the events recounted in those traditions.
I’ve given examples of national traditions, experiences, and memories that don’t seem reliably conducive of true beliefs in their adherents. But this doesn’t mean that NO national traditions reliably transmit true beliefs. Indeed, highly disputatious, strongly textualist religious traditions — not just Judaism, but the Vedic traditions, Mahayana Buddhism, etc. — might arguably be better at including and accounting for dissenting opinions within the nation than some other national traditions. And this, in turn, might lead to a more reliably truth conducive connection between national experience, national memory, and the underlying facts.
If this is true, however, the focus in the Kuzari Principle on nationhood is misplaced. We would be better served by focusing on the specific characteristics of the social belief-forming processes that make them reliably truth-conducive.
So the Kuzari Argument doesn’t succeed. The traditions around Exodus, to the extent that they can be ascribed to a Jewish national experience and memory, don’t provide evidence for the belief in the truth of the miracles associated with Exodus.
In introducing the discussion of the Kuzari Argument, however, I suggested that assessing national traditions on the basis of whether they provide evidence for belief is not the only dimension along which to evaluate those traditions.
One of the first pieces of specifically Jewish evidence for the significance of testimony that Goldschmidt cites is Exodus 13:7-10: “Unleavened bread will be eaten throughout the seven days … You will tell your son on that day, saying: ‘This is because of what the Lord did for me when I went out of Egypt.’ It must be a sign for you on your hand and a memorial between your eyes, so that God’s law will be in your mouth, for with a strong hand the Lord brought you out of Egypt.”
A more nuanced reading of this passage suggests, however, that the significance of testimony in this case isn’t primarily *evidential*. Consider the statement, recognizable to anyone who’s attended a Passover seder, “This is what the Lord did for me when I went out of Egypt” (emphasis mine). As a literal statement, this is false: I was never in bondage in Egypt, so the Lord never freed me. And given this, it cannot be the case that the assertion of this statement is intended to be evidential.
What Goldschmidt ignores, in other words, is that the acquisition of knowledge or evidence is not the only purpose of testimonial exchanges. As I point out in my book on testimony, philosophers often fail to appreciate the extent to which people accept testimony for reasons that have nothing to do with a desire to acquire accurate information. We are only human! We sometimes accept testimony, for example. because we wish to maximize group agreement, or because we wish to avoid controversy, or because we wish to ingratiate ourselves with the testifier, or simply because believing the testimony leads to a more interesting outcome than not believing.
What the passage from Exodus suggests is that one of the central functions of national testimony is not transmitting information but promoting identity. “What the Lord did for ME” serves to make ME part of the tradition. It makes me identify with the Israelites’ deliverance out of the house of bondage.
Once you pause to reflect, you see that so much national testimony serves functions other than conveying reliably accurate information. Does it matter if Washington really chopped down a cherry tree? Not at all — the story not only teaches children the importance of telling the truth, but also implies a connection between the “founding fathers” and upright behavior.
National traditions are suffused with such identity-promoting elements. Consider the formula from the Pledge of Allegiance that the United States is “indivisible, with liberty and justice for all”. If this is intended as a statement of fact then we’d have to judge it to be false. It’s not intended in that way, however — it’s intended as aspirational … and inspirational.
We can see, in discussions of the 1/6/2021 insurrection attempt, the way that this identity-promoting element can run up against the goal of attempting reliably to transmit accurate information. As Paul Musgrave argues (h/t Lawyers, Guns & Money blog), the inability for many commentators accurately to describe what occurred last Wednesday involves a “reluctance to see what’s in front of their faces [that] has had less to do with scholarly integrity and more to do with wish-casting—making predictions because you want them to be true, not because the evidence supports it.”
This post is already too long, so I’ll close with two points about this second, non-truth-directed element of national tradition.
The first point is that I don’t see this element of traditiona as (merely) negative. Interpersonal ties, trust, and belonging are all important for the survival of traditions. And certain fables or falsehoods — perhaps the Washington-chopping-the-cherry-tree example? — are innocuous ways of seeking to achieve those important goals.
The second point, though, is that this element serves even further to weaken the extent to which we might suppose national traditional memory to be a reliably truth-conducive source of belief. This is because the reasons that contribute to the survival of national traditions very plausibly belong more to this second element of national tradition. It’s because a certain belief or practice fosters intra-group ties, trust, and belonging that the belief or practice survives. If it happens also to be true, that’s a happy by-product, but it’s not the reason that the belief or practice still exists. And this fact doesn’t inspire confidence that such national traditions reliably transmit accurate information.