Why the Pieces Do Not Fit
Before weighing the critics’ picture against what we have found, one more piece of their account is needed: when the later books were written.
On the critics’ own dating, several of the books we examined are early. The book of Judges belongs to what scholars call the Deuteronomistic History, which runs from Joshua through Kings. It is usually thought to have been compiled in the seventh or sixth century BCE, but from older material, and the Song of Deborah in Judges 5 is widely regarded as one of the oldest poems in the Hebrew Bible, perhaps from the twelfth or eleventh century BCE. The core of the book of Samuel is likewise thought to rest on early sources. The first chapters of Isaiah are generally assigned to the prophet himself, in the eighth century BCE. Other books are placed later. Ruth is dated by many scholars to the Persian period, and Ecclesiastes later still, while the Psalms are thought to span many centuries. But the oldest of these books are, on the critics’ own reckoning, older than the Priestly source, and in some cases older than Deuteronomy.
Now set the two pictures side by side.
Consider first the overall design. We found a single plan running through all five books: a purpose stated at the beginning, a history told in stages that match it, law placed where it drives the story forward, and the same organizing principle repeated at every level, from the five books down to the sections of a single story. This design does not respect the boundaries between the supposed sources. It runs through J, E, P and D alike. If the sources were written by different authors, centuries apart, without a shared plan, the design cannot have come from them. It can only have been imposed afterward, by the redactor. But a redactor who could take independent documents and shape them into a structure this intricate, while preserving them nearly intact, is no longer an editor in any ordinary sense. As the scholar R. N. Whybray observed, an editor capable of all that could just as well be the author. The philosopher Franz Rosenzweig once remarked that the letter R, which the critics use for the redactor, might as well stand for Rabbenu, “our teacher.” The redactor, on this showing, becomes the very author whose existence the theory set out to deny.
Consider next the details. We found that passages assigned to P in Genesis are tied by their words to the J and JE material around them, again and again, often through rare words. If, as the Neo-Documentarians hold, P was a wholly independent document, these ties should not exist at all. Two authors writing separately, without knowledge of each other, do not choose the same rare words at the same points in their stories. If instead P was written around J, as other scholars hold, then P’s author knew J intimately, and the theory has already given up the independence of its sources. Even that does not explain everything, because the links do not run only one way. A later author can echo an earlier one, but an earlier author cannot echo a text not yet written. Yet in the Flood story, for instance, the word basar, “flesh,” appears first in a verse assigned to J (Gen 6:3), and every other occurrence of it in the Flood narrative falls in passages assigned to P. The links run in both directions, across the supposed boundaries. Either the passages were written together, or a redactor reached into every source and chose its words to match the others. But if the redactor did that, the vocabulary of the sources is his, and vocabulary is one of the main tools the critics use to tell the sources apart.
This matters all the more because Genesis is the foundation of the theory. The criterion of the divine names, the first and classic test for separating the sources, applies only up to the revelation of the divine name in Exodus 6. The profiles of the sources, their characteristic words and concerns, were built largely on Genesis. If P cannot be cleanly separated from its surroundings in Genesis, the case for separating it elsewhere loses much of its ground.
Consider, too, the connections between distant passages. The chain of rare words that links the story of Joseph to the law of the heifer in Deuteronomy and the law of the red heifer in Numbers runs across sources the critics assign to different authors and different centuries. Connections of this kind only make sense if someone had all these texts before him together, as parts of one work, when the connections were made.
Consider the Tabernacle. The critics assign its plan to the Priestly writer, and many of them hold that it was never built, an ideal sanctuary that existed only on the page. On their theory the structure of the Torah as a whole was planned by no one, and the human body was designed by no author at all. Yet the three share a single structure, five members with a pair attached to the fourth, and the account of the Tabernacle’s making follows the formation of the human head in the womb, an order no ancient author could have known. As the chapter on the Tabernacle showed, the critics’ theory gives no reason why any of this should be so.
Consider, finally, the later books. Their authors alluded to the Torah through precisely the same kind of word connections, and they did so as though they expected their readers to know the Torah as a whole and to catch the allusions. Some of those allusions point to material the critics assign to P or to D. Gideon’s battle in Judges, for example, draws on the law of the trumpets in Numbers, the only passage in the Torah that commands sounding trumpets in war, with the promise that “you will be saved from your enemies” (Num 10:9), a passage assigned to P. And the opening chapter of Isaiah, traditionally read together with the opening of Deuteronomy and linked to it by the tradition itself, shares with it the lament “How” (eikhah) and the language of justice and judges. If these allusions are real, the texts they allude to already existed, and were known and revered, when the allusions were written. On the critics’ own dating, that is earlier than the material should have existed.
The critics have replies available. They can suggest that the allusions in the later books were added by still later editors, or that the links in the Torah are coincidences, or that a series of redactors added them over time. But each such reply requires another hypothesis, and another hidden hand working carefully through texts that, on the theory itself, were never meant to be read together. With each step the theory comes to look less like an explanation of the evidence and more like a means of preserving itself against it.
None of this, by itself, proves when the Torah was written or by whom. Unity of design shows that the Torah is one work. It does not, on its own, establish its date or its author. What it does show is that the theory of separate sources, cut and pasted by an editor, cannot account for what the text actually contains. The simplest explanation of a single, carefully designed text is that it is what the tradition has always said it is: one Torah, from one source.