Word Plays the Torah Explains
Some word plays are not hidden at all. The Torah stops and points to them, usually when it explains a name. When the builders of the tower are scattered, the text says: “Therefore its name was called Babel, because there the LORD confounded the language of all the earth” (Gen 11:9). The Hebrew word for “confound” is balal, root B-L-L, and Babel in Hebrew is Bavel, B-B-L. The two words are not identical, but they are close enough that no Hebrew reader could miss the connection.
When Noah is born, his father says: “This one will comfort us in our work and in the toil of our hands” (Gen 5:29). The name Noah has two letters, N-CH. The word “comfort” comes from the root N-CH-M, and its first two letters are Noah’s name. When Tamar’s son forces his way out ahead of his twin, the midwife says, “Why have you made a breach for yourself?” (Gen 38:29). “Breach” is paretz, root P-R-TS, and the child is named Peretz, with exactly the same letters.
In cases like these nobody doubts that the word play is intended, because the text says so. These explicit cases teach us what the Torah’s own word plays look like. The match need not be perfect, a two-letter name can echo a three-letter root, and what counts is the letters rather than the vowels.
Word Plays the Torah Does Not Explain
The more interesting cases are the silent ones, where the text says nothing and the reader must notice the connection for himself. Here is one of the clearest.
In Genesis 14 Abraham, still called Avram, rescues his nephew Lot by defeating four invading kings. In doing so he also recovers everything those kings had carried off from Sodom. The king of Sodom then makes him an offer: “Give me the persons, and take the goods for yourself” (14:21). “Goods” is rekhush, root R-K-SH. Avram refuses absolutely. He will not take “a thread or a sandal strap or anything that is yours” (14:23). A sandal strap is sarokh, root Sh-R-K.
R-K-SH and Sh-R-K are the same three letters in a different order. (SH and Sh are one and the same letter in the Torah scroll; the dot that later distinguished their pronunciation was added by readers long after the text was written. This book keeps the distinction only so that the reader knows how each word is pronounced.) The richest thing the king can offer and the cheapest thing Avram can name are made of the same letters, two verses apart, and the refusal is built out of the very wealth being refused. The text never remarks on it.
Real, or in the Eye of the Beholder?
A careful reader should be suspicious at this point, and the suspicion is fair. Hebrew builds its whole vocabulary from only twenty-two letters, mostly in groups of three. In a language built like that, some words are bound to share letters by pure chance. Anyone who looks hard enough will find “connections” everywhere, and the eye that is looking for them will find them most of all. How can we tell a real word play from a coincidence?
Our instincts are poor judges of coincidence. Suppose a word appears eleven times in Genesis, which has fifty chapters. How likely is it that two of the eleven fall in the same chapter purely by chance? Most people guess that it is unlikely. In fact the odds are about seven in ten. This is the same arithmetic behind the well-known fact that in a room of just twenty-three people, two probably share a birthday. If our sense of what is “too close to be chance” fails on a question this simple, it cannot be trusted with the subtler question of word plays.
The answer is to count rather than to feel, and this is what I did in a study of Genesis 14. First I fixed the rules for what counts as a match, based on the cases the Torah itself explains. Then I applied those rules to every word in the chapter. Finally I applied the same rules to a list of words drawn at random from the Hebrew Bible, matched to the chapter in length, in word lengths, and in its mix of rare and common words. The random list has every property of the chapter except one: no author chose its words. Running the comparison many times shows what chance alone typically produces.
The difference is striking. The real chapter produced 33 groups of words sharing three letters, while the random lists produced around 20. More telling was the size of the groups. In the random lists almost every group had only two members, and the largest had three. In Genesis 14 most groups had three or more members, and the largest had seven. Language alone does not produce this, and something more is at work.
The names in the chapter are the strongest part of the evidence. Genesis 14 is crowded with names: nine kings, their cities, the peoples they conquer, and Avram’s three allies. A story needs certain ordinary words. A battle in a valley must mention a valley, and a war of kings must mention kings. No story, however, requires a king to have a particular name, or any name at all. Yet nearly every name in Genesis 14 shares its letters with another name, or with a word in the chapter, or with a word in the chapters on either side. One coincidence proves nothing, but a coincidence repeated for nearly every name in a chapter is very hard to call chance.
None of this proves that any single word play was intended. A method of this kind sorts the evidence rather than settling every case. Some connections survive every test, others are probably chance, and many fall in between. What the method does show is that the overall pattern is real, and that the author of these verses chose words with their letters in mind.
Readers who still find the idea strange should wait until they have seen more. The examples in the coming chapters were chosen to be as clear as possible, and most can be followed without any Hebrew beyond the few letters each one needs. After that we will return to Genesis 14, and to the question of what its connections mean for the theory of sources.