What to Make of It
We are left with two codes. One is extremely dramatic. The other is certainly interesting, but not nearly as strong.
The first is the code for Sisera. It is the most compact occurrence of his name in the entire Torah, and it begins on the word for the deeps “covering” the Egyptians, in the very song to which the Book of Judges ties his defeat. A second Sisera code, entering from the other direction, lies there as well. It does not meet the first, and it passes through Miriam’s verse and the words “horse and his rider.” The Song of the Sea is less than nine hundred letters long, about one three-hundred-and-fiftieth of the Torah. By chance alone, the most compact occurrence would fall in it about once in 350 times, and two of the five occurrences would fall in it only about once in twelve thousand times. Those figures, however, measure only where the codes lie. There is no way to put a number on how closely the words and verses they pass through fit the story of Sisera.
The second is the code for Deborah. It is not the occurrence with the smallest interval; 27 of the 99 are smaller. So it lacks what makes the Sisera code striking. Yet it sits in a very interesting place, running into the verses of Miriam the prophetess. It also has to be weighed against the other names. Barak, Jabin and Jael yield nothing at all, and Deborah’s own closest codes fall in places with no connection to her. The method does not produce a meaningful result for every name one tries. That makes it harder to say that one can always find what one is looking for.
Taken at face value, these results look dramatic. Read this way, the connections that the Book of Judges makes with the parting of the Sea are more than a later author’s literary choices. The Torah itself would contain allusions to the later text. The allusions would go in both directions: from Judges back to the Torah, and from the Torah forward to Judges.
Even if these codes are genuine, however, they do not validate the codes presented by others. The two are very different in what they do and in what they show.
The codes presented here agree exactly with what a plain text implies. Without the Book of Judges they would mean nothing. A hidden “Sisera” in Exodus 15 would be one more random arrangement of letters among hundreds. Their force comes entirely from the fact that they support what anyone can see openly in Judges: one connection after another, in words and phrases, to the parting of the Sea and its song. The search did not set out to find something new. It tested whether the hidden letters would agree with what the plain text already shows, and in this case they did.
The codes generally presented to the public are of a different kind. They are words found in the text with no indication that they belong there. As we saw, any word that is not too long will eventually be found by this method of searching, and usually many times.
It is often said that several related words gather together in one place, and that is the most striking claim. On examination, though, it is usually far less than it appears. The familiar pictures of codes show the letters of the Torah laid out in rows, with the hidden words picked out across them, one above another. The letters are real. The text has not been changed or deliberately manipulated. But the picture is not the text as it is written. Each row is a slice of the text cut to a length chosen to line up the code, and only part of each row is shown. Hundreds of letters may lie on either side of what appears on the page. Moreover, many words are usually searched for, and only those that fit the picture are shown. The others were either not found at all or could not be fitted into the same frame. A reader sees the hits and never sees the misses.
The researchers who began this work understood the problem, and their formal statistical test was an attempt to answer it. Whether it succeeded is still disputed. What I have shown is difficult to measure statistically. What is shown in most popular presentations is far more difficult to measure.
Is the code for Sisera real? My own sense is that it is. It does what the text itself does: it points to the same place that the author of Judges pointed to, by entirely different means, and there are too many coincidences in it for me to regard it as a fluke. But I do not hold this with full confidence. It is a single example, a very strong one, but still a single example. Nor do I know whether the much weaker result for Deborah counts against it. Have I proven it? No. I am sure that many readers will not think so, and there is little more I can say to make the case more convincing. In large measure the answer may depend on whether such codes exist elsewhere, in other places where a later book clearly alludes to the Torah, and on how strong they seem there. I have no way of quantifying this, and no way of saying “this has to be true.” So I suspend full judgment, while my sense remains that this code is real. Others, I am sure, will disagree.
So what I have shown should not be taken as evidence that the codes in general are true. Perhaps they are; perhaps some are and others are not. Each must be judged on its own merits, not by whether what I have shown here is valid or invalid. I present this one code for one reason only: it points to the same connection the plain text has already made.
One question remains. Any search of this kind depends on every letter standing where it was first written. Whether the Torah could have come down to us in that state is the subject of the next chapter.