The Baltimore Stockbroker and the Bible Code
Ellenberg introduces attempts to find hidden messages in the Torah through equidistant letter sequences. The broader statistical question is how to distinguish evidence of a code from patterns found by searching many possibilities; this excerpt stops before the worked example and critique.
THE IDEA
The Baltimore Stockbroker and the Bible Code
People use mathematics to get a handle on problems ranging from the everyday (“How long should I expect to wait for the next bus?”) to the cosmic (“What did the universe look like three trillionths of a second after the Big Bang?”).
But there’s a realm of questions out beyond cosmic, questions about The Meaning and Origin of It All, questions you might think mathematics could have no purchase on.
Never underestimate the territorial ambitions of mathematics! You want to know about God? There are mathematicians on the case.
The idea that earthly humans can learn about the divine world by rational observation is a very old one, as old, according to the twelfth-century Jewish scholar Maimonides, as monotheism itself. Maimonides’s central work, the Mishneh Torah, gives this account of Abraham’s revelation:
After Abraham was weaned, while still an infant, his mind began to reflect. By day and by night he was thinking and wondering: “How is it possible that this [celestial] sphere should continuously be guiding the world and have no one to guide it and cause it to turn round; for it cannot be that it turns round of itself?” . . . His mind was busily working and reflecting until he had attained the way of truth, apprehended the correct line of thought, and knew that there is one God, that He guides the celestial sphere and created everything, and that among all that exist, there is no god besides Him. . . . He then began to proclaim to the whole world with great power and to instruct the people that the entire universe had but one Creator and that Him it was right to worship. . . . When the people flocked to him and questioned him regarding his assertions, he would instruct each one according to his capacity till he had brought him to the way of truth, and thus thousands and tens of thousands joined him.
This vision of religious belief is extremely congenial to the mathematical mind. You believe in God not because you were touched by an angel, not because your heart opened up one day and let the sunshine in, and certainly not because of something your parents told you, but because God is a thing that must be, as surely as 8 times 6 must be the same as 6 times 8.
Nowadays, the Abrahamic argument—just look at everything, how could it all be so awesome if there weren’t a designer behind it?—has been judged wanting, at least in most scientific circles. But then again, now we have microscopes and telescopes and computers. We are not restricted to gaping at the moon from our cribs. We have data, lots of data, and we have the tools to mess with it.
The favorite data set of the rabbinical scholar is the Torah, which is, after all, a sequentially arranged string of characters drawn from a finite alphabet, which we attempt faithfully to transmit without error from synagogue to synagogue. Despite being written on parchment, it’s the original digital signal.
And when a group of researchers at the Hebrew University in Jerusalem started analyzing that signal, in the mid-1990s, they found something very strange; or, depending on your theological perspective, not strange at all. The researchers came from different disciplines: Eliyahu Rips was a senior professor of mathematics, a well-known group theorist; Yoav Rosenberg a graduate student in computer science; and Doron Witztum a former student with a master’s degree in physics. But all shared a taste for the strand of Torah study that searches for esoteric texts hidden beneath the stories, genealogies, and admonitions that make up the Torah’s surface. Their tool of choice was the “equidistant letter sequence,” henceforth ELS, a string of text obtained by plucking characters from the Torah at regular intervals. The excerpt ends before the worked letter-sequence example and the Baltimore stockbroker story.
When to use it
When analyzing big data sets, consider whether the patterns you find are real or artifacts of multiple comparisons.
In evaluating miracle claims or conspiracy theories, ask whether the evidence is merely a result of cherry-picking coincidences.
When learning a new skill, beware of seeing progress in random fluctuations; use objective metrics to track improvement.
FROM THEORY TO EVERYDAY LIFE
In practice
A Hypothetical Stock-Market Pattern
Suppose a stockbroker searches many trading patterns and finds that a stock has risen the day after dropping more than 2% on a Tuesday in its last 10 occurrences. Reporting only that successful pattern hides how many possibilities were searched. Testing it on new data would help distinguish a useful relationship from a chance fit.
Torah code debate
Rips, Rosenberg, and Witztum searched for names encoded in the Torah through equidistant letter sequences. One way to assess such a claim is to apply comparable search rules to another long text, such as Moby Dick. The search’s flexibility and the number of attempted patterns matter when evaluating an apparent match.
A MOMENT TO REFLECT
Put it to work.
Can you recall a time when you thought you saw a meaningful pattern in random events? How did you confirm it was real?
What statistical safeguards could the Torah code researchers have used to avoid false positives?
Why is the human brain so prone to seeing patterns, even when none exist? How does this bias affect decision-making in everyday life?
Often confused with
References
- Equidistant Letter Sequences in the Book of Genesis — Doron Witztum, Eliyahu Rips, Yoav Rosenberg (1994)
- Mishneh Torah — Maimonides (1180)
CONTINUE THE CONVERSATION
Source texts
- How Not to Be Wrong The Power of Mathematical Thinking (Ellenberg, Jordan)