Chai, and the gift of eighteen

Why the cheque is written for 36 and not 35, and what the number eighteen was doing in Jewish hands long before it was a donation.

The arithmetic is the shortest in Jewish life. Chet is eight, yod is ten, and the word the two letters spell is חי, alive. Eighteen. It comes out at eighteen under the ordinal reckoning as well, since chet is the eighth letter of the alphabet and yod the tenth, and a number that survives being counted two different ways tends to feel settled.

חי=18 and חי=18 Siduri (ordinal)

Out of that comes a habit almost every Jewish family recognises. The wedding cheque is written for 180 rather than 175. The building fund is given 3,600. A child receives 18 dollars from a grandmother who has never in her life explained why, and does not need to.

What the sources do not say

It is worth being plain about this before going any further. No verse commands it. The Talmud, which has a great deal to say about the number eighteen, says nothing at all about giving money in multiples of it. The practice is a custom, and it belongs to the class of customs that grew out of a feeling for a word rather than out of a ruling. That is not a criticism. Judaism is full of them, and they are load-bearing. But a custom described as a law is a small dishonesty, and the honest version is more interesting anyway.

What the sources do carry is a long and careful attention to eighteen in its own right.

The eighteen the Talmud counts

The weekday Amidah is called Shemoneh Esreh, eighteen, and the Gemara in Berachot asks the obvious question: why eighteen? Three answers are given, one after another. Rabbi Hillel the son of Rabbi Shmuel bar Nachmani says they correspond to the eighteen mentions of the Divine Name that David set in the psalm beginning Havu laShem bnei elim. Rav Yosef says they correspond to the eighteen mentions in the reading of the Shema. Rabbi Tanchum, in the name of Rabbi Yehoshua ben Levi, says they correspond to the eighteen vertebrae of the spine.

The psalm Rabbi Hillel names is the twenty-ninth, and it opens like this.

The claim is checkable, and it is worth checking, because this is the sort of number that gets repeated for centuries without anyone opening the book. Counting every occurrence of the four-letter Name in Psalm 29 in our own index, bare or carrying a single prefixed lamed, gives 18. The Gemara is right, and it is right on the nose.

That is a different kind of eighteen from the one on the cheque. It is not a number read out of a word; it is a number counted in a text, and then used to explain the shape of something else. The rabbis who made that count had no index. They knew the psalm.

The same page raises the difficulty any davener will already have noticed, that the prayer as it is actually said runs to nineteen blessings, and answers it: the blessing concerning the heretics was instituted afterwards, at Yavneh. The name kept the older count.

Choosing life

The two-letter word itself carries most of its weight in Devarim, in the sentence that has probably been quoted more than any other line in the book.

And the next verse says what choosing it consists of.

Neither verse uses חי. Both use the plural, חיים, and the plural is not eighteen. חיים=68 The number belongs to the two-letter word, and only to it. Anyone who tells you the Torah instructs us to choose eighteen has done the arithmetic on a word the verse does not contain.

A coincidence worth refusing

There is a well-travelled remark that the verse read as calling the Torah a tree of life stands at Mishlei 3:18. It does.

The remark should still be refused, and the reason is a good one to have in hand. The division of the Hebrew Bible into numbered chapters is not Masoretic. It came into Hebrew Bibles late, by way of the Latin manuscript tradition, for the practical purpose of letting two people find the same place in a book. The verse divisions themselves are ancient and are the Masoretes’ own, but their numbering is not. So a chapter-and-verse coincidence is arithmetic performed on a printer’s apparatus rather than on the text, and it proves precisely nothing. A gematria is a claim about letters. Everything else is a claim about page furniture.

What eighteen actually looks like in the corpus

Eighteen is not a rare value. 703 word tokens in the twenty-four books carry it under the standard method, across 37 different spellings. חי itself stands 129 times. That is worth knowing before anyone builds an argument on a coincidence of eighteens: in a corpus of this size, the coincidence is the default.

The custom survives all of this intact. It was never a claim about the text. It is a way of saying, with money, the thing the word says, and it has the good sense not to argue.