daf bit: Bava Batra 172

[cellio]

The mishna teaches: if there are two men in the same town and both are named Yosef ben Shimon, neither may produce a bond of indebtedness against the other. Further, nobody else may produce a bond of indebtedness against either of them. And if a man finds among his possessions a quittance showing that the bond of Yosef ben Shimon was discharged, it applies to both of them. So how should they proceed, since we want Yosef to be able to borrow money? When writing the documents (both bond and quittance) they should write the names to the third generation (e.g. Yosef ben Shimon ben Reuven). If their names are the same to the third generation, then they should add a description (e.g. Yosef ben Shimon ben Reuven, the tall one). And if those are like too but one is a kohein or levi and the other not, they should indicate that. (I can't tell if they keep the description in this last case.) (172a)

Neither the mishna nor the g'mara here addresses the case where Yosef ben Shimon was unique and then another one moved into town.

I assume we're talking about small towns here, where it's not implausible for names to be unique and for people to know that. I'm a little surprised that a description (which could be subjective or mutable) has higher precedence than kohein/levi status (which is neither).

When I shared this at minyan this morning, somebody told me that one of her family members has a last name that means "limp" (as in "has a", not as in "floppy"), which seemed peculiar to her. She said she was going to go teach him this mishna.

Comments

[richardf8]

Has it occurred to you how like trying to come up with a compound key that can serve as a unique identifier this is?

And what a rich field talmudic information theory could be?

[siderea]

I was thinking "This is the oldest documentation I know of how awful it is to try using personal names as primary keys."

Now I am also thinking maybe the Jewish tradition of never naming a child after a still-living relative is not superstitious, but out of a reasonable abhorance of namespace collisions.

[metahacker]

+1
"Such naming, it angers the hashing function gods"
Would read that story

[cellio]

Ha, yes! And a couple thousand years later we still haven't learned to fix our identifiers!

Now I am also thinking maybe the Jewish tradition of never naming a child after a still-living relative is not superstitious, but out of a reasonable abhorance of namespace collisions.

Oh, I hadn't thought of that. Interesting idea. (I understand that this is specifically an Ashkenazi tradition, by the way. I'm told that S'fardim name people after living relatives. Now I wonder if they were more prone to descriptive nicknames.)

[cellio]

It has! What I was actually thinking was along the lines of "UUIDs, man! Not names!"

And even today we still have this problem, and I don't just mean in badly-designed database schemas. When we worked with a lawyer to write our wills, we weren't asked for something like Social Security numbers for the people we named. For relatives maybe that's not needed ("my brother Bob" is pretty clear), but I named a few non-relatives in mine. I guess we rely on the executor to figure it out? ("Oh, the Mary she's assigning custody of her cats to is probably the Mary who takes care of them when she's away...") I don't think we were even asked for addresses, which would at least provide some clarity for one point in time ("in 1998 Joe lived at $address in $town").

[thnidu]

Yeesh.