Why aren't you reading "Harry Potter and the Methods of Rationality" yet?

[pozorvlak]

Eliezer Yudkowsky just kicked it up a notch: "Blood for the blood god!" screamed Neville. "Skulls for the skull throne! Ia! Shub-Niggurath! The enemy's gate is sideways!"

...

"I am Neville, the last scion of Longbottom!" screamed Neville to the sky above, holding his wand pointed straight up as though to challenge the blazing blue heaven itself,

Comments

[alexey_rom]

Well, the question is ill-posed: I am reading it and advertising it to everyone willing to listen :)

[pozorvlak]

Excellent :-)

Obvious howler?

[anonymous]

You really should say what the "obvious howler" is. Two wizards shouldn't give birth to a Squib?

Re: Obvious howler?

[pozorvlak]

I was deliberately not saying so anyone who hadn't spotted it could go back and look if they felt so inclined. But since you ask: (highlight to read spoilers)

If wizardry is controlled by one gene, and the wizarding allele W is dominant over the muggle allele M, then all muggles must be MM and there can be no muggleborns. If it's recessive, however, then all wizards and witches are WW, so there can be no Squibs. Since both Squibs and muggleborns exist, wizardry can't be controlled by one gene

Re: Obvious howler?

[squid314]

I thought I saw Eli say somewhere that in his theory the parents of Muggleborns were all Squibs who had gotten assimilated into the Muggle population a long time ago and didn't remember their heritage.

Re: Obvious howler?

[pozorvlak]

Still doesn't work, AFAICT. Every MW child must have at least one parent with the W allele, who (if W is dominant) would be a full-blown wizard or witch. Squibs would be MM like most people, and thus could only have muggle children. Dually, if W is recessive, then every Squib would have to have at least one M parent, who would be a muggle.

Possible alternative theory: W is dominant, but all "muggleborns" are actually the products of extramarital liaisons between muggle women and wizards. That would account for a lot more children than random mutation on a particular locus, I'll bet.