(Your response, of course: it is right. And stop calling me “Shirley.”).

Re: 2, I missed your article, thanks for the pointer. I think there’s a few different issues here, and we should distinguish between them. One is how to measure the “effect of the use upon the potential market,” and whether that includes or does not include the use at issue, which raises the circularity problem. A second is how tangible or certain a “potential,” but currently unexploited, market has to be. A third, and the one I really meant to raise in my comment, is WHEN is that analysis locked in, if ever? I.e., at time 1, there is no market for the use that passes whatever test you use. Defendant wins, at least on factor 4. Now at Time 2 there’s a market — someone’s figured out how to make money off of thumbnails. Are you locked in by the holding at Time 1? Or do defendants at Time 1+n now lose on factor 4? I.e., when if ever in a new technological field do expectations become settled as to what the scope of the owner’s market is?

In terms of measuring the value of the use: that is a real imponderable, because it’s all circular.

Let me throw back a few (perhaps unfair) questions:

1) Should courts be promoting new markets for porn? Admittedly, Perfect 10 is probably a lot better than its competitors, but isn’t it bizarre that courts should stretch to try to encourage this business?

This may seem overly “substantive,” and not content-neutral enough for 1A scrutiny, but there was a great note in the Harv. L. Rev. recently against copyright in ads…precisely because they often constitute a “bad,” not a “good.” Bartow’s also questioned whether copyright law protects porn too much…in a very recent article (whihc I sadly can’t remember the title of).

2) Where do licensing markets stop? And should we at least agree with Lemley that judicial recognition of a licensing market should at least be predicated on the existence of such a market?