John lies to Mary if he says X, believes X to be false, and intends that Mary believe X.
As it stands, this doesn't work, since I may say X without representing X as true, or in a context in which it need not be taken as true. We also get complications with regard to in-persona statements -- perhaps John says X, believing it to be false, because he has a legal or moral responsibility to convey a message from somebody else, but no obligation to believe the message; perhaps he even makes clear that he himself doesn't believe it. If he did however intend that Mary believe X, is that enough to make this a case of lying? That is, lying has to be a way of speaking falsely, and there seems good reason to think that says-but-believes-false accounts are not adequate accounts of what it is to speak falsely about something. Dworkin himself goes on to characterize lying as "saying what you believe to be false in order to make other people believe something false"; but the definition he gives has no 'in order to' -- the intent that Mary believe could be incidental to the saying, for instance. (Another potential problem: what if John says X to Mary, knowing it to be false, but knowing that Mary will know he is lying, and he deliberately does this with the intention that Mary will believe some other likely alternative Y, which John also knows to be false! That is, what if the deception is oblique, rather than straightforward?)
In any case, he goes on to give a list of possible scenarios, asking his commenters at The Stone to say whether they regard them as permissible or impermissible. Whenever we deal with scenarios, we have to be on the lookout for sleight-of-hand or illusions created by wording. In several cases here, for instance, we are dealing with situations in which just a slight change of wording could change entirely whether there was even anything false. (10) is a really good example of this:
10. We heap exaggerated praise on our children all the time about their earliest attempts to sing or dance or paint or write poems. For some children this encouragement leads to future practice, which in turn promotes the development–in some — of genuine achievement.
We obviously heap praise on children for attempting difficult things, and do so all the time, and there's nothing obviously false about that, so it could sound like it's perfectly fine. And sometimes we might call this 'exaggerated praise', where that means, 'we work to make it very, very clear', like giving an exaggerated emphasis to something so that nobody misses it. But 'exaggerated praise' in this context has to mean that we are not praising the children for doing well given the difficulty, but in ways that we have reason to think they don't deserve. It's not clear that we actually do this all the time; and it's not clear how lying to children about how good they are at something actually relates to future practice in real life. (It is notorious, for instance, that children who are over-praised for things that they didn't work hard for tend to expect that things will come easily and stop working hard.)
And there are the usual issues of interpretation:
9. I am negotiating for a car with a salesperson. He asks me what the maximum I am prepared to pay is. I say $15,000. It is actually $20,000.
In the case of (9), I don't think people usually have so precise a notion of what the maximum they are prepared to pay are, and we are explicitly dealing with a negotiation. The answer could just as easily be interpreted as "let's assume for the sake of this negotiation, or at this stage of the negotiation, that it's $15000" rather than "in absolute terms, $15000".
(8) is particularly interesting:
8. In order to test whether arthroscopic surgery improved the conditions of patients’ knees a study was done in which half the patients were told the procedure was being done but it was not. Little cuts were made in the knees, the doctors talked as if it were being done, sounds were produced as if the operation were being done. The patients were under light anesthesia. It turned out that the same percentage of patients reported pain relief and increased mobility in the real and sham operations. The patients were informed in advance that they either would receive a real or a sham operation.
In (8) we are dealing with the same kind of situation as if someone were to say, "What's going to happen to you next will be either a play or a real-life situation, and we won't tell you which, because we want to compare the two." There's no clear intent for anyone to believe something false -- we've already stated that it might be false. And for what we're doing it doesn't really seem to matter whether patients believe it's a sham or not. This is another case where the distinctions noted above can matter.
Another issue that comes up in several cases is that the lying is actually just an element of a more general activity (comforting someone, for instance) which is obviously good, so that we have to be wary of confusing "The general kind of action is good" with "Doing it this particular way is good".