> Certainly, an unbiased symmetrical appraisal should be the goal?I have not read Seligman's work, but I can see at least one way in which the asymmetry in response to positive vs. negative can be justified.
If someone is doing something positive, you want them to keep doing it--so you want it to be permanent. And you want them to internalize whatever it is that makes them do that thing, make it part of themselves--so you want it to be personal. And you want them to expand whatever that positive factor is to other domains--so you want it to be general.
If someone is doing something negative, you want them to stop doing it--so you want it to be temporary. And you want them to not make it part of themselves--so you want it to be impersonal. And you want them to keep it confined and not let it affect other domains--so you want it to be specific.
In other words, the "optimistic bias" is an asymmetric strategy to achieve an asymmetric goal--maintain positives, but change negatives.