The main theses of this paper are 1) that there are such things as "negotiating strategies," 2) that negotiating strategies are acquired from experience from a variety of often dissimilar events, and 3) that when those strategies are brought together at the bargaining table, they usually bring about agreements that are sub-optimal from an overall social welfare standpoint. This perspective is set up to contrast with a variety of models and views that tend to treat negotiations as self-contained events, as though they were individual "games" that could be analyzed solely with respect to their particular parameters - in isolation from other negotiations and experiences that may have shaped the behavior and expectations of the bargainers. A by-product of this approach is the proposition that there is no single answer to the question "how should one negotiate?" and that instead we should expect to find a variety of dissimilar strategies in use, and that the effectiveness of a strategy will depend upon which other strategies it encounters in practice. The paper is constructed around the concept of "learning," and builds its theses from explorations of the various definitions and interpretations of that word. The paper begins with a short explanatory background, proceeds to explorations of three different applications of the term "learning" in negotiation theory, and then goes on to develop a model of adaptive learning that is relevant to negotiating strategies. Following this is a discussion of the social "optimality" of the negotiating strategies that can be expected to evolve in a society. Then there are a few numerical examples that are intended to highlight the properties of the theory, and, finally, a few comments on insights that may be useful for negotiators themselves.