ABSTRACT
All of the features that, on Dreyfus's view, plague the Artificial Intelligence field in general are on display in Lenat's comment. The first point to observe about Lenat's assessment is the language in which Lenat characterizes the difficulties in NL: the grasping of a sentence's sense from among the multifarious possibilities, given the ways in which the given speech structure might occur in practice, might require almost any pieceof prior presumed knowledge. Lenat conceives, then, of the problem of understanding something like the contextualized situation in terms of mentally accessing pieces of prior knowledge. The marked difference between the way a symbol-system and a person respond to knowing more about a situation or a person suggests difference in the ways in which each entity arrives at a determination of the relevant aspects of a given state of affairs. Additionally, however, Lenat mentions the temptation and proclivity to define down success in the NL field, a temptation that he laudably resists.
