Between two opposite conceptions about meaning, the traditional one of “literal meaning” and what I have called “Contextualism “ (Dascal, 1981), there is an alternative which I have dubbed “Moderate Literalism”, and defended in another place (Dascal, 1983).
 Contextualism as I understand it is the view that linguistic expressions lack “literal” meaning established by rules of language, and asserting that they get their meanings only as a function of the context within which they are used.
 My view of moderate literalism does not exclude literalism but it does modify this view taking off its excess weight. According to my view, the excessive demands made by the traditional conception on literal meaning are left aside, such as: it be sufficient to determine truth conditions and illocutionary force; that is always be a part of the message transmitted by the speaker; that it cannot be cancelled or neutralized in any context of use, etc.
 In this paper I do not intend to expound my proposal —this I have done elsewhere (Dascal, 1983)— but to discuss some recent criticisms against the notion of literal meaning. Gibbs (1984) tried to show that from the viewpoint of the psychology of language understanding, this notion does not play the role which had been attributed to it.
 A detailed discussion of the criticism will have as an outcome —according to me— that it is unnecessary to go from the traditional conception to radical contextualism, and that the notion of literal meaning, modified in due form, retains its validity and a crucial relevance for the psychological explanation of the process of understanding utterances.
 I will consider certain of Gibbs’ theoretical objections to the notion of literal meaning (section 2); afterwards I will take his empirical arguments (section 3), and conclude with a discussion of the implications that the author attributes to his criticism.
 [J.A. Robles]