Levels of Meaning and Moral Discourse

Abstract

In a brief paper, in Hebrew, Professor Bar-Hillel has addressed himself to the task of explaining the often noticed peculiar characteristics of ‘moral sentences’1. He points out the peculiarities in question by contrasting, with respect to certain properties, a typical ‘moral sentence’ (e.g., ‘Stealing is wrong’) with a typical non–moral one (e.g. ‘All ravens are black’)2. Although both sentences are grammatically declarative, the latter is verifiable, its predicate is ‘directly observable’ and there is no doubt that it has a truth–value, whereas, whether the former has any of these three properties is open to serious question. These differences are sometimes presented as characterizing a more general opposition or ‘gap’ between two types of discourse, described by pairs of labels such as ‘evaluative’ vs. ‘factual’, ‘prescriptive’ vs. ‘descriptive’, ‘practical’ vs. ‘theoretical’, etc. ‘Moral sentences’ constitute only a subcategory - perhaps the most important one - of the first member of each pair, but the analysis of their peculiarities may be considered as paradigmatic for the whole type of discourse to which they belong. In order to explain these peculiarities, some philosophers simply deny that moral sentences have ‘cognitive meaning’ (since they cannot be true or false), and proceed to assign to them another type of ‘meaning’, the most usual one being the so-called ‘emotive meaning’. Others defend the cognitive meaningfulness of moral sentences, but postulate that they refer to a peculiar (‘non– natural’) kind of properties or to a special realm of ‘reality’ (the realm of ‘ values’).