In a recent response to a question, Michael Lacewing writes: "Blackburn’s quasi-realism argues that ethical language is rather more complex than either emotivist theory claims [Ayer's and Stevenson's]. First, ethical language does express propositions, such as ‘what she did was courageous’ or ‘his remark was unkind’ as well as ‘murder is wrong’. The predicates ‘was courageous’, ‘was unkind’, ‘is wrong’, attribute a property to something (what she did, his remark, murder). However, second, these predicates aren’t genuine descriptions of what she did, etc. but ‘projections’ of our evaluations. In using ethical language, we don’t speak of and think in terms our personal evaluations, but in terms of the properties of things in the world. We treat our evaluative commitments (to courage, to kindness etc.) as though they were judgments about how the world is. This is enormously useful, because it is much easier to coordinate our attitudes with other people if we think in terms of an intersubjective world of moral properties. Third, this isn’t simply a mistake or illusion. Quasi-realism argues that we can meaningfully talk of moral judgments being true or false." My questions: does the desire to separate our evaluations from descriptions attributing (evaluate) predicates to people and actions in the world really make sense, especially if it seems that distinguishing them, as most people don't, would not be "useful."? Does the notion of subjective projections onto I guess the flat screen of reality make sense - at least without assuming some metaphysical fact/value dichotomy? What is wrong with thinking that 'what she did was courageous' is both a fact about what she did and an expression of admiration for that fact? Thanks for your time.