One often reads today (see below) that certain properties of natural language (NL) or certain methods of describing NLs are (VIRTUALLY) CONCEPTUALLY NECESSARY. I take the negative view that such usages are an illegitimate, purely rhetorical way of seeking to justify assumptions which cannot be supported on genuine factual or theoretical grounds. Put differently, such claims are a way of suggesting that specific and actually highly controversial assumptions more or less follow from some very general and uncontroversial features of NL. The problem is that the suggestion is not, and, I claim, could not be, backed up with genuine evidence or argument showing that it is true.