I have since come across this passage in The Times review of David Rynin's edition of Alexander Bryan Johnson's Treatise on Language (Lit. Supp., 19th July,1947). It encourages me to think that I have not overstated the case against abstract words.
"In this Treatise it is assumed that since there is nothing objective corresponding to anything but a proper name, and no symbol can function as a proper name unless we are acquainted with what it means, all other words purporting to designate universals, abstractions or insensibilia are meaningless, or at best can only be attempts to refer to something associated, presumably by way of feeling, with designata for which concrete evidence is offered".
I did not know it was as bad as that.