Following your urging, William, I've spent some time on Roy Harris's teachings. (One of the many notes I've scribbled as I read Harris runs to over 5,000 words.) I reached a glum and unlovely-to-hear conclusion: Harris does not have sufficient acuity of intellect to do the job he assigns himself.
He doesn't see key distinctions. He doesn't notice inconsistencies in his own usages, allowed in large part by an apparent fuzziness in his own mind. He announces platitudes as though they were profundities. He unfairly condemns lexicographers because he apparently doesn't grasp what they are trying to do. For example, he writes, " Words are not temporal invariants (as dictionaries like to present them). . . . . The static abstractions listed by lexicographers thus impose yet another linguistic misconception on the unfortunate student. The lexicographer's view of language is a case of trying to impose a normative straitjacket on an open-ended flux of relations between linguistic acts and the world." Harris does not seem to grasp that most intelligent lexicographers do not think of what they call "words" as "temporal invariants". First, about variants. Harris reacts as though lexicographers insist a given definition is THE, and the ONLY, invariant description of the notion that arises in the mind of someone "familiar" with a given "word". I claim they don't do that: My dictionary has 126 variants in the entry for 'run'. "Oh, but that's not "temporal" variation - variation over time. Lexicographers act as though they don't realize the notion stirred by a given "word" varies over time." Yes, they do. They know that in composing a dictionary-definition they are attempting only this: to write something such that reading it will cause certain notion to rise in the reader's mind -- notion that's commonly, roughly, like what's in minds of people of this era familiar with the word. The "of this era" is important. They know they are writing a dictionary today for today's readers. In a good modern dictionary, lexicographers will often supply definitions that they label 'archaic'. They obviously would not do that if they didn't believe a 'definition' of three hundred years ago could be misleading today. I myself will agree that I'd be in error if I go to a dictionary to find THE meaning. But I love the fact that the dictionary is there when I want to find out what today's doctor has in mind when he mentions otocunia. I certainly don't feel that by trying to describe the notion in the minds of such doctors the lexicographer is trying "to impose a normative straitjacket" on me. The people who think of a dictionary definition as "normative", as "THE meaning", are wrong. It is merely "descriptive" - of the notion likely to arise these days in the minds of people familiar with the "word". Harris writes: ""The term integrational alludes to the recognition that the linguistic sign alone cannot function as the basis of an independent, self-sufficient form of communication. . . A sign acquires a meaning only when occurring in a specific context. " His apparent conviction that this "integrational" observation is a profound and novel insight seems to me unwarranted. Who of intelligence would ever question that at least some of the notion occasioned by hearing an utterance can be affected by the situation in which it's uttered? Notice his locution, "A sound acquires a meaningb&" That's bad phrasing. It's too blurry in serious linguistics or philosophy ever to suggest that a sound "acquires" or "has" a "meaning". I'm not bent on annihilating the term "meaning". The notion(s) that arises in Jones's mind when he hears a phrase may be usefully termed the phrase's "meaning for him". On this forum I've regularly stressed that error creeps in when we assume that any utterance/scription has a "THE" meaning, which many of us fuzzily think of as a mind-independent entity. You'll recall my repeatedly saying that all notion is indeterminate, indefinite, multiplex and TRANSITORY. The 'transitory' is something Harris and I seem to agree on: What today comes to mind when we hear an utterance (text) will be different, however slightly, from what came to mind when we heard it yesterday. When Harris says, "Every episode of communication, however trivial, necessarily involves creative activity by the participants, including their own interpretation of the situation in which it occurs," I claim I know what he is after, but it too feels almost platitudinous. My guess is he's counting on the word 'creative' to give the assertion a glow of originality. When Harris says, "Every utterance is a new utterance, no matter how many times someone may have 'said it before'", he doesn't seem to realize the ambiguity behind that 'utterance' and that 'it'. Maybe this conveys the difficulty: an "utterance" can be thought of as either an act of speaking, or as the text of that speech. Technicians might cite the difference between a token and type. If I say "my coffee mug" five times in a row, of course each set of sound waves is new, another instance of uttering. But almost all of us would agree I've used "the same phrase" over and over again. I also resist a bit when Harris says, "A linguistic act, in any case, does not necessarily require the utterance or inscription of words." This announcement reminds me too much of those philosophers (and critics) who feel they have achieved a breakthrough in thought when all they've done is merely stretched a word's scope beyond its conventional reach. (As a youth I remember being irked by hearing the word 'obscene' applied to both the holocaust and a lady's extravagant Easter Parade hat.) If we grant that Harris (whose title is Emeritus Professor of General Linguistics at Oxford) is entitled to stretch 'linguistics' any way he wants, I feel he is then required to come up with another label for what it is his more restricted "linguistics" colleagues are studying and teaching. Innovative usages of familiar terms seldom enlighten us; more often, they simply kill a formerly useful term. Harris would call "linguistic" such things as traffic lights, paintings, fire alarms, photographs, nasty finger gestures, doorbells and other items. Harris seems to call 'linguistic' anything that's used to "communicate". If that's right, I'd prefer he save the old usage of 'linguistic' and, for what he has in mind, simply use the adjective 'communicative".
