Ferdinand de Saussure (185701813) is generally considered the person who formalized many common sense ideas about language and thereby turned philoogy into linguistics.
The editors of the online journal Biolinguistics are a bold lot, publishing articles that question the journal’s chief axiom that there is such a thing as a biological, language faculty. The current issue has a long paper (available here) by Dutch linguist Jan Koster that rejects the whole idea of any kind of biological organ or special faculty for language. Instead it presents a classical humanist portrait of language. Classical humanism can be summed up the pithy statement of Protagoras, “man is the measure of all things.” It contrasts with the religious notion that God is the measure of all things, and also with the more modern doctrine that physics, chemistry, and computation is the measure of all things. He sums up his case in the paper’s main title, “Ceaseless, Unpredictable Creativity,” and argues that “language is primarily a cultural phenomenon … [and like] the biosphere … human culture is ceaselessly creative in ways that are fundamentally unpredictable and presumably non-algorithmic or machinelike” [pp. 61, 69].
I’m sympathetic to the proposition, although when taken to Koster’s extreme it pretty much does away with this blog’s role. The origin of speech in his eyes is as biological as the origin of computer programming.
The phrase about ceaseless, unpredictable creativity comes from Stuart Kauffman whose paper “Beyond Reductionism: Reinventing the Sacred” (here) has a section headed, “The Biosphere and Human Culture are Ceaselessly Creative in Ways that Cannot be Foretold.” In fact, much of Koster’s paper reads like the application of Kauffman’s ideas to linguistics. Make that, applying ideas to classical linguistics. and frankly I enjoyed seeing somebody take out those old, common sense ideas and take them for a spin. How do they hold up?
- Classical idea #1: “Language…is a set of invented tools that enable us to give public expression to our inner feelings and thoughts.” 
- Classical idea #2: Language is a “system of signs [which have] an external aspect (significant) and a conceptual aspect (signifié) [and cannot] be reduced to individual psychology.” 
- Classical idea #3: “Language plays a dominant role in determining the common culture of those who speak a language.” 
- Classical idea #4: Language is arbitrary, both in the relation between sound and sense, and in the way language divides reality. 
- Classical idea #5: “Language…is the fruit of human agency.” 
- Classical idea #6: “A language minimally contains words… man-made, public cultural objects” [67, 66]
Anybody familiar with the linguistics of the past 50 years will immediately notice that the principle classical ideas had nothing to say about syntax. And it is not just an oversight. The investigation of syntax proposes alternative views to the classical ones.
- Counter idea #1: Language is a set of rules that organize and generate sentences.
- Counter idea #2: The rules of language can be formalized and explained by individual psychology.
- Counter idea #3: Culture determines the way rules are expressed in any particular language.
- Counter idea #4: The rules of language are not arbitrary.
- Counter idea #5: Language is the fruit of the way the brain is organized.
- Counter idea #6: A language minimally contains a predicate and a verb.
These two sets of ideas seem to talk past one another and never really grapple with the other’s position, so it might seem that a synthesis should be possible. Just view each one of the systems as incomplete and combine them. Indeed, at one point Koster refers to “the synthesis of Saussurean and Chomskyan ideas that I advocate” [78-79], although most of Chomsky goes out with his approach. I think, however, that neither system gets at the heart of what makes language so remarkable, so combining the two will still miss the fundamental mystery.
Give Koster credit for arguing effectively that language is ceaselessly creative and not entirely subject to mechanical rules. In other words he rejects the original Chomskyan ambition of writing a set of rules that can generate “all and only” the sentences of a particular language. Koster denies that a syntax and dictionary are sufficient to determine meaning. An ever-changing environment is also part of the story. Thus, it might seem that the most fixed part of a language is the proper noun such as Franz Schubert, but Koster provides a series of sentences that show it will not suffice to know that the name refers to an Austrian composer who perfected the form of the German art song. Even the meaning of proper nouns can be ceaselessly creative:
- Schubert is difficult.
- Schubert will take 30 pages.
- Schubert is for sale.
- Schubert will be reburied next yer.
- Schubert can be downloaded everywhere.
- Schubert will be burned on request.
I confess to being so old-fashioned that I paused a moment before understanding that last usage, but that’s Koster’s point. Environments change and usages change with it. But the usage change cannot be biologically based; it must be a man-made [sic] bit of creativity, an adaptation to our new, home-burnt-CD environment. Since I have been making similar arguments for years, I’m easily persuaded by Koster. He sums it up in direct rejection of “ultra-Darwinians,” “We do not live a life dictated by our genes.” 
At the same time, I find that I have moved well away
from the common sense of the classical line. I once, but no longer, I did think
of language as a tool for making public what is privately felt and thought.
Yes, it can do that, but there are counterexamples: — The president gave a speech today. (Reports what is public.) —My wife is of two minds on the
matter. (Reports what is private in another.) —I have to get a grip. (A thought that remains private.) And anybody
who has written a story or essay knows that the use of language commonly
results in ideas that first appear on paper that are then internalized by the
writer. So the process goes from outside to in, not in the classical direction.
So should I go with the counter idea that language is a set of rules? That
won’t work either, because of language’s ceaseless creativity. Schubert will be burned on request is an
English sentence, but the definitions of both Schubert and to burn are
outside the dictionary. And if we change on
request to something more modern like upon
tweeting* the whole sentence steps outside the dictionary.
I’m sticking with the definition that I have used before on this blog: Language is a tool for governing perception by piloting attention. I may want to pilot attention to my personal feelings and thoughts, and I may be content to use the existing rules and definitions, but language allows much more.
Koster comes closest to grasping my point when he refers to “awareness” as “part of the process that interprets coded word information … [and] eventually will contribute to subjectively experienced understanding,” but he misses. The subjectively experienced understanding doesn’t come “eventually,” and it doesn’t come from some general awareness. It requires attention from the outset and perception governs the interpretation. It might seem a strange miss for an author who rejects the idea of a biolinguistics, rejects the mind=brain notion, and who sees Chomsky’s fundamental error as conflating computational and natural languages. But the one non-classical idea Koster accepts is that the brain (not the mind) is all computer. In that computational model attention, perception, and sensation are, at best, a side effect.
Which brings us to the old mystery, the conundrum that so baffled classical linguists that they forbad the question: where did language come from? If language is entirely a man-made, word-dependent technology, how could it have possibly begun? Who could have said stick and had somebody else understand that the sound just made was supposed to refer to a stick? Returning to square one and staying there is not going to win many adherents on this blog.
* To tweet: to send a message via Twitter.