Page 441 - Encyclopedia of Philosophy of Language

time, publicly and in writing, opposes thewell-known Chomskyan criterion of 'well-formedness' as the ulti- mate standard by which to judge a linguistic production.
In the Chomskyan linguistic tradition, well- formedness plays the role of the decision-maker in questions of linguistic 'belonging.' A language consists of a set of well-formed sentences: it is these that 'belong' in the language; no others do. This is the definition that—assumed implicitly or explicitly invoked—has been the bulwark of the Chomskyan system since the late 1950s; it is also the definition that, from the earliest times, has most often come under attack from the quarters of so-called 'Ordinary Working Linguists' (often called 'OWLS'), and the one that makes least sense if, for a moment, consideration is given to what it is that people really say, and how they judge their own language's 'well-formedness' or 'correctness.'
As Lakoff points out, this latter notion is a highly relativistic one; it has to do (and a lot to do) with what speakers know about themselves, about their conversational partners (often called 'interlocutors'), about the topic of their conversation, and about its 'progress' (or what is felt as such, versus 'not getting anywhere'; the notion of 'progress' in conversation is discussed extensively in Stalpers 1993). In the follow- ing, this problem is discussed on the basis of a concrete example.
In grammar, correctness, as prescribed by the gram- marians, often collides with what the language user perceives as correct. Classical examples include the constructTM adsensum, by which a noun in the singular denoting a collective body takes a plural verb form, since the plurality of the 'sense' is perceived as more important than the command of the grammar to use the singular (e.g., 'The Board of Directors have decided not to pay dividends this year,' and similar constructions). The following example is another case in point.
There is a rule of English grammar that says the relative pronoun who should be used when dealing with a noun which is human (and animate, of course), whereas which should be used for a nonanimate (and usually also for a nonhuman) referent. Here are some examples:
The man who kissed my daughter ran away (who for a human subject).
The car which hit John's bicycle disappeared around the corner (which for a nonanimate (and nonhuman) subject).
and
The bird which shat on my nose flew away (which for a nonhuman, although animate, subject).
Such are the rules. However, questions then arise as to how they are maintained, and whether they are
always obeyed, or whether there are cases where rule observation is less 'correct' than breaking the rule. Consider some additional examples.
My cat, who believes that I'm a fool, enjoys tormenting me.
This sentence, due to George Lakoff (1971b), is not all bad, or always bad. It all depends on the cat, on the speaker, and on their relationship. Given a special, intimate connection between human and pet, it may even be the case that which, for a cat of a certain quality and lineage, is totally inappropriate, and even unthinkable; thus Lakoff: 'if I happen to have such a cunning feline' (1971b: 330).
The same is the case in the following extract, descri- bing a program (called 'CREANIMATE') that will allow children to create animals of their choice, using the computer.
In a typical interaction, a student may indicate that he wants to create a bird that swims. The system may respond by discussing some existing aquatic birds such as ducks and penguins. It could display video sequences demonstrating how these birds use the ability to swim to help them survive in the wild. The tutor would try to get the student to refine his design by asking whether the bird will use his wings to swim, the way a penguin does, or its feet, the way a duck does.
(Schank and Edelson (1990:9); emphasis added)
Strictly speaking, the above is not only ungram- matical: reference ('anaphora,' as it is called tech- nically) is made to a nonhuman being (a penguin) by the human pronoun who, but, moreover, incon- sistently: ducks are also nonhuman, but still they are referred to as it. However, this then raises the question of why a duck is 'it' and a penguin 'he.'
This is not a matter of mere humanlikeness in gen- eral (such as: penguins are 'dressed up' in black ties, like noble corporate gentlemen at a social occasion). The real clue to the different conceptualizations is in the total (not just linguistic) context and its attendant conceptualization. According to this, 'hand- swimming' is considered to be typically human, as opposed to 'doggie-style swimming' (with all four feet). Penguins, who swim with their 'hands,' are therefore practicing 'human-swim'; ducks, that swim with their feet, are like doggies, therefore they 'animal- swim.'
Notice furthermore that this is not a reference to the 'real world,' since humans do not swim only with their hands or arms, as everybody knows who prac- tices the breast stroke. It is the legs that really provide the motive power, but they are not seen as such: visu- ally and conceptually, the arms are what is somehow characteristic of human swimming. The language reflects this conceptual dichotomy by its different ref- erential pronominalization; and the moral is that attention to this should be paid in categorizing or indexing 'cases,' as when an attempt is made to treat
Pragmatics
419