Review of Assigned Readings

The Language Instinct

You should have read chapters 1-3, which basically cover the following material:

1) Argue for language as an instinct as natural to us as web spinning is for spiders (chapters 1-2)

2) Argue against the Whorfian Hypothesis and the implications of Orwellian New Speak (Chapter 3)

1) File 1: Introduction

A. General Introductory Material

This section covers the material in Pinker's chapters 1-3, as well as File 1 in Language Files. I'll review here some important points from Pinker's Chapters 1-2 (that I went over in class) and at the end will highlight some additional points that you should make sure to study.

a) language is a cultural artefact
b) the language we speak determines the way we view the world; (the Hopi's have no concept of time)
c) children learn to talk from role models and caregivers, especially from the careful speech they receive from their mother or primary caregiver
d) the English language is in decline; people simply don't use language as well as they used to, in large measure due to the influence of popular culture
e) English is an illogical langauge and English spelling makes no sense whatsoever
 
 

a) Is English in decline? What does decline mean? People who complain about the decline of English are really complaining for the most part about language change. They say, "We don't use English as well as we used to," or "We don't use the word X correctly." What these people fail to grasp is that English is not in decline. Rather, English is changing, as English has always changed. Just try to read an Old English text and you'll see this for yourself. Here's a link to the Old English epic poem Beowulf, for example. And for fun, here's a bit of the text itself:

Ða wæs on burgum         Beowulf Scyldinga,
      leof leodcyning,         longe þrage
 55 folcum gefræge         (fæder ellor hwearf,
      aldor of earde),         oþþæt him eft onwoc
      heah Healfdene;         heold þenden lifde,
      gamol ond guðreouw,         glæde Scyldingas.
      ðæm feower bearn         forð gerimed

Say what? Sure doesn't look much like the language we call English today.

For a more pedestrian example, only a couple hundred years ago, the past of "catch" was regular, but speakers changed it to "caught" on analogy with buy/bought. Now, it's not hard to imagine a whole bunch of people running around somewhat exercised about all the folks who were misusing the past of "catch". (For every generation, of course, the world is always going to hell in a handbasket.) Today's decline becomes tomorrow's norm. More about this later in the course.

b) English is illogical. Well, what does it mean for a language to be logical? Examples like driveway (where you park) don't show that English is illogical. They just show that English has compound words which don't necessarily mean exactly what the sum of their parts suggests. If we mean that English doesn't work like Prolog (a programming language), well, that's true. But, NO NATURAL LANGUAGE BEHAVES LIKE A PROGRAMMING LANGUAGE, so whatever English is guilty of, all human languages are also guilty of.

c) English spelling? Sure, it's a pain. But it's not as crazy as you think. Much of the problem has to do with the fact that English spelling is OLD and, as a consequence, reflects the way many words sounded at an earlier time.

Question: would a revolutionary overhaul of the spelling system do much good?
Answer: in the short run, maybe, but remember than languages are always changing, so in a few hundred years, things would look pretty messy again.

d) As for misconceptions (a-c) above, here's a combined answer. Language isn't a cultural artefact. As Pinker points out, We don't learn language the way we learn to tell time or to negotiate for the purchase of a car (a truly remarkable and perplexing experience with its own covert rules that is rooted in attempting to make you feel good about being fleeced and may still leave you without floor mats). Consider that 3 year olds are grammatical wizards but incompetent at other cognitive skills. Just try to get a three year old to do math or evaluate the pros and cons of buying a Mac, a PC, or a Sun station and you'll see this for yourself. This should tell us something. It should tell us, Pinker argues, that we're actually predisposed  to be language learners. The interesting fact is that language arises spontaneously in the human child, without effort or formal instruction. Sure, we get input from our primary caregiver, but think about how American children of parents with foreign accents almost never grow up speaking English with that same accent. If they were simply learning English from Mom or Dad, why would this happen? (Again, more on language acquisition later in the course.) The bottom line is that language is qualitatively the same in all of us. It's what Pinker calls an "instinct" and compares to the web spinning capabilities of spiders. Language is intimately a part of what we are, of what it means to have a human brain and is a biological adaptation to communicate information. All of this should strongly indicate that language isn't a cultural construct which dictates our world view. For more detailed discussion of why the so-called Whorfian hypothesis is not accurate, see Pinker's chapter 4. For the rest of this section, I'll focus on the kind of evidence that Pinker points to in order to argue that language is an instinct (and not a cultural artefact), i.e. that language arises naturally and spontaneously in the human mind.
 
 

a) pidgins and creoles

Pidgins are a makeshift jargon allowing for communication between speakers of different languages. They exhibit no regular syntactic patterning; i.e. no consistent word order, lack of inflectional material like prefixes and suffixes, show no tense marking and have no consistent way of identifying who did what to whom. So, what this means is that among people communicating in a pidgin,  much of the communicative burden falls on context.

Importantly, a funny thing has been shown to happen to the children of pidgin speakers. They don't acquire pidgin. Instead, they spontaneously change the pidgin into a full fledged language w/ auxiliary verbs, consistent syntax, prepositions, case markers and so forth. This process is called creolization, and the language that arises spontaneously from a pidgin is called a creole.

What does this tell us? It tells us that language acquisition isn't simply a matter of repeating what you are exposed to. Instead, creolization involves constructing a grammar in the mind, despite the fact that that grammar was not present as input to the child. So here we have evidence that children are not simply immitators, they are creators of language.

Question: Can you go back into Pinker and find a couple of examples of creolization that he discusses?

Another argument for language as instinct comes from so-called poverty of the input reasoning. This means that children acquiring language are able to produce and understand sentences which they have never before heard. Pinker discusses how children are able to make questions out of constructions they've never heard before. Consider for example, the following sentences.

• a unicorn is in the garden
• is a unicorn in the garden?

• a unicorn that is eating a flower is in the garden
• is a unicorn that is eating a flower in the garden?

What's the big deal, you ask. The answer lies in the relevance of the second pair of sentences, and this is why. In the first pair, we can see that making a question involves reversing the order of the copula (is) and the subject (a unicorn). Presumably, the simplest rule for learning how to make questions is to just move the first "is" to the other side of the subject. This becomes a problem in the second pair, though. Why? Obviously, if we move the first "is" to the left of the subject, we'll get: "is a unicorn that eating a flower is in the garden." Quite a mouthful, but more importantly, a) it's a "bad" English construction that nobody says and b) it's a mistake that very young children NEVER make, even though the particular type of construction is so rare as to be non-existent in the speech that very young children are exposed to. This is crucial, because it again shows that children are building their own grammars and that the "rule" of question formation is sensitive not just to what the first "is" in the sentence is, but rather to the hierarchical structure of the sentence. They know which "is" to move, because they are grammar building machines, even though they may never have heard this type of construction previously.

c) regularization

If you've been around little kids, you may have noticed that they say things like this:

• I goed to the bathroom.

What's the big deal? They're kids and they haven't mastered English yet. True, but this type of error is actually very illuminating because, again, it's something that kids do even though they NEVER hear it from the adults around them. It's more evidence that kids are not just imitators. Instead, they are grammar builders. This process is called regularizationbecause kids are taking an irregular verb and regularizing it. How does this show that they are building a grammar? If we think about it, we can see that what's happening is that the child has learned a general rule for past tense formation in English and simply applied the rule. The problem is that the verb "go" is a suppletive (put this in your memory buffer  for morphology) verb whose past tense form happens to be "went". Kids figure this out quickly enough, but the big picture is that such regularization calls our attention to the way that grammar emerges from the child's mind naturally. In this case, they've just overapplied the past rule.

Okay, if language is an instinct that arises spontaneously in us as a part of the structure of our brains, we might expect to find it in the brain and perhaps even find grammar genes (since genes for everything seem to be in vogue these days). Pinker points out that there is actually some evidence for this. Here some of it is in a nutshell.

Logically, if language is located in the brain, if you disrupt the relevant part of the brain, you might expect to disrupt language but leave other aspects of intelligence in place. Similarly, we should expect to find people who are mentally retarded but have perfect language, people who have been referred to as linguistic idiot savants.

Brain lesions can often lead to something called aphasia (again, more on aphasia later in the semester). One form of aphasia is called Broca's aphasia and it involves a loss of function words (but not content words). Pinker cites the striking example of a man incapable of using the function words OR and BE, but who had no trouble with the identical sounding (homophonous) content words OAR and BEE. Additionally, non-language related tasks involving general intelligence were not a problem for this individual. What's the point? The point is that the man's injury affected a part of his brain that had the effect of screwing up a particular part of his grammar but not the rest of his intelligence.

There's a disease called SLI (Specific language impairment) that runs in families and doesn't affect overall intelligence. But speech is enormously difficult for people who have SLI. To speak, for them, requires exhaustive mental work. The rest of their linguistic abilities remain intact, and they have no trouble with other aspects of general intelligence. What's interesting here is that this problem runs in families, strongly suggesting that the problem may result from a defective gene that is passed on through generations.

Finally, we have to turn to the relevance people who are linguistic idiot savants. Pinker discusses the case of Denyse, born with spina bifidia (split spine) which results often in hydrocephalus, an increase in pressure in the cerebro-spinal fluid filling the ventricles of the brain. Anyway, gory details aside, Denyse is a prolific conversationalist but is also significantly retarded. Here then, we have the other side of the coin. If brain lesions can affect language without affecting intelligence, we now have a case of retarded intelligence but a fully developed linguistic grammar.

BACK TO THE BIG PICTURE: Pinker develops the view that Language IS an instinct in the sense that language emerges naturally in humans just as nest building does in certain birds. Pursuing a computer metaphor, we're hard-wired to acquire language from the time we are born. And learning a language involves not simply repetition of what we hear but rather the internal creation of a grammar.

• arguments against the Whorfian hypothesis
• why Pinker says Orwellian Newspeak would never work
• what Pinker means by the term mentalese