61 61

Chapter 45: The Absence of Absolute Reference

How do words and things connect? As we pointed out in chapter 3, it seems totally obvious to many people, even inevitable, that the connection could not be anything more than a "point and shoot" verbal camera. We "point and name" and that is all that it takes. This has been a common sense truth for centuries. Nonetheless, in philosophical and linguistic circles, it is a devilish challenge to state how we actually refer to objects, not to mention situations, feelings, and ideas. The rapid stream of language and inchoate demands of life force us, every moment, to combine old words in order to pick out new references.

Must a child learn “words” first? It is not so obvious. Suppose a child first recognizes commotion and activity. And he hears a parent saying “look there is singing, dancing, and eating”. The child might first learn “ing” as equal to activity and march into a busy room and shout “-ING”. No one reports such a result. If we want to build a good theory of acquisition, we need to not merely observe that this does not happen, but build a theory that explains why it does not happen. If a child has a rule that says “learn words in isolation before more complex structures” then the fact that no adult says “-ING” by itself would make it not a child’s first hypothesis. But now we have articulated a principle for how inflectons appear: learn words first, then learn affixes that alter their meaning. How do we know that “ing” is not a word? It never occurs by itself. But here is already a grammar-specific constraint “occur by itself = word.” Thus we have grammatical assumptions from the outset.

We will wander through the most obvious terms that orient reference: pronouns (it, that) pointer words (this, that) place words (here, there), identity (same) and non-existence (no hats). Each term has individual wrinkles that are part of the repertoire of reference. What is, step by step, the ladder of abstraction for reference in the acquisition process? It remains dim, the sequence hard to see. Our explorations are designed to extract the latent ambiguity in each of them. The ambiguities are slightly contrived, but they reflect ambiguities that arise every day, that silently confound children, that flummox foreigners, and that must be mastered to be full adult speakers..

The overarching idea is this: grammar is laden with abstraction. Yet children may be innately designed to grasp exactly the abstractness of reference (milk is good) before they see how specific reference works (this milk is good). But much remains to be puzzled out.

Truly New Words

Picking out the meaning of a new word in the first place is the hardest. Lila Gleitman (nn1) has developed a simple method to show what an enigma it can be: she just asks people to learn a new word in an action situation, which should be easier than understanding an emotion term (like embarrass). Suppose I say “John loves the bloothe” and it has a picture of him hitting a baseball with a bat. What does “bloothe” mean? It could mean “bat,” “ball,” “game,” “swing,” “sport,” “challenge.” If I use a verb, things become more obscure: "John bloothes it completely" and there is a picture of John cutting a tree with a saw, we don't know whether "blooth" means "remove,” "saw,” "cut,” "hate,” “enjoy” or what. How we use both circumstance and syntax to triangulate meanings remains a mystery. There is no reason to think that children are different from adults in their mental ability to refer. Yet each grammar marks its own routes to reference, so there is much to acquire.

Are Words Good Pointers? What’s the Point of Pointing?

Even adding gestures will not do the trick. The physical act of pointing itself can be mystifying. Young nursery school teachers sometimes point at something in the distance, like a tree, unaware that their charges are stubbornly looking at the teacher's finger, not the tree. (nn2) So pointing is itself an ambiguous gesture. Pointing words (deictics) like this, that, here, there, nearby seem exact but tolerate more ambiguity than pointing fingers. Still these pop up quickly in the language of children as if children already knew how those ambiguities work. We will try to see if they do.

Inevitable Subjectivity

Reference is always half subjective, half reality. Words inevitably include more meaning than just physical features. Suppose I say, "have you read that?" pointing at an object, using one of the words that seems designed to point at a specific object. But I do mean not literally that book, nor even one with the same print or the same size. If I read Moby Dick and you did too, you could answer "yes,” even though mine was paperback and yours hardback. (nn3) "Read that book" means the content of the book, not the physical paper and print. If I read it out loud mindlessly, like a parrot, then I would not have really read it either. To "read" means to read and to understand. And to understand is unavoidably linked to slightly different understandings because each person's experience is different. Thus we rapidly move from a definition of a book to a definition of book which includes parts of ourselves.

That's actually how we use the word "understand" too. We say things like "you can't really understand a storm until you have been in one in a ship on the ocean.” (nn4) This equates understanding with internal sensual experience, not intellectual awareness of, say, weather maps. This is a caricature of the results of a century of philosophical discussion about reference. Although those discussions are abstruse, to say the least, they contain an insight we must be aware of when we speak with children. Every object word is partly subjective, for adults as well as for children. No words give pure, absolute reference.

A three-year-old, even a two-year-old, visiting a friend, might easily say "oh I have that at my house" looking at a toy. This demonstrates that the child’s version of that involves reference to an abstraction, not an actual specific object. Were it otherwise every parent would have to explain "oh he means that he has a toy just like your toy, don't worry, that toy will not magically move from here to there." No such instruction is needed because the child grasps this abstraction immediately. We no more need to instruct a child about the abstract nature of words than we need to tell a child how to make a three-dimensional reality out of a two-dimensional retinal image. One could easily imagine an organism where referentiality is actual and not abstract. Perhaps salmon returning to their place of birth have a brain marking that refers to that place and nothing else. Humans do not happen to be such an organism.

How to Catch the Emergence of Reference

To see how children learn about reference, I think one should go right out and ask some pertinent questions of children. Here are some simple and obvious ones that can show that even the word this refers to an idea of an object and not an actual object.

Exploration 5.1: thisness, hereness
Suppose we just try to follow in the steps of our example:
Ask a child: "Does your friend have this book at her house?"
in a situation where one knows that she does and see what the child says. She might say "well she has one like it,” or she might just say "yes.” Or she might say, "that's silly, how could she, the book is right here.”

Caption: Deictic pointers without fixed reference

Another probe of this would be to use pictures. A picture of a dog can be in two places, but the same particular picture cannot be in two places. (If you think it can, then you have seen just the ambiguity in language that we are trying to talk our way around, but cannot really escape.)

Exploration 5.2: Doggedness
Look at these two pictures which have the same dog in two places.
Ask a child: “Can you see a cat next to this dog?”
[Pointing to Picture 1 but identical dog is in Picture 2]
Picture 1 of Picture 2 of
Dog Dog +Cat
Bird+Cat
If they allow the abstract notion of a picture of a of dog to equal “this dog,” then they will notice that picture 2 allows a “yes” answer. If they understand “this dog” to be only picture 1—then the child will answer “no.”

Caption: Abstract reference for ordinary nouns

If this can mean only what is literally pointed to--then the Dog in picture (1) is clearly not next to a cat. It seems quite unlikely that the child will say “No it is right there all by itself,” pointing to the same Picture 1, but it is not impossible, particularly for a young child or perhaps an autistic one that may understand context differently. Use of this for two different objects might be seen as a confusing switch, much like the jumping back and forth between "I" and "you" that many children stumble over. An adult says "Are you hungry" and the child says "yes, you are hungry" accepting "you" as his name. Would the same child resolutely point to Picture 1?

Exploration 5.3: Variable Hereness
The word here is subject to the same kind of variation. The exploration could not be simpler:
Say “put your finger here” while you touch your own nose.
Will the child touch your nose or her own nose?

Caption: Self-referential “here”

Would the autistic child only touch your nose? Both responses are right, but if the child touches her own nose, then she has construed “here” as an abstraction with a hidden capacity for variation. “Here” is not exclusively where the Speaker is, but it can cover the Hearer’s “here” too. (To be sure that the child does not just imitate the gesture, one could point with a finger to your nose, but say: “put a penny here” and see what happens.)

It means that there is something like an algebraic variable X hidden inside the word here that allows the listener to introduce a new context. This is not an odd wrinkle of language--perhaps the source of a family laugh—-but it is really the essence of grammar and to use any human grammar, a child must be comfortable with the variable nature of reference.

Abstract the-ness

Abstract reference is not rare. It is present constantly. The use of a introduces a conceptual set that is unbounded. “Have you got a pencil or a cigarette?” means any one we can imagine. Once introduced, a turns into the, but still may not pick out a unique object, as we can see here:

Exploration 5.4: How a turns into the
Suppose someone says at dinner: “Everybody has a spoon.”
[Speaker points at his spoon.]
Picture: people and spoons around a table
“Can you pick up the spoon and put it in your mouth?”
Now “the spoon” in fact refers to a whole set of spoons.
Prediction: child will pick up his own spoon, not the speaker’s

Caption: Non-specific the

Note also that “the spoon” is not any of the spoons on the table either: you cannot reach over and take your neighbor's. It must be linked back to the "a spoon" of the previous sentence, which was yours.

Extension:

Now for contrast one could try:

“Can you pick up THIS fork and put it in your mouth?”

Then we would expect exactly that the child would lean over and get your fork.

Pedagogically, perhaps situations, not explanations, are all one needs to get the right. Working as a United Nations volunteer with my family in Sarajevo, I was thrust into a Second Language English class where they were explaining English articles to eleven-year-olds—-since Serbo-Croatian has none. The teacher said “a marks an unspecified person place or thing.” Asked to teach on the spot, I just used all my acquisition experiments instead and said, look when something is new, put in “a,” and as soon as it is there, it will magically turn into “the” like this: “I have a hat. The hat is red.” Although they generally refused to do homework, I said “come back tomorrow with ten examples like this.” They did, seemed to understand it perfectly, and loved it, which indicated once again that children just need exposure to distinctive situations, not explanation, to learn grammar. At the right moment, it would be a good game for any child, especially for children who seem to have a language deficit.

We still need to ask: why does a turn into the? It is actually a profound mental transformation. A child must understand that "a" turns into "the" in most sentences, but that it fundamentally introduces a mental set. The conceptual move from choice to chosen is a move from a general perspective to a particular perspective. That is, an arbitrary member of a set “a” turns into a very specific “the” (though it is still not necessarily a single physical object). Language finesses a major mental shift with tiny words. A child may have the concepts, but he must connect them to the tiny indicator words too.

All this exposition is needed to give a partial, non-technical explanation of something that feels like common-sense. It is because, indeed, common-sense itself has abstract principles buried in it just like vision has principles of geometry buried in it, which are much more complex to explain in words than just opening your eyes.

Michael Maratsos (nn5) did some clever experiments demonstrating the contrast between definite and indefinite. First he had a child knock over one of a row of cups. Then he asked "what did he do" and the children would say "knocked over a cup,” indicating that they were aware that it was one of a set. Then he put more toys before the child:

“Here is a row of ducks”

[put in a row of ducks]

adult: "take a duck" [child chooses one]

a. “Now give me a duck”

b. “Now give me the duck”

Answer (a) naturally invites the child to select a new duck and give it to you. But (b), irritatingly, calls for the child to surrender exactly the duck he just acquired, even though a whole row of them are present. Maratsos found that children at 2 years 9 months readily understood (b), but this result has not been consistently replicated (some children choose a new duck for “the”). This method is easily explored with anything handy: