might be, or the principles behind these rules, is that sentences in languages are not just

linear strings of words, or if we think about it not so much in terms of their representation

on a page but in terms of speech, they’re not just temporal sequences of words, not

just one word after another.

It turns out that it’s a crucial fact about the syntax of language that sentences have

an internal structure. Words form groups, and these groups of words can form larger

groups. These groupings of words are called “constituents”. And it seems that children

are very sensitive to this from the earliest stages of acquisition. So when children postulate

possible rules for their language, something that we can see from the kinds of sentences

that they produce, which are not necessarily the same as those that adults produce, given

the stages that children go through, we see that children make certain mistakes, that

is they go through stages where the system of rules they’ve developed is different

from that of the adults, but there are certain types of mistakes, certain types of other

systems that they come up with, and certain types that they don’t come up with. And

it seems that in general all the rules that children postulate are sensitive to the kind

of structure that you have in sentences. That is, children are building on the idea that

sentences do not consist of just one word after another, but rather that there’s an

internal structure.

We can give one example of this, which is a classic demonstration from the literature,

which you’ll see in many places. If you think about how a child might try to acquire

the rule that produces yes/no sentences in English, that is, sentences that are also

called “polar questions”, questions that can be answered yes or no, suppose for example

that the child hears the sentence “Is the girl tall?” or the sentence “Can the girl

see the boy who is holding the plate?”. And supposing also that the child realises

that these sentences are the yes/no questions that correspond, respectively, to “The girl

is tall.” and “The girl can see the boy who is holding the plate.” What kind of

rule might a child come up with to derive the question from the declarative, from the

statement? So what might a child propose is a possible rule of English to relate questions,

polar questions, to statements?

If sentences are just sequences of words, one word after another, then an entirely plausible

rule that we could come up with on the basis of data like that would be that you take the

first auxiliary word in the declarative sentence and you move it to the beginning of the sentence

to form the polar question, you “front” it to the beginning of the sentence. That

rule would absolutely characterise what we’ve already seen. But it would also mean that

if you had a sentence like “The boy that is holding the plate can see the girl.”,

the corresponding question would be “Is the boy who holding the plate can see the

girl?”. That, however, is an impossible sentence of English. It’s ungrammatical.

What that shows us, for one thing, is that that isn’t the correct rule for forming

yes/no questions in English. Not only that, but it seems to be the case that children

do not entertain that as a possible rule for forming yes/no questions in English. What

actually is the rule for forming yes/no questions in English is not that you take the first

auxiliary in the declarative ad move it to the beginning, but that you take the first

auxiliary after a certain group of words, and you move that to the beginning of the

sentence. So it’s actually the first auxiliary after the subject of the sentence, where the

subject of the sentence might be a single word or it might be a sequence of words, such

that all of those words form a constituent. So the rule for forming yes/no questions is

sensitive to the structure of the sentence, not just to the sequence of words just one

after another. So what we see from this, then, is that the rules of syntax are sensitive

to the structure of sentences, where by structure we mean that words form groups, groups that

are typically called in the syntactic literature “constituents”. So we talk about sentences

being composed by a number of constituents. And those constituents are referenced in the

rules of syntax in any language, and the rules that children hypothesise make reference to

these constituents.

It’s important to realise that constituents in a sentence can themselves contain smaller

constituents. So, if you take a phrase like “The boy who is holding the plate.”, that

itself is a constituent, but it contains smaller ones. For example, the plate is also a constituent.

So we can look at the structure of sentences, and a sentence, for example, might consist

of two constituents, each one of which itself might consist of two constituents, so that

you wind up with a total of four that the sentence broke down first into two major constituents,

and each one of those into two smaller constituents. It’s possible one of those might consist

itself of two smaller constituents, and so on.

If you took a very simple sentence like “The girl read the book.”, that sentence contains

a number of smaller constituents: “the girl” itself is a constituent, “the book” is

a constituent”, it turns out that “read the book” is a constituent, but it seems

there is no constituent which consists of just “The girl read”, and nothing else.

That is, there’s a constituent that contains those (“the girl read”), but it’s the

whole sentence. So, some words within that sentence form constituents, some of those

constituents combine into larger ones, and some words do not form constituents. Notice

this is a claim about how the language actually works. Not the way it should work. What we’re

claiming here is this actually reflects the reality about the syntax of the language.

At the moment I’ve just stated it as a fact. What you’re trying to do as a syntactician

is to find out what the facts are, and that position I’ve just taken, I would need to

be able to give arguments that those are the facts and not something else. That’s not

something I’m going to do now, but we’ll be coming back to this later in the course.