What is the best way of looking at them?
If a child is not talking and understanding in the expected way it is obviously worth taking account of all aspects of his or her life –brothers and sisters, grandparents, the family heritage, and more. It is obvious that long experience of speech and language is crucial for development – as shown by the tragic case of children brought up by wolves or dogs or in criminal isolation. The effects are typically life-long. And in obvious ways there is variation between the forms of speech and language in adjacent communities where the only barrier is by geographical address or cultural identity. All aspects of life have a bearing on speech and language.
But is it useful to regard a word as just the shortest possible sort of utterance? Are there just growingly complex utterances and other interactions involving family structure? Resources? Sex and gender? Social heritage?
Or are there phenomena which it makes sense to look at just in terms of speech and language, as linguistic structures? If the first words which a child puts together are “In bath” as he or she is being put into the bath, with two words of different sorts in a particular order, the order dictated by English, is this a first step towards an amazingly complex but discrete capacity? I believe that it is. But that is not a universal view. On a different view to mine, “In bath” just repeats a commonly used phrase.
Child Directed Speech and Language
Children sometimes hear things like: “Put your army-warmy in there,” perhaps as they are being dressed. This is often called ‘Child Directed Speech and Language’ (CDSL) or ‘Infant Directed Speech’ (IDS) or ‘motherese’, Such talk has distinct characteristics. Sentences are kept short. Subordinate clauses are avoided. Words are remade so as to conform to stereo-typical templates.
Personally, I have observed such talk only from mothers. It is sometimes also used with pets, but not in the expectation of speech or language, and only by women owners, as far as I know.
CDSL or IDS is believed by some theorists of speech and language acquisition to give children criterial building-blocks from which they can go on to learn more complex forms. But in some homes and cultures there is no CDSL or IDS. And yet whether or not children hear it, they still learn to talk.
I believe that that the purpose of CDSL or IDS is to express love and affection, and that it is entirely marginal for speech and language. So if a child’s speech or language is delayed or disordered, I don’t think this style of talk will help solve the problem. It may strengthen the bonds of love. But if it is used in a non-spontaneous, artificial way, this may get in the way of more spontaneous interactions. It seems to me these natural, loving, spontaneous interactions are by far and away the most important sort of linguistic experience for children to have.
You might say: My three-year-old child still hasn’t started putting two words together and is said to have a serious language delay. Isn’t the priority just to get my child putting two words together, and never mind the complexities?
Yes, indeed, enabling such a child to put words together is a priority. But it is also important to understand the whole process.
Objectivity and reliability
It is sometimes said that in linguistics there are no real facts, as facts are understood in physics and chemistry, just a highly-subjective and quite novel methodology, one that compares poorly with older and supposedly more reliable procedures and methodologies.
But that both muddles the time-line and understates the descriptive accuracy. The world’s first linguist was Panini who wrote an elegantly compact description of Sanskrit, the language of Indian Brahmins, as spoken around 500 years before the Christian era. Sanskrit was and similar to both Classical Latin and Ancient Greek. By one of Panini’s 3,959 observations, some consonants changed before a vowel. We have something similar in the two commonest words in English a and the. A gets the sound N as in an before a vowel. And the vowel in the changes.
Lexical words and functors
Updating Panini’s discoveries to the modern era, the distinctive property of a and the is that they belong to a class of elements known as ‘functors’ which approximately alternate with what are variously known as ‘lexical’ or ‘encyclopaedic’ or ‘information carrying’ or ‘content’ words, i.e. nouns, verbs, adjectives, adverbs and prepositions. The fact that we know this is part of What makes us human. Unlike lexical words, functors have no meaning other than in relation to a lexical word. All languages have both functors and lexical words, and in actual speech they tend to alternate with one another. I believe that the child comes to the task of learning language ‘looking for’ this sort of contrast between lexical words and the purely formal role of functors. The difference is flagged up for the learner by the fact that in a way quite unlike lexical words:
• Functors can vary according to whethe the next sound is a vowel as in the and a;
• Or lose everything apart from the final consonant like have in “We’ve done the washing.”
•. Or get stuck on to another functor and lose the vowel, with a change to both of the sound structures, as in don’t from do not or, in Texan English, y’all from you all;
•. Or do this and hop across another functor as in the Scots name of the lament ‘Will ye no come back again’. In the original form there is no hopping, but with the contraction, it becomes ‘Won’t you come back again’ with the n’t hopping over the Scots ye.
Not all of these things happen in every language. Some languages have things which don’t happen in English. Many languages have functors to express the difference between certainty and uncertainty. These interactions between the lexical words and the functors are the key points in the grammar, not just of English, but of every human language. Children are paying close attention to them at least as early as two. As pointed out by Jamal Ouhalla, these are the very words which children mostly leave out. But they are not details.
There are key points for children learning English. But crucial though everyday experience obviously is, it is not enough for children to learn things like the close relation between not and n’t in don’t and won’t, no matter how far this is broken down into child-friendly chunks. This and a number of other phenomena in English and all other languages that have been studied so far suggest that there has to be something which is now ‘hard-wired’ into the human genome. One key aspect of this is the way the learner ‘knows’ that words get ‘merged’ together on the basis of contrast, as between lexical words and functors.