In certain of these original formations we may detect a fine feeling for verbal analogy. Thus a French boy, after killing the ‘limaces’ (snails) which were eating the plants in the garden, dignified his office by styling himself a ‘limarcier’; where the inventive faculty was no doubt led by the analogy of ‘voiturier’ formed from ‘voiture’.^[99]

In other verbal formations it is difficult to determine the model which is followed. Signorina Lombroso gives a good example. A little girl of two and a half years had observed that when her mother allowed her to take, eat, or drink something, she would say ‘prendilo’ (take it), ‘bevilo’ (drink it), or ‘mangialo’ (eat it). She proceeded to make a kind of adjective or substantive out of each of these, asking ‘é prendilo?’ ‘é bevilo?’ ‘é mangialo?’ i.e., ‘Is it takable or a case of taking?’ etc., when she wanted to take, drink, or eat something.^[100] By such skilful artifices does the little word-builder find his way to the names which he has need of.

In certain cases these original constructions are of a more clumsy order and due to a partial forgetfulness of a word and an effort to complete it. Thus a boy of four spoke of being ‘sorrified,’ where he was evidently led out of the right track by the analogy of ‘horrified’. The same little boy who talked of his ‘digs’ used the word ‘magnicious’ for ‘magnificent’. This is a choice example of word-transformation. No doubt the child was led by the feeling for the sound of this termination in other grand words, as ‘ambitious’. Possible, too, he might have heard the form ‘magnesia’ and been influenced by a reminiscence of this sound-complex. The talk of ‘Jeames’ with which Mr. Punch makes us acquainted is full of just such delightful missings of the mark in trying to reproduce big words.

Sentence-building.

We may now follow the child in his later and more ambitious linguistic efforts. The transition to this higher plane is marked by the use of the completed form of thought, the sentence.

At first, as already pointed out, there is no sentence-structure. The child begins to talk by using single words. These words consist of what we call substantives, as ‘Mamma,’ ‘nurse,’ ‘milk,’ a few adjectives, as ‘hot,’ ‘nice,’ ‘good,’ a still smaller number of adverbial signs, as ‘ta-ta,’ or ‘away,’ ‘over,’ ‘down,’ ‘up,’ and one or two verb-forms, apparently imperatives, as ‘go’. The exact order in which these appear, and the proportion between the different classes of constituents at a particular age, say two and a half or three, appear to vary greatly. Words descriptive of actions, though very few at first, appear to grow numerous in a later stage.^[101]

In speaking of these words as substantives, adjectives, and so forth, I am merely adopting a convenient mode of description. We must not suppose that the words as used in this simple disjointed talk have their full grammatical value. It is not generally recognised that the single-worded utterance of the child is an abbreviated sentence or ‘sentence-word’ analogous to the sentence-words found in the simplest known stage of adult language. As with the race so with the child, the sentence precedes the word. Moreover, each of the child’s so-called words in his single-worded talk stands for a considerable variety of sentence-forms. Thus the words in the child’s vocabulary which we call substantives do duty for verbs and so forth. As Preyer remarks, ‘chair’ (stuhl) means ‘There is no chair,’ ‘I want to be put in the chair,’ ‘The chair is broken,’ and so forth. In like manner ‘dow’ (down) may mean ‘The spoon has fallen down,’ ‘I am down,’ ‘I want to go down,’ etc.^[102] The particular shade of meaning intended is indicated by intonation and gesture.

This sentence-construction begins with a certain timidity. The age at which it is first observed varies greatly. It seems in most cases to be somewhere about the twenty-first month, yet I find good observers among my correspondents giving as dates eighteen and a half and nineteen months; and a friend of mine, a Professor of Literature, tells me that his boy formed simple sentences as early as fifteen months. We commonly have at first quite short sentences formed by two words in apposition. These may consist of what we should call an adjective added to and qualifying a substantive, as in the simple utterance of the child C., ‘Big bir’ (bird), or the exclamation, ‘Papa no’ (Papa’s nose); or they may arise by a combination of substantives, as in the sentence given by Tracy, ‘Papa cacker,’ i.e., ‘Papa has crackers,’ and one quoted by Preyer, ‘Auntie cake’ (German, ‘Danna Kuha,’ i.e., ‘Tante Kuche’) for ‘Auntie has given me cake’; and in a somewhat different example of a compound sentence also given by Preyer, ‘Home milk’ (German, ‘Haim Mimi’), interpreted as ‘I want to go home and have milk’. In the case of one child about the age of twenty-three months most of the sentences were composed of two words, one of which was a verb in the imperative. The love of commanding, so strong in the child, makes the use of the imperative, as is seen in this case, very common. M.’s first performance in sentence-building (at eighteen and a half months) was, ‘Mamma, tie,’ i.e., ‘tie gloves’.

Little by little the learner manages longer sentences, economising his resources to the utmost, troubling nothing about inflections or the insertion of prepositions so as to indicate precise relations, but leaving his hearer to discover his meaning as best he may; and it is truly wonderful how much the child manages to express in this rude fashion. A boy nineteen and a half months old gave this elaborate order to his father: ‘Dada toe toe ba,’ that is, ‘Dada is to go and put his toes in the bath’. Pollock’s little girl in the first essay at sentence-building, recorded at the age of twenty-one and a half months, actually managed a neat antithesis: ‘Cabs dati, clam clin,’ that is to say, ‘Cabs are dirty, and the perambulator is clean’. Preyer’s boy in the beginning of the third year brought out the following, ‘Mimi atta teppa papa oi,’ that is to say, ‘Milch atta Teppich Papa fui,’ which appears to have signified, “The milk is gone, it is on the carpet, and papa said ‘Fie’”. It may be added that the difficulties of deciphering these early sentences is aggravated by the frequent resort to slurs, as when a child says, ‘m’ out’ for ‘take me out,’ ‘’t on’ for ‘put it on’.

The order of words in these first tentative sentences is noticeable. Sometimes the subject is placed after the predicate, as in an example given by Pollock, ‘Run away man,’ i.e., ‘The man runs (or has run) away,’ and in the still quainter example given by the same writer, ‘Out-pull-baby ’pecs (spectacles),’ i.e., ‘Baby pulls or will pull out the spectacles’. In like manner the adjective used as predicate may precede the subject, as in the examples given by Maillet, ‘Jolie la fleur,’ etc.^[103] Sometimes, again, the object comes before the verb, as apparently in the following example given by Miss Shinn: a little girl delighted at the prospect of going out to see the moon exclaimed, “Moo-ky (sky), baby shee (see)”.^[104] Here is a delightful example of a transposition of subject and object. A boy two years and three months asked, ‘Did Ack (Alec) chocke an apple?’ i.e., ‘Did an apple choke Alec?’ though in this case we very probably have to do with a misunderstanding of the action choke. Other kinds of inversion occur when more complex experiments are attempted, as in connecting ‘my’ with an adjective. Thus one child said prettily, ‘Poor my friends’;^[105] which archaic form may be compared with the following Gallic-looking idiom used by M. at the age of one year ten months: ‘How Babba (baby, i.e., herself) does feed nicely!’ The same little girl put the auxiliary out of its place, saying, ‘Tan (can) Babba wite’ for ‘Baby can write,’ though this was probably a reminiscence of the question-form.

These inversions of our familiar order are suggestive. They have some resemblance to the curious order which appears in the spontaneous sign-making of deaf-mutes. Thus a deaf-mute answered the question, ‘Who made God?’ by saying, “God made nothing,” i.e., “nothing made God”. Similarly the deaf-mute Laura Bridgman expressed the petition, ‘Give Laura bread,’ by the form, ‘Laura bread give.’^[106] Such inversions, as we know, are allowable and common in certain languages, e.g., Latin. The study of the syntax of child-language and of the sign-making of deaf-mutes might suggest that our English order is not in certain cases the most natural one.

A somewhat similar inversion of what seems to us the proper order appears in the child’s first attempts at negation. The child C. early in his third year expressed the idea that he was not going into the sea thus: ‘N. (his own name) go in water, no’. Similarly Pollock’s child expressed acquiescence in a prohibition in this manner, ‘Baby have papa (pepper) no,’ where the ‘no’ followed without a pause. The same order appears in the case of French children, e.g., ‘Papa non,’ i.e., ‘It is not Papa,’ and seems to be a common, if not a universal form of the first half-spontaneous sentence-building. Here again we see an analogy to the syntax of deaf-mutes, who appear to append the sign of negation in a similar way, e.g., ‘Teacher I beat, deceive, scold no,’ i.e., ‘I must not beat, deceive, scold my teacher’. We see something like it, too, in the formations of savage-languages, as when ‘fool no’ comes to be the sign of ‘not fool,’ that is of wise.^[107] When ‘not’ comes into use it is apt to be put in a wrong place, as when the little girl M. said, ‘No Babba look’ (i.e., ‘Babba will not look’), and ‘Mr. Dill not did tum’ for ‘Mr. Gill did not come’.^[108]

Another closely related characteristic of this early childish sentence-building is the love of antithesis under the form of two balancing statements. Thus a child will often oppose an affirmative to a negative statement as a means of bringing out the full meaning of the former. The boy C., for example, would say, ‘This a nice bow-wow, not nasty bow-wow’. The little girl M. said, ‘Boo (the name of her cat) dot (got) tail; poor Babba dot no tail,’ proceeding to search for a tail under her skirts. This use of a negative statement by way of contrast or opposition to an affirmative grew in the case of one child aged two years and two months into a habit of description by negations. Thus an orange was described by the saying, ‘No, ’tisn’t apple,’ porridge by ‘No, ’tisn’t bread and milk’. It is interesting to note that deaf-mutes proceed in a similar fashion by way of antithetic negative statement. Thus one of these expressed the thought, ‘I must love and honour my teacher,’ by the order, ‘Teacher I beat, deceive, scold no!—I love honour yes!’^[109]

These first essays in the construction of sentences illustrate the skill of the child in eking out his scanty vocabulary by help of a metaphorical transference of meaning. Taine gives a charming example of this device. A little girl of eighteen months had acquired the word ‘Coucou’ as used by her mother or nurse when playfully hiding behind a door or chair, and the expression ‘ça brûle’ as employed to warn her that her dinner was too hot, or that she must put on her hat in the garden to keep off the hot sun. One day on seeing the sun disappear behind a hill she exclaimed, ‘A bûle coucou’.^[110]

It is a fearful moment when the child first tries his hand at inflections, and, more especially in our language, those of verbs. Pollock’s child made the attempt, and successfully, at the age of twenty-two months. Such first essays are probably examples of pure imitation, the precise forms used having been previously heard from others. Hence while they show a growing power of thought, of a differencing of the relations of number and time, they do not involve verbal construction properly so called. This last appears as soon as the child carries over his knowledge of particular cases of verbal inflection and applies it to new words. This involves a nascent appreciation of the reason or rule according to which words are modified. The development of this feeling for the general mode of verbal change underlies all the later advance in correct speaking.

While the little explorer in the terra incognita of language can proceed safely in this direction up to a certain point he is apt, as we all know, to stumble now and again; nor is this to be wondered at when we remember the intricacies, the irregularities, which characterise a language like ours. In trying, for example, to manage the preterite of an English verb he is certain, as, indeed, is the foreigner, to go wrong. The direction of the error is often in the transformation of the weak to the strong form; as when ‘screamed’ becomes ‘scram,’ ‘split’ (preterite) ‘splat’ or ‘splut,’ and so forth. In other cases the child wall convert a strong into a weak form, as when Laura Bridgman, like many another child, would say, ‘I eated,’ ‘I seed,’ and so forth.^[111] Sometimes, again, delightful doublings of the past tense occur, as ‘sawed’ for ‘saw,’ ‘eatened’ for ‘eaten,’ ‘didn’t saw’ for ‘didn’t see,’ ‘did you gave me?’ for ‘did you give me?’ Active and passive forms are sometimes confused, as when M. said ‘not yike being picking up’ for ‘not like being picked up,’ etc. It is curious to note the different lines of imitative construction followed out in these cases.

One thing seems clear here: the child’s instinct is to simplify our forms, to get rid of irregularities. This is strikingly illustrated in the use of the heterogeneous assemblage of forms known as the verb ‘to be’. It is really hard on a child to expect him to answer the question, ‘Are you good now?’ by saying, ‘Yes, I am’. He says, of course, ‘Yes, I are’. Perhaps the poor verb ‘to be’ has suffered every kind of violence at the hands of children.^[112] Thus the child M. used the form ‘bēd’ for ‘was’. Professor Max Müller somewhere says that children are the purifiers of language. Would it not be well if they could become its simplifiers also, and give us in place of this congeries of unrelated sounds one good decent verb-form?

Other quaint transformations occur when the child begins to combine words, as when M. joining adverb to verb invented the form of past tense ‘fall downed’ for ‘fell down’. Another queer form is ‘Am’t I?’ used for ‘am I not?’ after the pattern of ‘aren’t we?’ An even finer linguistic stroke than this, is ‘Bettern’t you?’ for ‘Had you not better?’ where the child was evidently trying to get in the form ‘hadn’t you,’ along with the awkward ‘better,’ which seemed to belong to the ‘had,’ and solved the problem by treating ‘better’ as the verb, and dropping ‘had’ altogether.

A study of these solecisms, which are nearly always amusing, and sometimes daintily pretty, is useful to mothers and young teachers by way of showing how much hard work, how much of real conjectural inference, enters into children’s essays in talking. We ought not to wonder that they now and again slip; rather ought we to wonder that, with all the intricacies and pitfalls of our language—this applies of course with especial force to the motley irregular English tongue—they slip so rarely. As a matter of fact, the latter and more ‘correct’ talk—which is correct just because the child has stored up a good stock of particular word-forms, and consequently has a much wider range of pure uninventive imitation—is less admirable than the early inventive imitation; for this last not only has the quality of originality, but shows the germ of a truly grammatical feeling for the general types or norms of the language.

The English child is not much troubled by inflections of substantives. The pronouns, however, as intelligent mothers know, are apt to cause much heart-burning to the little linguist. The mastery of ‘I’ and ‘you,’ ‘me,’ ‘mine,’‘mine,’ etc., forms an epoch in the development of the linguistic faculty and of the power of thought which is so closely correlated with this. Hence it will repay a brief inspection.

As is well known, children begin by speaking of themselves and of those whom they address by names, as when they say, ‘Baby good,’ ‘Mamma come’. This is sometimes described as speaking “in the third person,” yet this is not quite accurate, seeing that there is as yet no distinction of person at all in the child’s language.

The first use of ‘I’ and ‘you’ between two and three years is apt to be erroneous. The child proceeds imitatively to use ‘I,’ ‘me,’ ‘my’ for ‘you’ and ‘your’. Thus one child said, ‘What I’m going to do,’ for, ‘What are you going to do?’ In this case, it is plain, there is no clear grasp of what we mean by subject, or of the exact relation of this subject to the person he is addressing.

Yet along with this mechanical repetition of the pronominal forms we see the beginnings of an intelligent use of them. So far as I can ascertain most children begin to say ‘me’ or ‘my’ before they say ‘you’. Yet I have met with one or two apparent exceptions to this rule. Thus the boy C. certainly seemed to get hold of the form of the second person before that of the first, and the priority of ‘you’ is attested in another case sent to me. It is desirable to get more observations on this point.

To determine the exact date at which an intelligent use of the first person appears, is much less easy than it looks. The ‘I’ is apt to appear momentarily and then disappear, as when M. at the age of nineteen months three weeks was observed to say ‘I did’ once, though she did not use ‘I’ again until some time afterwards. Allowing for these difficulties it may be said with some degree of confidence that the great transition from ‘baby’ to ‘I’ is wont to take place in favourable cases early in the first half of the third year. Thus among the dates assigned by different observers I find, twenty-four months, twenty-five months (cases given by Preyer), between twenty-five and twenty-six (Pollock), twenty-seven months (the boy C.). A lady friend tells me that her boy began to use ‘I’ at twenty-four months. In the case of a certain number of precocious children this point is attained at an earlier date. Thus Preyer quotes a case of a child speaking in the first person at twenty months. Schultze gives a case at nineteen months. A friend of mine, a Professor of English Literature, whose boy showed great precocity in sentence-building, reports that he used the forms ‘me’ and ‘I’ within the sixteenth month. Preyer’s boy, on the other hand, who was evidently somewhat slow in lingual development, first used the form of the first person ‘to me’ (mir) at the age of twenty-nine months.

The precise way in which these pronominal forms first appear is very curious. Many children use ‘me’ before ‘I’. Preyer’s boy appears to have first used the form ‘to me’ (mir). ‘My’ too is apt to appear among the earliest forms. In such different ways does the child pass to the new and difficult region of pronominal speech.

The meaning of this transition has given rise to much discussion. It is plain, to begin with, that a child cannot acquire these forms as he acquires the name ‘papa,’ ‘nurse,’ by a direct and comparatively mechanical mode of imitation. When he does imitate in this fashion he produces, as we have seen, the absurdity of speaking of himself as ‘you’. Hence during the first year or so of speech he makes no use of these forms. He speaks of himself as ‘baby’ or some equivalent name, others coming down to his level and setting him the example.

The transition seems to be due in part, as I have elsewhere pointed out, to a growing self-consciousness, to a clearer singling out of the ego or self as the centre of thought and activity, and the understanding of the other ‘persons’ in relation to this centre. Not that self-consciousness begins with the use of ‘I’. The child has no doubt a rudimentary self-consciousness when he talks about himself as about another object: yet the use of the forms ‘I,’ ‘me,’ may be taken to mark the greater precision of the idea of ‘self’ as not merely a bodily object and nameable just like other sensible things, but as something distinct from and opposed to all objects of sense, as what we call the ‘subject’ or ego.

While, however, we may set down this exchange of the proper name for the forms ‘I’ and ‘me’ as due to the spontaneous growth of the child’s intelligence, it is possible that education exerts its influence too. It is conjecturable that as a child’s intelligence grows, others in speaking to him tend unknowingly to introduce the forms ‘I’ and ‘you’ more frequently. Yet I am disposed to think that the child commonly takes the lead here. However this be, it is clear that growing intelligence, involving greater interest in others’ words, will lead to a closer attention to these pronominal forms as employed by others. In this way the environment works on the growing mind of the child, stimulating it to direct its thoughts to these subtle relations of the ‘me and not me,’ ‘mine and thine’. The more intelligent the environment the greater will be the stimulating influence: hence, in part at least, the difference of age when the new style of speech is attained.^[113]

The acquirement of these pronominal forms is a slow and irksome business. At first they are introduced hesitatingly, and alongside of the proper name; the child, for example, saying sometimes, ‘Baby’ or ‘Ilda,’ sometimes ‘I’ or ‘me’. In some cases, again, the two forms are used at the same time in apposition, as in the delightful form not unknown in older folk’s language, ‘Hilda, my book’. The forms ‘I’ and ‘me’ are, moreover, confined at first to a few expressions, as ‘I am,’ ‘I went,’ and so forth. The dropping of the old forms, as may be seen by a glance at the notes on the child C., and at Preyer’s methodical diary, is a gradual process.

Quaint solecisms mark the first stages of the use of these pronouns. As in the case of the earlier use of substantives, one and the same form will be used economically for a variety of meanings, as when ‘me’ was by the boy C. used to do duty for ‘mine’ also, and ‘us’ for ‘ours’. Here it is probable there is a lack of perfect discrimination. The connexion between the self and its belongings is for all of us of the closest. When a child of two, who was about to be deprived of her doll, shouted, ‘Me, me!’ may we not suppose that the doll was taken up into the inner circle of the self?^[114] Sometimes in this enrichment of the vocabulary by pronouns new and delightful forms are struck off, as when the little experimenter invents the possessive form ‘she’s’.

The perfect unfettered use of these puzzling forms comes much later. Preyer quotes a case in which a child Olga, aged four years, would say, ‘She has made me wet,’ meaning that she herself had done it. But this perhaps points to that tendency to split up the self into a number of personalities, to which reference was made in an earlier essay.

The third year, which witnesses the important addition of the pronouns, sees other refinements introduced. Thus the definite article was introduced in the case of Preyer’s boy in the twenty-eighth month, in that of an English boy at the age of two years eight months. Prepositions are introduced about the same time. In this way childish talk begins to lose its primitive disjointed character, and to grow into an articulated structure.^[115] Yet the perfect mastery of these takes time. A feeling for analogy easily leads the little explorer astray at first, as when the child M. said ‘far to’ after the model ‘near to’.

Through this whole period of language-learning the child continues to show his originality, his inventiveness. He is rarely at a loss, and though the gaps in his verbal acquisitions are great he is very skilful in filling them up. If, for example, our bright little linguist M., at the age of one year eight and a half months, after being jumped by her father, wants him to jump her mother also, she says, in default of the word ‘jump,’ “Make mamma high”. A boy of twenty-seven months ingeniously said, ‘It rains off,’ for ‘The rain has left off’. Forms are sometimes combined, as when a boy of three years three months used ‘my lone,’ ‘your lone,’ for ‘me alone’ or ‘by myself,’ ‘you alone’ or ‘by yourself’. Another girl, two years ten months, said, ‘No two ’tatoes left,’ meaning ‘only one potato is left’. Pleonasms occur in abundance, as when a boy of two would say, ‘Another one bicca (biscuit),’ and, better still, ‘another more’.

There is one part of this child’s work of learning our language of which I have said hardly anything, viz., the divining of the verbal content, of the meaning we put or try to put into our words. A brief reference to this may well bring this study of childish linguistics to a close.

The least attention to a child in the act of language-learning will show how much of downright hard work goes to the understanding of language. If we are to judge by the effort required we might say that the child does as much in deciphering his mother-tongue as an Oriental scholar in deciphering a system of hieroglyphics. Just think, for example, how many careful comparisons the small child-brain has to carry out, comparisons of the several uses of the word by others in varying circumstances, before he can get anything approaching to a clear idea, answering even to such seemingly simple words as ‘clean,’ ‘old’ or ‘clever’. The way in which inquiring children plague us with questions of the form, ‘What does such and such a word mean?’ sufficiently shows how much thought-activity goes in the trying to get at meanings. This difficulty, moreover, persists, reappearing in new forms as the child pushes his way onwards into the more tangled tracts of the lingual terrain. It is felt, and felt keenly, too, when most of the torments of articulation are over and forgotten. Many of us can remember how certain words haunted us as uncanny forms into the nature of which we tried hard, but in vain, to penetrate.

Owing to these difficulties the little learner is always drifting into misunderstanding of words. Such misapprehensions will arise in a passive way by the mere play of association in attaching the word especially to some striking feature or circumstance which is apt to present itself when the word is used in the child’s hearing. In this way, for example, general terms may become terribly restricted in range by the incorporation of accidentals into their meaning, as when a Sunday school scholar rendered the story of the good Samaritan by saying that a gentleman came by and poured some paraffin (i.e., oil) over the poor man. A word may have its meaning funnily transformed by such associative suggestions, as when a little girl, being told that a thing was a secret, remarked, ‘Well, mamma, ’ou (you) can whisper it in my ear’. As this example shows, a child in his ‘concreting’ fashion tries to get sensible realities out of our names. A mask was called by a boy of six a ‘grimace,’ this abstract name standing to his mind for the grinning face. A like tendency shows itself in the following quaint story. A boy and a girl, twins, had been dressed alike. Later on the boy was put into a ‘suit’. A lady asked the girl about this time whether they were not the twins, when she replied: ‘No, we used to be’. ‘Twin’ was inseparably associated in her mind with the similarity in dress. A somewhat similar effect of association of ideas is seen in the quaint request of the little girl M. that her mamma should ‘smell’ the pudding and make it cool. The action of bringing the face near an object yet so as not to touch it was associated with smelling, as in the little girl who, according to Mr. Punch, had her sense of propriety shocked by some irreverent person who did not “smell his hat” when he took his seat in church. Moral expressions get misunderstood in much the same manner. A little girl of three and a half years, pretending that her mother was her little girl, said: ‘You mustn’t do anything on purpose’. The usual verbal context of this highly-respectable phrase (e.g., ‘You did it on purpose’) had in the child’s mind given it a naughty meaning.

With these losings of the verbal road through associative by-paths may be taken the host of misapprehensions into which children are apt to fall through the ambiguities of our words and expressions, and our short and elliptical modes of speaking. Thus an American child, noting that children were ‘half price’ at a certain show, wanted his mother to get a baby now that they were cheap.^[116] With this may be compared the following: Jean Ingelow tells us she can well remember how sad she was made by her father telling her one day after dancing her on his knee that he must put her down as he ‘had a bone in his leg’.^[117] Much misapprehension arises, too, from our figurative use of language, which the little listener is apt to interpret in a very literal way. It would be worth knowing what odd renderings the child-brain has given to such expressions as ‘an upright man,’ ‘a fish out of water,’ and the like.

In addition to these comparatively passive misapprehensions there are others which are the outcome of an intellectual effort, the endeavour to penetrate into the mystery of some new and puzzling words or expression. Many of us have had our special horror, our bête noire among words, which tormented us for months and years. I remember how I was plagued by the word ‘wean,’ the explanation of which was very properly, no doubt, denied me by the authorities, and by what quaint fancies I tried to fill in a meaning.

As with words, so with whole expressions and sayings. It was a natural movement of childish thought when a little school-girl answered the question of the Inspector, ‘What is an average?’ by saying ‘What the hen lays eggs on’. She had heard her mother say, “The hen lays so many eggs ‘on the average’ every week,” and had no doubt imagined a little myth about this ‘average’. Again, most of us know what queer renderings the child-mind has given to Scripture language. Mr. James Payn tells us that he knew a boy who for years substituted for the words, ‘Hallowed be thy name,’ ‘Harold be thy name’.^[118] In this and similar cases it is not, as might be supposed, defective hearing—children hear words as a rule with great exactness—it is the impulse to give a familiar and significant rendering to what is strange and meaningless.^[119] A friend of mine when a boy was accustomed on hearing the passage, ‘If I say peradventure the darkness shall cover me,’ etc., to insert a pause after ‘peradventure,’ apprehending the passage in this wise: "If I say ‘Peradventure!’—the darkness," etc. In this way he turned the mysterious ‘peradventure’ into a mystic ‘open sesame,’ and added a thrilling touch of magic to the passage. My friend’s daughter tells me that on hearing the passage, “I ... visit the sins of the fathers upon the children unto the third and fourth generation, ... and show mercy unto thousands,” she construed the strange word ‘generation’ to mean an immense number like ‘billion,’ and was thus led to trouble herself about God’s seeming to be more cruel than kind.^[120]

In some cases, too, where the language is simple enough a child’s brain will find our meaning unsuitable and follow a line of interpretation of its own. Mr. Canton relates that his little heroine, who knew the lines in Strumpelpeter—

was told that she would catch a cold, and that she at once replied, “And will the doctor come and shook my head?”^[121] It was so much more natural to suppose that when the doctor came and did something this was carried out on the person of the patient.

There is nothing more instructive in this connexion than the talk of children among themselves about words. They build up quaint speculations about meanings, and try their hand bravely at definitions. Here is an example: A boy of five was instructing his comrade as to the puzzling word ‘home-sick’. He did it in quite a scientific fashion. “It’s like sea-sick, you know: you are sea-sick when you are sick at sea, and so you’re home-sick when you’re sick at home”.

There is something of this same desire to get behind words in children’s word-play, as we call it, their discovery of odd affinities in verbal sounds, and their punning. Though no doubt this contains a genuine element of childish fun, it betokens a more serious trait also, an interest in word-sounds as such, and a curiosity about their origin and purpose. It is difficult for grown-up people to go back in thought to the attitude of the child-mind towards verbal sounds. Just as children show ‘the innocence of the eye’ in seeing the colours of objects as they are and not as our habits of interpretation tend to make them, so they show an innocence of the ear, catching the intrinsic sensuous qualities of a word or a group of words, in a way which has become impossible for us.

This half-playful, half-serious scrutiny of word-sounds leads to the attempt to find by analysis and analogy a familiar meaning in strange words. For example, a little boy about four years old heard his mother speak of nurse’s neuralgia, from which she had been suffering for some time. He thereupon exclaimed, ‘I don’t think it’s new ralgia, I call it old ralgia’. A child called his doll ‘Shakespeare’‘Shakespeare’ because its spear-like legs could be shaken. Another boy of three explained ‘gaiters’ as things ‘to go out of the gate with’. Another said that the ‘Master’ which he prefixed to his name meant that he was master of his dog. A little girl in her third year called ‘anchovies’ ‘ham-chovies’‘ham-chovies’ ‘mermaid’ ‘worm-maid,’ ‘whirlwind’ ‘world-wind,’ ‘gnomes’ ‘no-mans’ (un-menschen), taking pleasure apparently in bringing some familiar element—even when this seems to other ears at least not very explanatory—into the strange jumble of word-sound that surrounded her. A child may know that he is ‘fooling’ in such cases, yet the word-play brings a certain satisfaction, which is at least akin to the pleasure of the older linguist.

This quasi-punning transformation of words is curiously like what may be called folk-etymology, where a foreign word is altered by a people so as to be made to appear significant and suitable for its purpose, as in the oft-quoted forms ‘sparrow grass’ (asparagus) and ‘cray-fish’ (from the French écrevisse, cf. the O. H. German Krebiz), where the attempt to suit the form to the thing is still more apparent.^[122] When, for example, a boy calls a holiday a ‘hollorday,’ because it is a day ‘to holloa in,’ we may say that he is reflecting the process by which adults try to put meaning into strange words, as when a cabman I overheard a few days ago spoke about putting down ashphalt (for ‘asphalt’). Some children carry out such transformation and invention of derivation on a large scale, often resorting to pretty myths, as when the butterflies are said to make butter, or to eat butter, grasshoppers to give grass, honeysuckles to yield all the honey, and so forth.^[123]

A child will even go further, and, prying into the forms of gender, invent explanatory myths in which words are personified and sexualised. Thus a little boy of five years and three months who had learned German and Italian as well as English was much troubled about the gender of the sun and moon. So he set about myth-making on this wise: “I suppose people^[124] think the sun is the husband, the moon is the wife, and all the stars the little children, and Jupiter the maid”. A German girl of six was thus addressed by her teacher: “‘Der’ ist männlich; Was sind ‘Die’ und ‘Das’?” To which she replied prettily: "Die ist dämlich (i.e., ‘ladyish’) und das ist kindlich". The tendency to attribute differences of sex and age to names observable in this last is seen in other ways. An Italian child asked why ‘barba’ (beard) was not called ‘barbo’. With this may be compared the pretty myth of another Italian child that ‘barca’ (boat) was the little girl of ‘barcainolo’ (boatman).^[125]

One other characteristic feature in the child’s attitude towards words must be touched on, because it looks like the opposite of the impulse to tamper with words just dealt with. A child is a great stickler for accuracy in the repetition of all familiar word-forms. The zeal of a child in correcting others’ language, and the comical errors he will now and again fall into in exercising his pedagogic function, are well known to parents. Sometimes he shows himself the most absurd of pedants. ‘Shall I read to you out of this book, baby?’ asked a mother of her boy, about two and a half years old. ‘No,’ replied the infant, ‘not out of dot book, but somepy inside of it.’ The same little stickler for verbal accuracy, when his nurse asked him, ‘Are you going to build your bricks, baby?’ replied solemnly, ‘We don’t build bricks, we make them and then build with them’. In the notes on the boy C. we find an example of how jealously the child-mind insists on the ipsissima verba in the recounting of his familiar stories.

Are these little sticklers for verbal correctness, who object to everything figurative in our language, who, when they learn that a person or an animal has ‘lost his head,’ take the expression literally, and who love nothing better than tying us down to literal exactness, themselves given to ‘word-play’ and verbal myth-making, or have we here to do with two varieties of childish mind? My observations do not enable me to pronounce on this point.

I have in this essay confined myself to some of the more common and elementary features of the child’s linguistic experience. Others present themselves when the reading stage is reached, and the new strange stupid-looking word-symbol on the printed page has to do duty for the living sound, which for the child, as we have seen, seems to belong to the object and to share in its life. But this subject, tempting as it is, must be left. And the same must be said of those special difficulties and problems which arise for the child-mind when two or more languages are spoken. This is a branch of child-linguistics which, so far as I know, has never been explored.