Any serious educational theory must consist of two parts: a conception of the ends of life, and a science of psychological dynamics, i.e. of the laws of mental change. Two men who differ as to the ends of life cannot hope to agree about education. The educational machine, throughout Western civilization, is dominated by two ethical theories: that of Christianity, and that of nationalism. These two, when taken seriously, are incompatible, as is becoming evident in Germany. For my part, I hold that, where they differ, Christianity is preferable, but where they agree, both are mistaken. The conception which I should substitute as the purpose of education is civilization, a term which, as I mean it, has a definition which is partly individual, partly social. It consists, in the individual, of both intellectual and moral qualities: intellectually, a certain minimum of general knowledge, technical skill in one's own profession, and a habit of forming opinions on evidence; morally, of impartiality, kindliness, and a modicum of self-control. I should add a quality which is neither moral nor intellectual, but perhaps physiological: zest and joy of life. In communities, civilization demands respect for law, justice as between man and man, purposes not involving permanent injury to any section of the human race, and intelligent adaptation of means to ends. If these are to be the purpose of education, it is a question for the science of psychology to consider what can be done towards realizing them, and, in particular, what degree of freedom is likely to prove most effective.

On the question of freedom in education there are at present three main schools of thought, deriving partly from differences as to ends and partly from differences in psychological theory. There are those who say that children should be completely free, however bad they may be; there are those who say they should be completely subject to authority, however good they may be; and there are those who say they should be free, but in spite of freedom they should be always good. This last party is larger than it has any logical right to be; children, like adults, will not all be virtuous if they are all free. The belief that liberty will ensure moral perfection is a relic of Rousseauism, and would not survive a study of animals and babies. Those who hold this belief think that education should have no positive purpose, but should merely offer an environment suitable for spontaneous development. I cannot agree with this school, which seems to me too individualistic, and unduly indifferent to the importance of knowledge. We live in communities which require co-operation, and it would be utopian to expect all the necessary co-operation to result from spontaneous impulse. The existence of a large population on a limited area is only possible owing to science and technique; education must, therefore, hand on the necessary minimum of these. The educators who allow most freedom are men whose success depends upon a degree of benevolence, self-control, and trained intelligence which can hardly be generated where every impulse is left unchecked; their merits, therefore, are not likely to be perpetuated if their methods are undiluted. Education, viewed from a social standpoint, must be something more positive than a mere opportunity for growth. It must, of course, provide this, but it must also provide a mental and moral equipment which children cannot acquire entirely for themselves.

The arguments in favour of a great degree of freedom in education are derived not from man's natural goodness, but from the effects of authority, both on those who suffer it and on those who exercise it. Those who are subject to authority become either submissive or rebellious, and each attitude has its drawbacks.

The submissive lose initiative, both in thought and action; moreover, the anger generated by the feeling of being thwarted tends to find an outlet in bullying those who are weaker. That is why tyrannical institutions are self-perpetuating: what a man has suffered from his father he inflicts upon his son, and the humiliations which he remembers having endured at his public school he passes on to Ònatives" when he becomes an empire-builder. Thus an unduly authoritative education turns the pupils into timid tyrants, incapable of either claiming or tolerating originality in word or deed. The effect upon the educators is even worse: they tend to become sadistic disciplinarians, glad to inspire terror, and content to inspire nothing else. As these men represent knowledge, the pupils acquire a horror of knowledge, which, among the English upper-class, is supposed to be part of human nature, but is really part of the well-grounded hatred of the authoritarian pedagogue.

Rebels, on the other hand,, though they may be necessary, can hardly be just to what exists. Moreover, there are many ways of rebelling, and only a small minority of these are wise. Galileo was a rebel and was wise; believers in the flat-earth theory are equally rebels, but are foolish. There is a great danger in the tendency to suppose that opposition to authority is essentially meritorious and that unconventional opinions are bound to be correct: no useful purpose is served by smashing lamp-posts or maintaining Shakespeare to be no poet. Yet this excessive rebelliousness is often the effect that too much authority has on spirited pupils. And when rebels become educators, they sometimes encourage defiance in their pupils, for whom at the same time they are trying to produce a perfect environment, although these two aims are scarcely compatible.

What is wanted is neither submissiveness nor rebellion, but good nature, and general friendliness both to people and to new ideas. These qualities are due in part to physical causes, to which old-fashioned educators paid too little attention; but they are due still more to freedom from the feeling of baffled impotence which arises when vital impulses are thwarted. If the young are to grow into friendly adults, it is necessary, in most cases, that they should feel their environment friendly. This requires that there should be a certain sympathy with the child's important desires, and not merely an attempt to use him for some abstract end such as the glory of God or the greatness of one's country. And, in teaching, every attempt should be made to cause the pupil to feel that it is worth his while to know what is being taught-at least when this is true. When the pupil co-operates willingly, he learns twice as fast and with half the fatigue. All these are valid reasons for a very great degree of freedom.

It is easy, however, to carry the argument too far. It is not desirable that children, in avoiding the vices of the slave, should acquire those of the aristocrat. Consideration for others, not only in great matters, but also in little everyday things, is an essential element in civilization, without which social life would be intolerable. I am not thinking of mere forms of politeness, such as saying "please" and "thank you": formal manners are most fully developed among barbarians, and diminish with every advance in culture. I am thinking rather of willingness to take a fair share of necessary work, to be obliging in small ways that save trouble on the balance. Sanity itself is a form of politeness and it is not desirable to give a child a sense of omnipotence, or a belief that adults exist only to minister to the pleasures of the young. And those who disapprove of the existence of the idle rich are hardly consistent if they bring up their children without any sense that work is necessary, and without the habits that make continuous application possible.

There is another consideration to which some advocates of freedom attach too little importance. In a community of children which is left without adult interference there is a tyranny of the stronger, which is likely to be far more brutal than most adult tyranny. If two children of two or three years old are left to play together, they will, after a few fights, discover which is bound to be the victor, and the other will then become a slave. Where the number of children is larger, one or two acquire complete mastery, and the others have far less liberty than they would have if the adults interfered to protect the weaker and less pugnacious. Consideration for others does not, with most children, arise spontaneously, but has to be taught, and can hardly be taught except by the exercise of authority. This is perhaps the most important argument against the abdication of the adults.

I do not think that educators have yet solved the problem of combining the desirable forms of freedom with the necessary minimum of moral training. The right solution, it must be admitted, is often made impossible by parents before the child is brought to an enlightened school. just as psychoanalysts, from their clinical experience, conclude that we are all mad, so the authorities in modern schools, from their contact with pupils whose parents have made them unmanageable, are disposed to conclude that all children are "difficult" and all parents utterly foolish. Children who have been driven wild by parental tyranny (which often takes the form of solicitous affection) may require a longer or shorter period of complete liberty before they can view any adult without suspicion. But children who have been sensibly handled at home can bear to be checked in minor ways, so long as they feel that they are being helped in the ways that they themselves regard as important. Adults who like children, and are not reduced to a condition of nervous exhaustion by their company, can achieve a great deal in the way of discipline without ceasing to be regarded with friendly feelings by their pupils.

I think modern educational theorists are inclined to attach too much importance to the negative virtue of not interfering with children, and too little to the positive merit of enjoying their company. If you have the sort of liking for children that many people have for horses or dogs, they will be apt to respond to your suggestions, and to accept prohibitions, perhaps with some good-humoured grumbling, but without resentment. It is no use to have the sort of liking that consists in regarding them as a field for valuable social endeavour, orÑwhat amounts to the same thingÑas an outlet for power-impulses. No child will be grateful for an interest in him that springs from the thought that he will have a vote to be secured for your party or a body to be sacrificed to king and country. The desirable sort of interest is that which consists in spontaneous pleasure in the presence of children, without any ulterior purpose. Teachers who have this quality will seldom need to interfere with children's freedom, but will be able to do so, when necessary, without causing psychological damage.

Unfortunately, it is utterly impossible for over-worked teachers to preserve an instinctive liking for children; they are bound to come to feel towards them as the proverbial confectioner's apprentice does towards macaroons. I do not think that education ought to be anyone's whole profession: it should be undertaken for at most two hours a day by people whose remaining hours are spent away from children. The society of the young is fatiguing, especially when strict discipline is avoided. Fatigue, in the end, produces irritation, which is likely to express itself somehow, whatever theories the harassed teacher may have taught himself or herself to believe. The necessary friendliness cannot be preserved by self-control alone. But where it exists, it should be unnecessary to have rules in advance as to how "naughty" children are to be treated, since impulse is likely to lead to the right decision, and almost any decision will be right if the child feels that you like him. No rules, however wise, are a substitute for affection and tact.