To grasp the shadowy and fantasmal form of a book, to hold it fast, to turn it over and survey it at leisure—that is the effort of a critic of books, and it is perpetually defeated. Nothing, no power, will keep a book steady and motionless before us, so that we may have time to examine its shape and design. As quickly as we read, it melts and shifts in the memory; even at the moment when the last page is turned, a great part of the book, its finer detail, is already vague and doubtful. A little later, after a few days or months, how much is really left of it? A cluster of impressions, some clear points emerging from a mist of uncertainty, this is all we can hope to possess, generally speaking, in the name of a book. The experience of reading it has left something behind, and these relics we call by the book's name; but how can they be considered to give us the material for judging and appraising the book? Nobody would venture to criticize a building, a statue, a picture, with nothing before him but the memory of a single glimpse caught in passing; yet the critic of literature, on the whole, has to found his opinion upon little more. Sometimes it is possible to return to the book and renew the impression; to a few books we may come back again and again, till they do in the end become familiar sights. But of the hundreds and hundreds of books that a critic would wish to range in his memory, in order to scrutinize and compare them reflectively, how many can he expect to bring into a state of reasonable stability? Few indeed, at the best; as for the others, he must be content with the shapeless, incoherent visions that respond when the recollection of them is invoked.

It is scarcely to be wondered at if criticism is not very precise, not very exact in the use of its terms, when it has to work at such a disadvantage. Since we can never speak of a book with our eye on the object, never handle a book—the real book, which is to the volume as the symphony to the score—our phrases find nothing to check them, immediately and unmistakably, while they are formed. Of a novel, for instance, that I seem to know well, that I recall as an old acquaintance, I may confidently begin to express an opinion; but when, having expressed it, I would glance at the book once more, to be satisfied that my judgement fits it, I can only turn to the image, such as it is, that remains in a deceiving memory. The volume lies before me, no doubt, and if it is merely a question of detail, a name or a scene, I can find the page and verify my sentence. But I cannot catch a momentary sight of the book, the book itself; I cannot look up from my writing and sharpen my impression with a straight, unhampered view of the author's work; to glance at a book, though the phrase is so often in our mouths, is in fact an impossibility. The form of a novel—and how often a critic uses that expression too—is something that none of us, perhaps, has ever really contemplated. It is revealed little by little, page by page, and it is withdrawn as fast as it is revealed; as a whole, complete and perfect, it could only exist in a more tenacious memory than most of us have to rely on. Our critical faculty may be admirable; we may be thoroughly capable of judging a book justly, if only we could watch it at ease. But fine taste and keen perception are of no use to us if we cannot retain the image of the book; and the image escapes and evades us like a cloud.

We are so well accustomed to this disability that I may seem to make too much of it. In theory, certainly, the book is never present in the critic's mind, never there in all its completeness; but enough of it, in a commonly good memory, remains to be discussed and criticized—the book as we remember it, the book that survives, is sufficient for practical purposes. Such we assume to be the case, and our criticism is very little troubled by the thought that it is only directed at certain fragments of the book which the author wrote, the rest of it having ceased to exist for us. There is plenty to say of a book, even in this condition; for the hours of our actual exposure to it were full and eventful, and after living for a time with people like Clarissa Harlowe or Anna Karenina or Emma Bovary we have had a lasting experience, though the novels in which they figured may fall away into dimness and uncertainty. These women, with some of the scenes and episodes of their history, remain with us as vividly as though we had known them in life; and we still keep a general impression of their setting and their fortunes, a background more or less undefined, but associated with the thought of them. It all makes a very real and solid possession of a kind, and we readily accept it as the book itself. One does not need to remember the smaller detail of the story to perceive the truth and force of the characters; and if a great deal is forgotten, the most striking aspects of the case will linger in the mind as we look back. Dramatic episodes, fine pieces of description, above all the presence of many interesting and remarkable people—while there is so much that instantly springs to light when the book is mentioned, it seems perverse to say that the book is not before us as we write of it. The real heart and substance of the book, it might even be urged, stands out the more clearly for the obscurity into which the less essential parts of it subside.

And true it is that for criticism of the author's genius, of the power and quality of his imagination, the impressions we are able to save from oblivion are material in plenty. Of Richardson and Tolstoy and Flaubert we can say at once that their command of life, their grasp of character, their knowledge of human affections and manners, had a certain range and strength and depth; we can penetrate their minds and detect the ideas that ruled there. To have lived with their creations is to have lived with them as well; with so many hours of familiar intercourse behind us we have learnt to know them, and it matters little that at any particular moment our vision of their work is bound to be imperfect. The forgotten detail has all contributed to our sense of the genius which built up and elaborated the structure, and that sense abides. Clarissa and Anna and Emma are positive facts, and so are their authors; the criticism of fiction is securely founded upon its object, if by fiction we mean something more, something other, than the novel itself—if we mean its life-like effects, and the imaginative gifts which they imply in the novelist. These we can examine as long and as closely as we choose, for they persist and grow more definite as we cultivate the remembrance of them. And to these, accordingly, we find our criticism always tending; we discuss the writer, we discuss the people in his book, we discuss the kind of life he renders and his success in the rendering. But meanwhile the book, the thing he made, lies imprisoned in the volume, and our glimpse of it was too fleeting, it seems, to leave us with a lasting knowledge of its form. We soon reach the end of so much as we have to say on that subject.

Perhaps we should have more to say of it if we read the book differently in the first place. I scarcely think we could any of us claim that in reading a novel we deliberately watch the book itself, rather than the scenes and figures it suggests, or that we seek to construct an image of the book, page by page, while its form is gradually exposed to us. We are much more inclined to forget, if we can, that the book is an object of art, and to treat it as a piece of the life around us; we fashion for ourselves, we objectify, the elements in it that happen to strike us most keenly, such as an effective scene or a brilliant character. These things take shape in the mind of the reader; they are recreated and set up where the mind's eye can rest on them. They become works of art, no doubt, in their way, but they are not the book which the author offers us. That is a larger and more complex form, one that it is much more difficult to think of as a rounded thing. A novel, as we say, opens a new world to the imagination; and it is pleasant to discover that sometimes, in a few novels, it is a world which "creates an illusion"—so pleasant that we are content to be lost in it. When that happens there is no chance of our finding, perceiving, recreating, the form of the book. So far from losing ourselves in the world of the novel, we must hold it away from us, see it all in detachment, and use the whole of it to make the image we seek, the book itself.

It is difficult to treat a large and stirring piece of fiction in this way. The landscape opens out and surrounds us, and we proceed to create what is in effect a novel within the novel which the author wrote. When, for example, I try to consider closely the remnant that exists in my memory of a book read and admired years ago—of such a book as Clarissa Harlowe—I well understand that in reading it I was unconsciously making a selection of my own, choosing a little of the story here and there, to form a durable image, and that my selection only included such things as I could easily work into shape. The girl herself, first of all—if she, though so much of her story has faded away, is still visibly present, it is because nothing is simpler than to create for oneself the idea of a human being, a figure and a character, from a series of glimpses and anecdotes. Creation of this kind we practise every day; we are continually piecing together our fragmentary evidence about the people around us and moulding their images in thought. It is the way in which we make our world; partially, imperfectly, very much at haphazard, but still perpetually, everybody deals with his experience like an artist. And his talent, such as it may be, for rounding and detaching his experience of a man or a woman, so that the thing stands clear in his thought and takes the light on every side—this can never lie idle, it is exercised every hour of the day.

As soon as he begins to hear of Clarissa, therefore, on the first page of Richardson's book, the shaping, objectifying mind of the reader is at work on familiar material. It is so easy to construct the idea of the exquisite creature, that she seems to step from the pages of her own accord; I, as I read, am aware of nothing but that a new acquaintance is gradually becoming better and better known to me. No conscious effort is needed to make a recognizable woman of her, though in fact I am fitting a multitude of small details together, as I proceed to give her the body and mind that she presently possesses. And so, too, with the lesser people in the book, and with their surroundings; so, too, with the incidents that pass; a succession of moments are visualized, are wrought into form by the reader, though perhaps very few of them are so well made that they will last in memory. If they soon disappear, the fault may be the writer's or the reader's, Richardson's if he failed to describe them adequately, mine if my manner of reading has not been sufficiently creative. In any case the page that has been well read has the best chance of survival; it was soundly fashioned, to start with, out of the material given me by the writer, and at least it will resist the treachery of a poor memory more resolutely than a page that I did not thoroughly recreate.

But still, as I say, the aspects of a book that for the most part we detach and solidify are simply those which cost us no deliberate pains. We bring to the reading of a book certain imaginative faculties which are in use all the day long, faculties that enable us to complete, in our minds, the people and the scenes which the novelist describes—to give them dimensions, to see round them, to make them "real." And these faculties, no doubt, when they are combined with a trained taste, a sense of quality, seem to represent all that is needed for the criticism of fiction. The novel (and in these pages I speak only of the modern novel, the picture of life that we are in a position to understand without the knowledge of a student or a scholar)—the modern novel asks for no other equipment in its readers than this common gift, used as instinctively as the power of breathing, by which we turn the flat impressions of our senses into solid shapes: this gift, and nothing else except that other, certainly much less common, by which we discriminate between the thing that is good of its kind and the thing that is bad. Such, I should think, is very nearly the theory of our criticism in the matter of the art of fiction. A novel is a picture of life, and life is well known to us; let us first of all "realize" it, and then, using our taste, let us judge whether it is true, vivid, convincing—like life, in fact.

The theory does indeed go a little further, we know. A novel is a picture, a portrait, and we do not forget that there is more in a portrait than the "likeness." Form, design, composition, are to be sought in a novel, as in any other work of art; a novel is the better for possessing them. That we must own, if fiction is an art at all; and an art it must be, since a literal transcript of life is plainly impossible. The laws of art, therefore, apply to this object of our scrutiny, this novel, and it is the better, other things being equal, for obeying them. And yet, is it so very much the better? Is it not somehow true that fiction, among the arts, is a peculiar case, unusually exempt from the rules that bind the rest? Does the fact that a novel is well designed, well proportioned, really make a very great difference in its power to please?—and let us answer honestly, for if it does not, then it is pedantry to force these rules upon a novel. In other arts it may be otherwise, and no doubt a lop-sided statue or an ill-composed painting is a plain offence to the eye, however skilfully it may copy life. The same thing is true of a novel, perhaps, if the fault is very bad, very marked; yet it would be hard to say that even so it is necessarily fatal, or that a novel cannot triumphantly live down the worst aberrations of this kind. We know of novels which everybody admits to be badly constructed, but which are so full of life that it does not appear to matter. May we not conclude that form, design, composition, have a rather different bearing upon the art of fiction than any they may have elsewhere?

And, moreover, these expressions, applied to the viewless art of literature, must fit it loosely and insecurely at best—does it not seem so? They are words usurped from other arts, words that suppose a visible and measurable object, painted or carved. For criticizing the craft of fiction we have no other language than that which has been devised for the material arts; and though we may feel that to talk of the colours and values and perspective of a novel is natural and legitimate, yet these are only metaphors, after all, that cannot be closely pressed. A book starts a train of ideas in the head of the reader, ideas which are massed and arranged on some kind of system; but it is only by the help of fanciful analogies that we can treat the mass as a definite object. Such phrases may give hints and suggestions concerning the method of the novelist; the whole affair is too nebulous for more. Even if a critic's memory were infallible, as it can never be, still it would be impossible for him to give a really scientific account of the structure of the simplest book, since in the last resort he cannot lay his finger upon a single one of the effects to which he refers. When two men stand looking at a picture, at least their two lines of vision meet at a point upon the canvas; they may dispute about it, but the picture stands still. And even then they find that criticism has its difficulties, it would appear. The literary critic, with nothing to point to but the mere volume in his hand, must recognize that his wish to be precise, to be definite, to be clear and exact in his statements, is hopelessly vain.

It is all undeniable, no doubt; from every side we make out that the criticism of a book—not the people in the book, not the character of the author, but the book—is impossible. We cannot remember the book, and even if we could, we should still be unable to describe it in literal and unequivocal terms. It cannot be done; and the only thing to be said is that perhaps it can be approached, perhaps the book can be seen, a little more closely in one way than in another. It is a modest claim, and my own attempt to assert it will be still more modest. A few familiar novels, possibly a dozen, by still fewer writers—it will be enough if I can view this small handful with some particularity. And I shall consider them, too, with no idea of criticizing all their aspects, or even more than one. How they are made is the only question I shall ask; and though indeed that is a question which incidentally raises a good many others—questions of the intention of the novelist, his choice of a subject, the manner of his imagination, and so forth—these I shall follow no further than I can help. And as for the few novels that I shall speak of, they will be such as appear to illustrate most plainly the various elements of the craft; one need not range widely to find them, nor does it matter if the selection, from any other point of view, should seem arbitrary. Many great names may be passed over, for it is not always the greatest whose method of work gives the convenient example; on the other hand the best example is always to be found among the great, and it is essential to keep to their company.

But something may first be said of the reading of a novel. The beginning of criticism is to read aright, in other words to get into touch with the book as nearly as may be. It is a forlorn enterprise—that is admitted; but there are degrees of unsuccess.