CONTRIBUTIONS TO RHETORICAL THEORY Edited by Fred Newton Scott, Ph. D. Junior Professor of Rhetoric in the University of Michigan V. THE METAPHOR: A STUDY IN THE PSYCHOLOGY OF RHETORIC BY GERTRUDE BUCK, Ph. D. INSTRUCTOR IN ENGLISH IN VASSAR COLLEGE. THE INLAND PRESS ANN ARBOR MICH. 7 i (olTD C V This pamphlet may be ordered of Sheehan & Co., booksellers, Ann Arbor, Michigan, at the price of fifty cents, net. The preceding issues in the same series nave been as follows: I. Figures of Rhetoric: A Psychological Study, by Gertrude Buck (out of Print); II. Student Slang, by Willard C. Gore; III. Two Problems in Composition-Teaching, by Joseph V. Denney; IV. References on the Teaching of Rhetoric and Composition, by the editor. The price of each of these is fifteen cents, net. INTRODUCTION. The subject of metaphor bristles with problems. Is this figure a natural product or an artificial .'' How does it come to be ? How does it die? How is it related to plain statement .' How are "radical" meta- phors different from "poetical" ? Why does metaphor please the reader? How does it become ' ' mixed" ? These are questions which every serious consideration of the subject must at least attempt to answer. Such solutions as have hitherto been furnished these problems have been rooted in the philosophy of an earlier generation, now discredited. The purpose of this study is to explain metaphor in terms of the con- temporary psychology. In so doing it has, perhaps, been inevitable that a new face should be put upon this figure. From a mechanical structure it has become a biologic organism. It has come to stand as the linguistic representative of a certain stage in the development of thought, and thus an expression perfectly natural and universal, rather than as a literary device, somewhat artificial and wholly unique, obedi- ent to no laws save those empiric ones whose validity extends no farther than to itself. In this conception of metaphor the present study differs from the rhetorical treatises as a class, though building upon their foundations. It also deviates from the practice of the rhetoricians in distinguishing care- fully between the metaphor as viewed from the writer's and from the reader's standpoint. The fact of such a distinction has been often implied, but the essential difference between the activities set up in each case has not been explained and the one activity has often been confused with the other. The doctrines that metaphor is invariably antecedent to plain state- ment, that radical and poetic metaphor differ only in representing differ- ent stages in the development of a perception, that metaphor is pleasurable to the reader because of the harmoniously differentiated activities which it sets up in his mind and body, are doctrines not formulated by the rhetoricians. They grow, however, directly from the fundamental con- ception that metaphor is the expression in language of a certain stage in the development of perception. It has seemed natural to consider, first, the normal metaphor, and afterward some abnormal variations from the type. The term "pathol- ogy" in its application to rhetorical processes has been borrowed from Dr. Fred Newton Scott, Junior Professor of Rhetoric at the University of Michigan, who first used it in a paper entitled "Diseases of English Prose," read before the annual meeting of the Modern Language Asso- ciation, in December, 1896. To Dr. Scott I am also indebted for much stimulus and criticism in the preparation of this thesis; to Dr. John Dewey, now of the University of Chicago, for the fundamental philosophic conceptions embodied in it, to Professor Francis W. Kelsey of the University of Michigan, Professor Milton W. Humphreys of the University of Virginia, and Dr. Mary Gilmore Williams, now professor of Greek at Mt. Holyoke College, for a large number of classical metaphors which, though few of them appear upon these pages, have been of the greatest assistance to my own study of the subject. CONTENTS. Introduction Chapter I. Chapter II. Chapter III. Chapter IV. Chapter V. Genesis: The Radical Metaphor Genesis: The Poetic Metaphor The Evolution into Plain Statement The Aesthetics of Metaphor Pathological Forms of Metaphor Summary of Conclusions - - Appendices — A. Bibliography _ - - - - B. Former Definitions of Metaphor ni I 19 36 46 60 68 70 7S CHAPTER I. GENESIS: THE RADICAL METAPHOR. A comparative study of the existing^ definitions of metaphor, while not in itself without interest, yields to the theory of the subject far less than might fairly be expected. The results of such a study may be sum- marized' in the two general statements that the definitions examined are all essentially alike, and that . all are descriptive rather than genetic, static instead of dynamic. That all existing definitions of metaphor are practically identical, means that Aristotle's conception of this figure has simply repeated itself 4 in each succeeding generation of rhetorical treatises. To him metaphor was essentially a transference of meaning from one word to another over the bridge of analogy.^ Later definitions only phrase somewhat differ- ently the same idea. Metaphor is the use of a word denoting one object to indicate another object resembling the first in some particular; or it is the "comparison" of two objects; or it i&the affirmation of the identity of two objects.' In spite of superficial diversities, it is evident that all these definitions come in the end to the same thing. Each is based upon a recognition of the composite character of the metaphor. Each resolves the figure into at least two parts. There is the thing meant and the thing said; the object of which the writer is speaking and the object to which it is compared. What rhetorician has not seen that when one speaks of oak-trees as "green-robed senators'" two distinct elements are involved, (i) the trees and (2) the senators, to which they are compared ? And in addition to these two constituents a third is often noted, the resemblance which exists between the two objects and justi- fies their comparison. So much, then, can we learn about metaphor from the definitions delivered to us. It is not a simple structure, but duplex *, or even trip- licate. It is essentially a "gedoppelte Ausdruck."* Plain statement says one thing; metaphor says two. Plain statement is single, meta- phor is double. While the straightforward, prosaic method would be to express an idea in tetms of itself, the metaphor expresses it in terms of another idea. Yet, to express one idea in terms of another logically requires some connection between them. This connection is, for meta- phor, that of resemblance. Every metaphor may, accordingly, be analyzed into three parts, the two constituent objects or ideas and the connecting link between them. ' The quotation and classification of previous definitions will be found in Appendix B. ^ See Appendix B. 'Keats, Hyperion, Book I. * "Des Metaphorische," sajrs Hegel, "das in sich selbst diese Zweiheit ist." AesiA. I, 507 * Hegel, Aesth. I, 507. — 2 — Thus far our definitions have laid bare the structure of the meta- phor. They have taken the figure to pieces, and it lies apart. But at this point their service ends. They have dissected the dead body of the metaphor; but they have not told us how the living figure came to be, nor what it in essence is. And this is what we wish to know — not merely the anatomy of the metaphor, but its life. We must learn, not only that there are two major constituents in this figure, but how they came to co-exist there. Did they appear simultaneously .■" And if so, out of what were they born ? Or did one first come into being and draw the other to a position at its side .-' But in this case, whence came the one ? And by what means did it grapple to itself the other ? How do these two die out of the figure so that it ceases to be a metaphor ? What, in short, is the life-history of metaphor ? What process in the mind of the writer lies back of the figure as it stands in type ? What process is induced in the mind of the reader ? These are questions, it is plain, touching the psychology of the writer and of the reader of a metaphor. We can give no adequate answers to them from the merely descriptive definitions of this figure. Let us then put aside for the present these definitions and betake our- selves to such accounts as we have in rhetorical and philological treatises of the genesis and growth of metaphor. These accounts, we shall find, begin with Cicero, who advances two hypotheses as to the origin of metaphor. " Necessity," he says, "was the parent (of metaphorical speech) compelled by the sterility and narrowness of language; but after- wards delight and pleasure made it frequent; for as dress was first adopted for the sake of keeping off the cold, but in process of time began to be made an ornament of the body, so the metaphorical use of words was originally invented on account of their paucity, but became common from the delight which it afforded."* Later, Quintilian' concludes that meaning is transferred from one object to another, either because there is no proper word corresponding to the object to which transfer is made, or because, if there is a proper word, the transferred word is for some reason preferable. Both these explanations, it is evident, agree in explaining the use of metaphorical words as due to one of two causes: (i) The poverty of language, or (2) the desire to ornament speech. Transfer of meaning takes place either because it is necessary, or because it is pleasurable. If we include under the second hypothesis the considerations that a transfer of meaning makes speech more vivacious', more energetic*, or forcible', clearer', more *Cic. L>e Oratore^ III, 38, tr. by J. S. Watson, Bohn Ed. Bauer repeals lais explanauon almost verbatim: " Die Metaphorische Redeweise ist vielumfassend und von der Notwendig- keit erzeugt im Drange der .Armut und Verlegenheit, nachmals aber gesucht worden um ihrer Anmut und Lieblichkeit willen." Das Bild in der Sprache, II, p. 27. 'See definition. Appendix B. •CampBell, Philos. of Rhet., Bk. Ill, Ch. I, %YL. ♦Whately, Eh. of Rhet., Pt. Ill, Ch. II, §3. »Wendell, Eng. Comp., Ch. VII. •Oenung, Pract. Els. of Rhet., Ch. Ill, p. 90. _3 — vivid and picturesque', more economicaP, more stimulating', we may say that all explanations of the origin of metaphor are covered by the dilemma of Cicero, j Philologists and philosophers have, for the most part, empha- sized the hypothesis of the poverty of language, rhetoricians that of the desire for ornamenting speech. The separation between the two has thus grown with time, until it has come to be pretty generally accepted as dividing the genus metaphor into two distinct species, one of which comes into being as the result of the scarcity of proper words, the other as the expression of a tendency to ornament speech. This classification of metaphors goes back, by implication at least, to Quintilian, who shows that we use this figure sometimes perforce, but sometimes "with a view to significance or force of expression."* Hugh Blair states the rationale of this division somewhat decisively in the following passage: "But although the barrenness of language and the want of words be doubtless one cause of the invention of tropes, yet it is not the only nor' perhaps even the principal source of this form of speech. Tropes have arisen more frequently and spread themselves wider from the influence which imagination possesses over language."* But by far the most definitive separation of metaphors into two classes^ on the basis of their origin has been made by Max Miiller. "I call it radical metaphor, " he says, ' ' when a root which means to shine is applied to form the names not only of the fire or the sun, but of the spring of the year, the morning light, the brightness of thought, or the joyousout burst of hymns of praise From this we must distinguish poetical metaphor, namely, when a noun or verb ready made and assigned to one definite object or action is transferred poetically to another object or action."® 'Carpenter, Exs. in Rhet.^ Adv. Course, 2d Ed., Ch. XII, p. 199. ^Spencer, Philosophy of Style, Scott's Ed., p. 25. »A. S. Hill, Prins. of Rhet., Ed. 1896, pp. I18-II9. *"This change (transfer of meaning) we make either because it is necessary, adds to significance; or, as I said, because it is more ornamental From necessity the rustics speak of the gemma, ' bud ' of the vines (for how else could they express themselves ? ) and say that the corn thirsts and that the crops suffer. From necessity we say that a man is rough or hard, because there is no proper term for us to give to these dispositions of the mind. But we say that a man is inflamed with anger, burning with desire, and has fallen into error, with a view to significance or force of expression." Institutes of Oratory, Watson's Tr., Bk. VIII, Ch. 6, §§ 6 and 7. ^Lectures on Rhetoric and Belles Lettres, Lect. XII, p. 151. ®Miiller, Science of Language, Second Series, Lect. VIlI, p. 371. The same statement is made in varying phrases throughout Muller's writings: " Metaphor, therefore, ought no longer to be understood as simply the premeditated act of a poet, as a conscious transference of a word from one object to another. This is modern, fanciful, individual metaphor, while the old meta- phor was much more frequently a matter of necessity-''— v^<^^^«. 46:621-2. "This process (of radical metaphor) is different and ought to be distinguished from another, namely, the transfer- ence of ready-made words from one well-known object to another equally well known object, as when the poets call the rays of the sun arrows," etc. — Fortn. 46:617. In like manner Bauer and Von Raumer divide the genus metaphor — " Man muss," so says Bauer. " nun zwischen zwei Arten der Metapher unterscheiden, welche ich die radicale und die poetische nennen will." — Das Bild in der Sprache, I, p. 25. Then follows Miilier's distinction. The following is Von Raumer's division: "I. Metaphern, welche durch den Mangel eines treffenden Ausdrucks hervorgerufen wurde. II. Metaphern welche dem Schmucke der Rede dienen." — Die Metapher bet Lucrez^ p. i. See also Hegel, Aesth. I, pp. 505-6. — 4 — This classification of metaphors has, apparently, become pretty firmly entrenched. One species, that which has arisen from the neces- sity of borrowing words, has been handed over to the philologists and the philosophers. The other, which has grown out of the desire to ornament speech, is, by common consent, remanded to the rhetoricians. We shall, perhaps, do well to allow this division for the present, and betake our- selves in turn to the students of language and of rhetoric for light upon the origin of metaphor. From the philologists we shall obtain no explicit formulation of any theory for the genesis of the figure save that of the poverty of language, and hence the necessity of borrowing the designations of material objects to represent spiritual conceptions; or, as some prefer to state the same hypothesis, the extension of names properly applied to material objects until they should cover correspondent spiritual conceptions as well. Max Muller commonly stands as the philological sponsor of this idea, though we must turn to Locke for its ultimate source. The classic passage upon which all succeeding writers have drawn may again be quoted here: "It may also lead us a little towards the original of all our notions and knowledge, if we remark, how great a dependence our words have on common sensible ideas; and how those, which are made use of, to stand for actions and notions, quite removed from sense, have their rise from thence and from obvious, sensible ideas, are transferred to more abtruse significations, and made to stand for ideas that come not under the cognizance of our senses; e. g. to imagine, apprehend, comprehend, adhere, conceive, instil, disgust, disturbance, tranquility, etc. , are all words ta,ken from the operations of sensible things and applied to cer- tain modes of thinking. Spirit in its primary signification is breath; angel, a messenger; and I doubt not that if we could trace them to their sources, we should find, in all languages, the names which stand for things, that fall not under our senses, to have had their first rise from sensible ideas. By which we may give some kind of guess what kind of notions they were and whence derived, which filled their minds, who were the first beginners of language; and how nature, even in the nam- ing of things, unawares, suggested to men the originals and principals of all their knowledge; whilst, to give names that might make known to others any operations they felt in themselves, or any other ideas, that came not under their senses they were forced to borrow words from ordi- nary, known ideas of sensation, by that means to make others the more easily to conceive those operations, they experienced in themselves, which made no outward sensible appearance.'" The theory is rendered into Blair's vernacular without substantial change: "Men naturally sought to abridge this labor of multiplying words in infinitum; and, in order to lay less burden on their memories, made one word, which they had already appropriated to a certain idea or object, stand also for some other idea or object, between which and the primarv one they found, or fancied, pome relation, The i Locke, Essay on the Human Understanding. Bk. Ill, Ch. I, § 5. — 5 — operations of the mind, and affections, in particular, are, in most lan- guages, described by words taken from sensible objects. The reason is plain. The names of sensible objects were, in all languages, the words most early introduced; and were, by degrees, extended to those mental objects of which men had most obscure conceptions, and to which they found it more difficult to assign distinct names. They borrowed, there- fore, the name of some sensible idea where their imagination found some affinity. "> In the citations so far made we have simply the statement that the names of material objects are borrowed to designate spiritual concep- tions, this transaction being justified by the existence of a certain resem- blance or analogy between the two. We are not assured that the bor- rower saw this analogy clearly, or that his seeing it incited him to the borrowing. But Muller and Whitney set at rest all our doubts upon this point. "The mental process," says Muller, "which gave to the root mar the meaning of to propitiate was no other than this, that men perceived some analogy between the smooth surface produced by rub- bing and polishing and the smooth expression of countenance, the smoothness of voice, and the calmness of looks produced even in an enemy by kind and gentle words. Thus, when we speak of a crane we apply the name of a bird to an engine. People were struck with some kind of similarity between the long-legged bird picking up his food with his long beak and their rude engines for lifting weights."* And in like strain, Whitney: "Abstract words have," he says, "been won through the transfer to an ideal use of words and phrases which had before desig- nated something physical and sensible. And the transfer was made in the usage of individuals and communities who saw a resemblance or '^Lectures on Rhetoric and Belles Lettres, Lect. XIV. Compare the following: "All expressions of mental phenomena are borrowed from analogous material experiences." — Hoef- ding, Outlines of Psychology, p. 133, Lowndes tr. "Because language is developed under the influence of attention directed to the external world, we find that expressions for mental phenomena were originally taken from the material world. The inner world of mind is denoted by symbols borrowed from the outer world of space." — Hoeffding, Outlines 0/ Psychology', p. 2. '' Every word which is used to express a moral or intellectual fact, if traced to its root is found to be borrowed from some material appearance. Right means straight, wrong means twisted. Spirit primarily means wind; transgression, the crossing of a line; supercilious the raising of the eyebrow Thought and emotion are words borrowed from sensible things, and now appropriated to spiritual nature. "^Emerson, Nature, Ch. IV (Language). " Because of this radical correspondence between visible things and human thoughts, savages, who have only what is necessary, converse in figures. As we go back in history, language becomes more pic- turesque, until its infancy, when it is all poetry, or all spiritual facts are represented by natural symbols." — Emerson, iVa/wr^', Ch. IV (Language). " Man is at first by the action of all his faculties, carried out of himself and toward the external world; the phenomena of the external world strike him first, and hence these phenomena receive their first names. The first signs are borrowed from sensible objects, and they are tinged to a certain extent by their colors. When man afterwards turns back upon himself, and lays hold more or less distinctly of the intellectual phenomena which he has always, though somewhat vaguely perceived, if then, he wants to give expression to the new phenomena of mind and soul, analogy leads him to connect the signs he seeks with those he already possesses; for analogy is the law of all growing or devel- oped language. Hence the metaphors to which our analysis traces back most of the signs and names of the most abstract moral ideas." — Victor Cousin, quoted by Miiller, Science of Lan- guage, Second Series, p. 367. 2 Science of Language, Second Series, p. 369. analogy between the physical act and the mental and who were ingen- ious enough to make an application of material already familiar to new and needed uses."' "Every figurative transfer which ever made a suc- cessful designation for some non-sensible act or relation, before undesig- nated, rested upon a previous perception of analogy between the one thing and the other; no one said apprehend of an idea until he had felt the resemblance between the reaching out of the bodily organs after a physical object he wished to handle and the striving of the mental powers toward a like end." ^ Also Barrett Wendell: " Tendo means to stretch; ad means to or toward; attention really means a stretching out toward. In some remote past it was a metaphor used by some old speaker of Latin who perceived that the process of mind by which we attend to anything is very like the physical process by which we stretch out our hand to grasp a tangible object The original maker of the word (apprehend) saw the likeness between the mental process of what we now call apprehension and the physical process of grasping. He called the one by the name that really denoted the other." ' Now, let us consider for a moment what is implied in such a theory as this. In the mere statement its inherent fallacies start out. Miiller, whose doctrine we may instance as the completest formulation of the position we are discussing, holds substantially that there was a time in human history when only names for the material facts of life existed; what we now call abstract, intellectual, or spiritual facts not being known; that when these intellectual or spiritual conceptions did appear in the consciousness of the individual, each by its resemblance to a cer- tain well-known material entity, suggested that entity to the mind of the speaker, who thereupon borrowed the name of the material as an appropriate expression of the spiritual fact. The transaction was purely commercial. The demand was made for a label to affix upon a newly manufactured conception; and since none existed which had been designed specifically for this purpose, there was nothing for it but to use another label, such as properly appertained to a conception resembling in some ways the nameless species. Thus when spirit, a new concep- tion, came into the field of human consciousness, it stood there naked and shivering, until it was clothed upon by the word breath, which represented an analogous material object.* Now, if only we could be certain of two things, we should not hesitate to adopt this theory; the two things being, (i) that where the name of a material was first transferred to a spiritual fact, the one con- ception was so sharply distinguished from the other in the mind of the speaker as to make them in realitv two separate ideas; and (2) that the ^ Oriental and Linguistic Studies, "Schleicher and the Physical Theory of Language," p. 304- * Life and Growth of Language, Ch. VIII, pp. 137-8. '^English Composition, Ch. VII, p. 248. * Max Miiller, Science of Language, Second Series, p. 369. W. D. Whitney, Life and Growth of Language^ Ch. VIII, p. 137. UNI V i:..; .::i ' i ); resemblance between them was so clear to the borrower that it could serve as an incitement to unite the two under one name, or as a justifi- cation for the project of using the name of the one for the other. Are these points, however, unquestionable? May we safely assume that the spiritual fact, before it is named, has such a separate, clear-cut existence in the mind which assigns to it the name of the correspondent material fact as is implied by the borrowing theory? Does a new con- ception spring into language thus completely defined and distinct from all other conceptions?' The answer of modern psychology to this question is distinctly negative. The definition of an idea is a long and gradual process, much like that by which a formless sprawl of protoplasm becomes a firmly articulate creature; or like that by which a dark, irregular blur against the landscape grows little by little, as one approaches it, into the sharp outline of a clump of trees, with all their minute ramifications and inter- lockings. The individual consciousness, both as to the whole and as to its parts, grows by successive differentiations into greater and yet greater distinctness of outline and richness of detail. According to the accepted theories of mental development, the first state of the individual con- sciousness is like that of the primeval earth, without form and void. Gradually, however, out of the wide-weltering chaos, the vaguest and largest physical sensations roughly define themselves, those of comfort as against discomfort; hunger as opposed to satisfaction; light in con- trast to darkness; oneself versus others. Each of these large sensations can then differentiate itself still further; — that of light, for instance, into the different colors, these into tints or shades, and these again into half- tones. This process of differentiation is the task of the entire mental life. Says James: "The object which the numerous inpouring currents of the baby bring to his consciousness is one big, blooming, buzzing confusion. That confusion is the baby's universe; and the universe of all of us is still to a great extent such a confusion, potentially resolvable and demanding to be resolved, but not yet actually resolved, into parts. "^ The task of all succeeding experiences is to separate out from this chaos of sensation its particular elements. Continually the vague discriminations grow finer and sharper. The homogeneous sense of com- fort or discomfort finally becomes the epicure's keen pleasure in the ' Miiller, himself, in later writings, repudiates this idea, and asserts unequivocally: "To imagine in the earliest periods of language a real transference of name from a known thing to an unknown, would be contrary to one of the leading principles of the Sciences of Thought and Language, namely, that nothing can be to us without a name. The act of clothing naked con- cepts with old garments is an act of charity which we never perform. What really happens is that names vary in intension As the conceptions lose their full intension their names become larger, i. e., become applicable to new germinal concepts which are but waiting for a name to spring into life." — Fortn. 46:632. It may here be noted that this apparently new hypothesis of Miiller is really but a slight variation upon the extension-of-names theory. (See infra.) It differs slightly from the customary form of this theory, however, in its assertion that words become less specialized by use. If this were true, which unfortunately it is not, our problem would be far different. The psychological and philological considerations touched upon in the succeeding pages will be sufficient disproof of Miiller's astonishing dictum. * Psychology, Briefer Course, Ch. II., pp. 15-16. — 8 — complex flavors of a French soup, the musician's appreciation of each separate motif in a symphony, or the artist's sharp discrimination of a hundred different shades of green in the spring landscape. The mental life, like the physical, organizes itself from the whole to its parts, from simplicity to complexity, from homogeneousness to differentiation. Transferring this theory of the progress of intellectual experience to the philological field, we should expect to find the earliest speech-forms expressive not of clear-cut and well defined ideas, but of somewhat large and vague, complex but unanalyzed sensations. Let us, however, inquire of the philologists whether this expectation is answered by the fact. There are commonly held to be two opinions among students of language as to the psychological nature of the earliest forms of speech. An early school, of which we may take Miiller as the type, is said to affirm that root-words have always a general or an abstract significa- tion.' A later faction, among whom may be instanced Sayce and Bleek, are held to oppose this view by regarding the most primitive expressions as essentially concrete and particular in meaning.'"' ' " Most roots that have yet been discovered had originally a material meaning, and a meaning so general and comprehensive that they could easily be applied to many special objects." — Miiller, Science of Language, p. 370. "The Science of Language, by inquiring into the origin of general terms, has established two facts of the highest importance, namely, first, that all terms were originally general; and secondly, that they could not be anything but general." — Miiller, Science of Thotighl, p. 456. "If the science of language has proved anything, it has proved that every term which is applied to a particular idea or object (unless it be a proper name) is already a general term." —Miiller, Science of Thought, p. 549. " In order that any such name [as canis to dog] could be given, it was requisite that general ideas, such as roving, following, stealing, running, resting, should previously have been formed in the mind, and should have received expression in language. These general ideas are expressed by roots." — Miiller, Science of Language^ Second Series, Ch. VII, p. 329. For an equivalent statement see W. D. Whitney, Life and Growth of Language, Ch. IV, pp. 29S-9. ^ Sayce declares that the history of language development shows everywhere a progress from words representing particular to those corresponding to general ideas. He cites in support of this dictum the facts so universally noted in the vocabulary of modern savages, that "the individual objects of sense have names enough, while general terms are very rare. Thus the Mohicans have words for cutting various objects, but none to signify cutting simply, and the Society Islanders can talk of a dog's tail, a sheep's tail, or a man's tail, but not of tail itself. The Tasmanians were so utterly deficient in the power of forming abstract ideas, that they were obliged to say Mike the moon,' or some other round object, when they wanted to express the conception of roundness." — Prins. of Compar. Phihl., 2nd Ed., Ch. VI, pp. 221-2. The following statement is parallel: The aborigines of Tasmania could not "express abstract qualities, such as hard, soft, warm, cold, long, short, round, etc.: for 'hard,' they would say 'like a stone;' for 'tall,' they would say 'long legs,' etc.; for 'round,' they said 'like a ball,' 'like the moon,' and so on." — Milligan, Vocabulary of the Dialects of Some of the Aboriginal Tribes of Tasmania, p. 34. Quoted by Miiller, Chips from a German Workshop, Vol. V, p. 76 (On the Philosophy of Mythology). Bleek reaches the same conclusion from psychological considerations. "The perception of which one is conscious on hearing a sound, was designated by a word which arose from the imitation of the sound. This perception was not at all of an abstract or general character, but an altogether concrete and individual one. For instance, had a word been formed from imitation of the note of the cuckoo, its concept could not possibly have been limited to that of the bird, or to that of crying, or to any property of the animal or its utterance, etc., etc., but the whole situation, in so far as it came into the consciousness, was indicated by the word. The frequent hearing of the same sound was of itself enough to bring the salient points of the situation into consciousness, but the signification of the word still com- prehended the most heterogeneous elements, whereof one was made more prominent in one con- pection, another in another." — The Origin of Language^ p. 67. — 9 — It is quite possible, however, that the breach between these two Tepresentative points of view may be not so wide as at first it seems. Muller's position, as commonly understood, is quite self-contradictory. On the one hand he regards root meanings as abstract conceptions due to a combination of concrete impressions, and again he declares that "All roots, i. e., all the material elements of language, are expressive of sensuous impressions, and of sensuous impressions only."' He urges also ^ that most roots ' ' had originally a material meaning. " Plainly a "sensuous impression" and a "material meaning" are not synonymous with a "general " in the sense of an abstract idea.' It is not probable that Muller thought the matter out into this anti- thesis; if he had done so, he would undoubtedly have defined somewhat more narrowly his use of the word "general" so as to avoid the appar- ent contradiction; bnt it is quite possible, by reference to the situations in which he constantly employs the term, to understand it not in the philosophic sense of abstract or conceptual, but in the looser meaning of vague, largely inclusive, little defined. For instance, the citation made on a foregoing page,* in which it is stated that most roots had originally a meaning "so general and comprehensive that they could readily be applied to man}- special objects," seems to indicate such an understand- ing of the word.* ' Science of Language, Second Series, p. 372. 2 See Note i, p. 8. ^Romanes puts this inconsistency sharply in the following passage: "He (Max Miiller) has failed to distinguish between ideas as 'general,' and what we have called 'generic;' or between an idea that is general because it is born of an intenrional synthesis of the results of a previous analysis, and an idea that is generalized, because not yet differentiated by any inten- tional analysis, and therefore representing simply an absence of conceptual thought. My child, on first beginning to speak, had a generalized idea of similarity between all kinds of bright, shining objects, and therefore called them all by the one denotative name of 'star.' The astronomer has a general idea answering to his denominative name of ' star, ' but this has been arrived at after a prolonged course of mental evolution wherein conceptual analysis has been engaged in conceptional classification in many and various directions; it therefore represents the psychological antithesis of the generalized idea, which was due to the merely sensuous associa- tions of preconceptual thought." — Metttal Evolution in Man, p. 336. In pronounced anti- thesis to Miiller, Romanes sets the views of Bleek on the subject: "As a result of his (Bleek's) prolonged and first-hand study of the subject, he is strongly of opinion that aboriginal words were expressions ' not at all of an abstract or general character, but exclusively concrete or indi- vidual.' By this he means that primitive ideas were what I have called generic, for he says that had a word been formed from imitation of the sound of a cuckoo, for instance, it could not pos- sibly have had the meaning limited to the name of that bird; but would have been extended so as to embrace 'the whole situation so far as it came within the consciousness of the speaker.' That is to say, it would have become a generic name for the whole recept of bird, cry, flying, etc., etc., just as to our own children the word ba-a=sheep, bleating, grazing, etc. Now, this process of comprising under one denotative term the hitherto undifferentiated perceptions of ' a whole situation, so far as it comes within the consciousness of the speaker,' is the very opposite of the process whereby a denominative term is brought to unify, by an act of 'generalization,' the previously well-differentiated concepts between which some analogy is afterward discov- ered." — Mental Evolution in Man^ p. 337. *P. 8. * Compare the similar use by Leibnitz: " Les termes g^n^raux ne servent pas seulement a la perfection des langues, mais m^me ils sont necessaires pour leur constitution essentielle. Car si par les choses particuliires on entend les sndividuelles, il serait impossible de parler, s'il n"y avait que de noms /rij/rtfj et point d'appellatifs, fcsX. a dire, s'il n'y avait des mots que pour les individus, puis qu'a tout moment il en revient de nouveaux lorsqu'il s'agit des individus, des accidens et particulierement des actions, qui sont ce qu'on designe le plus: mais si par les — 10 — But this is also essentially what Bleek means when he says ' that the root-word indicated "the whole situation" presented to the speaker at the particular moment, and thus represented a perception ' ' altogether concrete and individual." "The whole situation" is as yet in the mind of the primitive man a vague and formless sensation,^ As Sayce himself puts it: "Language is the expression of thought; and the first ideas were as much undifferentiated embryos as the jelly fish on the shore or the beehive life of primeval man. There was no unity in them; idea had not been subordinated to idea; but each was the mere individual impression of the moment, with all the vagueness and complexity of a sensation."^ choses particulieres on entend les plus basses especes (species infimas) outre qu'il est difficile bien souvent de les determiner, il est manifeste que ce sent deja des universaux, fondes sur la similitude plus ou moins etendue, selon qu'on parle des genres ou des especes, il est naturel de marquer toute sorte de similitudes ou convenances et par consequent d'eraployer des termes gen- eraux, etant moins charges par rapport aux idees ou essences, qu'ils renferment, quoiqu'ils soient plus comprehensifs, par rapport aux individus,a qui ils conviennent, etaient bien souvent les plus aises a former, et sont les plus utiles. Aussi voyez vous que les enfans et ceux qui ne savent que peu la langue, qu'fls veulent parler, ou la matiere, dont ils parlent, se servent des termes generaux comme chose, plante, animal, au lieu d'employer les termes propres qui leur manquent. Et il est sur que tous les noms propres ou individuels ont ete originairement appellatif ou giniraux.^' — Opera Philosophica^ Erdman's ed. p. 297. Sayce very cleverly reconciles the self-contradiction of Miiller as follows: "There is, however, a truth in the pre- vailing theory (that original roots were abstract in meaning) The sentence comes before the word, the indefinite before th^ definite; and the root period, as we have seen, is char- acterized by the want of differentiation. The Aryan root, consequently, while primarily denot- ing an individual object, would have done so in a very different way from that in which we should denote the same. The individual can only be properly understood in relation to the general: when, therefore, the idea of the general has not yet been arrived at, the idea of the particular is at once vague and sensuous. The word which denotes it is merely a mark, nothing more; just as much as a proper name, and with no more subjective reference than the proper name has. So long as the object can be pointed out sensibly, the meaning and the reference of the word is unmistakable. We know exactly, for instance, who a particular John or Henry are when they are indicated by the finger; but when the object is not present, the signification and content of the word is wholly vague and uncertain. The jud'^ment which is summed up in it is not determined by immediate reference to such and such a thing; we can not think this is a tree; and accordingly each person forms his own judgment and attaches a different interpreta- tion to the vocable. The term is not yet defined by its external object and language has not yet arrived at the explication of its words by other means. In this way the Aryan roots might easily have come to have those vague general significations which are ascribed to them, although they properly represented individual objects and actions." — Prins. of Compar. Philol.^ 2d Ed., Ch. VI, pp. 225-6. iSee p. 8. 2 It seems more than probable that psychologically considered, this first unformulated impres- sion of a situation may have been in final analysis an embryo act, its most primitive species, per- haps, being simply that of eating or of turning from food, of facing an enemy or running from him. Such an incipient activity must undoubtedly precede the intellectual translation of the situation in terms of explicit perception. Suggestive of this hypothesis are the following state- ments: '* We may be tolerably certain that the things capable of satisfying hunger form a per- fectly distinct class in the mind of any of the more intelligent animals; quite as much as if they were able to use or understand the word food." — ^J. S. Mill, Exain. of Hamilton' s Philos., p. 403. (Quoted by Romanes, Mental Evolution in Man, p. 42.) ''The tendency to apply the same term to a large number of objects, ('ball' to ball, orange, moon, lamp-globe, etc.) can be understood, I think, only if we keep in mind the extent to which the formal noun ' ball ' has really an active sense. ' Ball ' is ' to throw' just as much as it is the round thing. I do not believe that the child either confuses the moon with his ball, or abstracts the roundness of it; the roundness suggests to him something he has thrown, so that the moon is something to throw — if he could only get hold of it." — John Dewey, The Psychology of Infant Language^ Psych. Rev. Vol. I, p. 63. " Wenn zuerst eine Wurzel ausgesprochen wurde, sie einen Vorgang in der Seele, einen Lebensakt dar- stellte." — Gerber, Die Sprache als Kunst, I, p. 229. * Prins. of Compar. Philol., 2d Ed., Ch. VI, pp. 243-4. — 11 — Both the philosophers, then, and the philologists have helped us toward the same conclusion, namely, that the first impressions in lang- uage were likely to be those of large, chaotic, little-differentiated percep- tions. But let us see how this a priori judgment is corroborated by the records which we have of the actual speech of children during their early years. M. Taine has given us some interesting data upon this point.* Of his young daughter at the age of twelve months he writes: Her grand- mother * ' often showed her a painted copy of a picture, by Luini, of the infant Jesus naked, saying at the same time, 'There's bebe.' A week ago, in another room where she was asked ' Where's bebe T meaning herself, she turned at once to the pictures and engravings that happened to be there. Bebe has, then, a general signification for her, namely, whatever she thinks is common to all pictures and engravings of figures and landscapes, that is to say, if I am not mistaken, something varie- gated in a shining frayne. In fact it is clear that the objects painted or drawn in the frame are as Greek to her; on the other hand the bright square inclosing any representation must have struck her. This is her first general word."^ "She was in the habit of seeing a little black dog belonging to the house, which often barks, and to which she first learned to apply the word oua-oua. Very quickly and with very little help she applied it ta dogs of all shapes and kinds that she saw in the streets and then, what is still more remarkable, to the bronze dogs near the staircase. Better still, the day before yesterday, when she saw a goat a month old that bleated, she said 'oua-oua,' calling it by the name of .the dog, which is most like it in form and not by that of the horse which is too big, or of the cat, which has quite a different gait."^ ' ' Cola (chocolate) is one of the first sweetmeats that was given her and it is the one she likes best Of herself .... she has extended the meaning of the word and applies it now to anything sweet; she says cola when sugar, tart, a grape, a peach, or a fig is given to her."* ' ' The above mentioned little boy of twenty months used the word teterre (pomme de terre) to designate potatoes, meat, beans, almost everything good to eat, except milk which he called lolo. Perhaps to him teterre meant everything solid or half solid that is good to eat."* 'Article in Revue Philosophique, No. i, Jan. 1876, translated in Mind, Vol. 11, pp. 252-259. Paul's general declaration is most interesting in this connection: A child " will more commonly err on the side of width than of narrowness of application, and the more so according as his stock of words is more limited. A child will include a sofa under the name of a chair, an um- brella under the name of a stick; and this repeatedly." — Frins. of Hist, of Lang. ^t^. 79-80. And Preyer's statement embodies precisely the thesis of this chapter: "The two com- prehensive, too indefinite concept atta (away, gone,) has broken up into more limited and more definite ones. It has become, as it were, differentiated, as in the embryo the separate tissues are differentiated out of the previously apparently homogeneous tissues." — The Mind of tke Child^ tr. by H. W. Brown, Pt. II, ch. XVIII, p. 123. * Mind, II, p. 254. ^ Mind, II, p. 255. *Mind, II, p. 256. '^ Mind, n, footnote p. 256. — 12 — " Yesterday means to her in the past, and tomorrow in the future, neither of these words denoting to her mind a precise day in relation to today, either preceding or following it."^ It will be noted that while Taine here uses the terms "extension of meaning" and "general words," his illustrations warrant neither phrase. The first instance shows simply a vague perception of the particular picture of Luini, an association of the name beb6 with this perception and a confusion, due to the vagueness of the first perception, between this particular picture and all others which presented to the child prac- tically the same sensation. The word bebe represented no general, in the sense of abstract, idea; but simply a complex perception as yet so cloudy and so undifferentiate that any similar situation could readily be mistaken for it. Upon this point Paul remarks: "For the simple and unreasoning mind of childhood, it is sufficient if the contents of the idea presented to it agree, to warrant an identifying process on his part, whether real identity exists or not. And further, such agreement needs to be partial only — indeed, under some circumstances, merely trivial — to cause this identification; that is, as long as the mental impression is still vague and confused. Thus it is that on the very commencement of the process of acquiring language, the custom grows of defining by the same word, not merely a single, but several, not merely objects which actually resemble each other, but such as bear even a remote resemblance."'' In regard to the phrase "extension of meaning;" the term cola was not stretched to cover other edibles which were perceived to resemble chocolate in sweetness, but the one sweetmeat was to the child's hazy sense, the other, as the picture of the Christ-child was any other simi- larly framed. Between the two no distinction of detail had yet arisen. The rise of a certain distinction is, however, to be noted in the case of the boy oi twenty months. To him the difference between solid things to eat and liquid things to eat had become so apparent as to pre- vent the identification of the sensation arising from the one with that occasioned by the other. Being recognized as different sensations, the two of necessity received different names, teterre for the solid, lolo for the liquid, food. For the same reason the girl could not call the young goat by the name for horse. The detail of size had emerged from the complex perception and stood out so clearly as to make quite impossible a confusion between the sensations horse and kid. Nor could she give to the kid the name of the cat, whose gait (so M. Taine) at once dis- tinguished the complex sensation of its appearance from that of the young goat.' Richard Grant White makes a typical misinterpretation of this vague use of words by the child when he remarks, of the application of * Mind, II, p. 259. ^ Prins. of Hist, of Lang., Ch. IV, p. 79. 'For other instances of the large, undifferentiate use of words on the part of children, see Preyer, The Mind of the Child, tr. by H. W. Brown, Pt. II, Ch. XVIII; K. C. Moore, Mental Development of a Child, Pt. IV. Sec. II; Barnard Perez, The First Three Years of Childhood, tr. by A. M. Christie, Ch. Xl; Sully, Studies of Childhood, Ch. V. — 13 — the term "pussy-cat" to all fur and velvet, "The child knows as well as its mother, that her muff or her gown has not four legs, claws, whisk- ers and a tail; and it has no purpose of concealing that knowledge. But its poverty of language enables it to speak of the muff and the velvet gown only by a name which expresses to the child the quality which the muff, the gown, and the animal have in common."* Now here lies the fallacy. The child does not know " as well as its mother " that the muff and the kitten differ in that the one has head, legs and claws, while the other has not. Out of its hazy perception of a soft furry something these details have not yet developed. They do not at the moment form any part of the child's consciousness. Its perception is of a soft furry something, not further defined, to which the generic term pussy-cat perfectly applies. We are here brought back to our point of departure, the borrowing theory of the genesis of metaphor. We have seen that in the early stages of thought-developmentj words stand not for some unique clearly- defined entity in the world, but for a somewhat inchoate perception of a whole situation, the details of which have not yet disentangled them- selves so as to be projected sharply against the consciousness. As a result of this vagueness of perception, situations in detail very different are seen as identical and the same name applied to each. But this pro- cess is essentially other than that asserted by Locke and his disciples. The two objects of thought do not stand side by side in the mind of the primitive man until he chances to note a resemblance between them, and is thereby warranted in applying the name of the one to the other. The two are one so far as his perception can testify. They have not yet been separated out of the original sensation. Gummere states the case precisely: "A child, even now, does not call a bird's nest a house on the basis of observed relations between a nest and a house; the nest is a house. There is no like about it until the child (i) increases his vocab- ulary with the word ' nest ' and its meaning, and (2) brings the new word into relations with the old word ('house'). The imagination of primitive man was not analytic. He did not watch some ship ride the waves and muse: ' How like yon craft is to a fiery steed ! I liken it to a fiery steed — in fact I shall save time by calling it a fiery steed.' His restless eye, subject to no fine tutorings of reason, saw an actual horse bound over the ' foaming fields ' (not in their turn based on any expanse of water = expanse of land)"" It is only after the nest has been called a house, the velvet a pussy- cat, that the one object can be discriminated from the other, and the real relationship of the two becomes evident as such. The theory that the material significance of a word is, in the early stages of language, stretched to cover an intellectual or spiritual concep- tion related to it by some resemblance, is but a restatement of the bor- rowing hypothesis. This form of the explanation, however, is more 1 Words and their Uses, 25th Ed., Ch. V, p. 82. * Metaphor and Poetry, Mod. Lang. Notes ^ I, 83. ^ OF TffE y UNIVFRRTTV — 14 — inclined to give due value to the looseness of the name which has already been applied to the material object. But it, too, fails in assuming an original separation between the material and the spiritual fact, and the speaker's explicit perception of an analogy between them, which shall justify his pulling over the one-half the other's cloak. The same argu- ments, then, by which the borrowing hypothesis has been condemned, destroy also the validity of this extension-of-meaning theory. Let us now go back to some of the radical metaphors whose origin has been accounted for on the borrowing theory and see how our own hypothesis would explain their existence. Instead of asserting that the root mar came to have the meaning of to propitiate because ' ' men pre- ceived some analogy between the smooth surface produced by rubbing and polishing and the smooth expression of countenance, the smoothness of voice, and the calmness of looks produced even in an enemy by kind and gentle words," * we must deny to the mental process involved any such clear definition and formal character. The man who first used the word mar in the sense of to propitiate undoubtedly did so because he saw darkly the mollifying effect upon a bristling enemy, which Miiller describes, as a situation in which something rough swiftly became some- thing smooth. The differing details did not obtrude themselves into his consciousness. In essence the two situations were identical. So in the case of the children whose "metaphorical" speech is noted by Sully. "A star," he says, "looked at, I suppose, as a small bright spot, was called by one child an eye. The child M. called the opal globe of a lighted lamp a 'moon.' 'Pin' was extended by another child to a crumb just picked up, a fly and a caterpillar, and seemed to be something little to be taken between the fingers. The same child used the sound 'at' (hat) for anything put on the head, including a hair-brush. Another child used the word 'key' for other bright metal things, as money. Romanes' child extended the word ' star, ' the first vocable learned after 'Mamma' and 'Papa,' to bright objects generally, candles, gas-flames, etc. Taine speaks of a child of one year who, after first applying the word 'fafer' (from 'chemin de fer') to railway engines, went on to transfer it to a steaming coffee-pot and everything that hissed or smoked or made a noise."'' The perception of the first child was that only of "a small bright spot," not of a star or an eye, in our sense of the two words; that of the second was of a pale, glowing sphere, out of which the separate objects "lamp-globe" and "moon" had not yet developed. The third applied the word ' ' pin " to the situa- tion of picking up something small in the fingers. ' 'At " was putting something to the head; "key," a small bright metal article; "star," "bright objects generally;" "fafer," "everything that hissed or made a noise." Each of these words represented, not two differentiate percep- ceptions united by the tie of resemblance, but a perception as yet ^ See p. 5. * Studies of Childhood, Ch. V, pp. 162-3. — 15 — homogeneous and undeveloped, which must later separate into its con- stituent elements.' Let the final instance be that of the word "spirit." Between our sense of the word and the primitive meaning "breath," there was at first no distinction of thought. Spiritus meant the animating principle, simply, the essence of life, not yet differentiated into physical and spiritual. The rise of such conscious differentiation was marked by the limitation of the word spirit to the immaterial meaning. We may say, then, that radical metaphors were not metaphorical ' when first used. They have become so only as we have learned to dis- \ criminate between the two situations to which the same term has been \ applied. Until men recognized a difference between breath and spirit, the word spirit had no metaphorical sense. In other words, the radical\ metaphor is psychologically a survival from a primitive stage of percep-/ tion, a vestige of the early homogeneous consciousness. It represents a state of mind which does not now exist in relation to these same objects or situations. Hence the puzzle; and hence the natural anachronism of assuming our own more differentiate perceptions as the basis for an explanation of this phenomenon. This is what Locke and all his fol- lowers have done. Because we, of a later age, see with some distinct- ness the essential difference between spirit and breath, therefore the primitive man did — so they argue. And having once postulated this statement, they are thereby driven to connect the two conceptions again, in order to bring them under the mantle of the same word. This they do by the obvious analogy between them, the perception of which is made at once the inciting cause and the justification for the primitive speaker's application of the name of the one to the other. But to assert that for the modern intelligence there exists a sense of "Der parallel- ismus in den Phanomenen der Welt,"^ is by no means proof that it was overtly perceived by the naive mind of the aborigine. On the contrary all the doctrines of recent psychology lead to the conclusion that the actual perception of resemblance is an activity comparatively late in its development. A vague feeling of the identity of two sensations or situa- tions must repeatedly be experienced before it can b^ analyzed into the intellectual recognition of a resemblance or analogy. This first hazy feeling is the original fire-mist, out of which solidifies but slowly the explicit comparison. This dictum of psychology is confirmed by the later philological revelations which declare that the early utterances of man were the expression, not of clear-cut finished conceptions of single objects, but of ^ Sully says of the case of Prayer's boy, who "confused 'too little' with 'too much,' and 'yesterday' with 'tomorrow,' it is easy to see that the child's mind had reached merely the vague idea unsuitable in quantity in the one case and time not present in the other, and that he failed to differentiate these ideas." The confusion between 'learn' and 'teach,' 'borrow' and 'lend,' is thus explained: "Such words as learn, teach, call up first a pictorial idea of an action in which two persons are seen to be concerned. But the exact nature of the relation, and the difference in its aspect as we start from the one or the other term are not per- ceived." — Studies of Childhood, Ch. V, p. 165. 2 Bauer, Das Bild in der Sprache, II, 28. Wie- the individual's hazy, ill-defined sense of a complex situation, out of which, one by one, the details were gradually to develop. Under this latter hypothesis the confusion under the same name of two situations potentially differing in detail becomes readily conceivable. Radical metaphor thus may be regarded as a natural and inevitable outcome of the conditions of the primitive consciousness; while the elder theory makes it not only artificial but psychologically impossible.' y We are thus compelled to discard altogether the borrowing theory of the origin of the radical metaphor, as based upon a false assumption of the psychology of the aboriginal mind, and are driven to the belief that this figure arose out of the vagueness of a first unorganized sensation; the same term being, thus, without sense of transfer, applied to situations now recognized as essentially different, but then felt to be identical. To paraphrase the common dictum, it may be said that radi- ( cal metaphor arises not from poverty of language but from poverty or l^immaturity of thought. As any situation is enriched by the emergence of its constituent details into the consciousness of the speaker, it becomes more and more impossible to identify it with another situation. The chances of dissonance are increased as by the multiplication of tones in a musical chord. It is only a simple, because an undeveloped con- ception, that can be confused with another. Thus radical metaphor is observable only in past stages of the growth of a language. It exists continually but can not be recognized as such until the stage of hom- ogeneity for any given perception has been passed. Only when the situations concerned have complexed until they have become discordant and incompatible, can they be recognized as two, and their expression under one form of language be felt as metaphorical. To summarize briefly, radical metaphor is a backward-looking term. It characterizes a past stage of homogeneous perception, now become heterogeneous. Having become heterogeneous, yet bearing a homoge- neous name, it offers a puzzle to linguists and philosophers, who have hitherto failed to recognize the past state of the perception represented by the figure; but assuming the present as having always existed, built up their explanation upon this assumption. On going back to the prim- itive consciousness, however, we must regard the radical metaphor as an expression loosely comprehending two or more situations, now quite distinct, but at the time of their inclusion under the same term, felt to be one. The term "radical metaphor" is, then, a milestone of society's intel- lectual progress. It marks the point at which an expression, felt by its maker to be single, the exact representative of a simple sensation, is recognized by others as dual, as suggestive not of one impression or sit- uation, but of two. Other people could not, however, thus recognize 'Miiller takes us to the lowest depths of absurdity when he asserts that "one of the first results "of an age in which there were no names save those applicable to the commonest experi- ences of a savage life ''would naturally be that objects in themselves quite distinct, and origin- ally conceived as distinct by the human intellect, would nevertheless receive the same name."" — Science 0/ Lang., Second Series, p. 373. — 17 — the double reference of the expression, had not their perceptions become keener than those of the original maker of the expression. To their sharpened intelligences the fire and the sun, "the spring of the year, the morning light, the brightness of thought, the joyous outburst of hymns of praise"' could no longer present the same sensation. The process of crumbling, sickness and death, evening and night, old age and the fall of the year, grew distinct from one another. The same name as applied to these different situations was not felt to be natural, inevi- table, but unusual, demanding explanation. The differences between the situations had dawned and the use of the same word to represent them had become metaphoric. In so far as this recognition of the metaphoric or dual sense of the word "crumbling," for instance, as applied to old age takes place, the niind which recognizes it has passed beyond the intellectual development of the first user of the expression. He made no metaphor, in the ordi- nary sense of the term. There was no doubleness in his thought or in his expression. He but applied to a single situation what seemed to him its ordinary, accepted name. It was a situation not before represented in language because it had not hitherto been recognized explicitly enough. Our primitive metaphor-maker, however, perceives it, although so vaguely that it is not distinguished from other situations later found to differ from it. He, accordingly, calls it by the name commonly applied to the differ- ing situations which produce upon him the same impression and neither he nor any one else perceives that he has made an unusual application of the name. His very naming of this new situation has, however, called attention to it. It begins to be noticed by others and thus gradually to educate the perceptions on which it plays, leading them to finer and yet finer discriminations. Little by little dawns the sense that this situation is, in some respects, different from the others grouped with it under the same designation. It breaks away from these other situations, and clamors for a separate name. The use of the single word to cover two different impressions is then characterized as radical metaphor. The social consciousness has passed, in respect to this word, from homo- geneity to complexity. The intellectual advance of society has carried it past the point of our primitive speaker's confused impressions to a finer discrimination. Here the situations to which he naively applied the same term, seeing them as undistinguished, have grown apart; and lo, his simple speech has become a figure ! He himself has, indeed, con- tributed to this intellectual advance of society, by the very expression of his sensations in language; the expression has acted to bring about a condition in which it shall be felt as inadequate. It is the old tragedy of progress. The perception of a radical metaphor, it may be said, is society's recognition of the fact that an idea has quite outgrown its habitual form of expression. When uttered the expression was sufficiently differentiate 'Muller, Science of Language^ Second Series, p. 371. — 18 — for the embryo thought; now the thought has grown into discriminations unrecorded by the word. The result is a tension within the expression itself which brings to light its diverse meanings and demands relief. This is accomplished usually by a gradual movement in the direction of limit- ing this word to one of its tugging significations, and remanding the other meaning to a different phrase. Thus the word spirit is now used only in its immaterial sense, the material signification being represented by the Saxon breath. Apprehend, attention, angel, right and wrong, (refer- ring to the examples used in this chapter), are purely intellectual terms now, their correlatives lay hold of, stretch toward, a messenger, straight and twisted, serving the physical meanings.* To put the whole matter into a sentence: Specialization in language follows at some distance specialization in thought'; and the recognition of any expression once simple as metaphorical marks the social demand for a division of labor on the part of language which shall make it adequate to the growing differentiation of the thought it represents. If a definition be required, radical metaphor arises when a thought has outgrown its form of expression. It is the bursting forth of a doubly branching sig- nificance from the single sheath of language once adequate to contain it. 1 "Thus as the general movement of human knowledge is from recognition of sensible objects to an ever finer analysis of their qualities and determination of their relations, and to the appre- hension of more recondite existences, objects of thought, so, as the accompaniment and neces- sary consequence, there is a movement in the whole vocabulary of language from the designation .of what is coarser, grosser, more material, to the designation of what is finer, more abstract and conceptual, more formal." — Whitney, Life attd Growth of Language, Ch. V, pp. S9-90. CHAPTER II. GENESIS: THE POETIC METAPHOR. The last chapter discussed the theory of the philologists, that met- aphor arises from poverty of language. We shall now examine the hypothesis propounded by the rhetoricians, that this figure comes into being as the result of a desire to beautify or to energize speech.' Broadly speaking, the difference between radical and poetic met- '<^ aphor as commonly understood, is the difference between artlessness and art. Radical metaphor is the naive act of a mind not yet conscious of its own processes, making a figure because it must. Poetic metaphor is the purposeful effort of a writer to produce a certain effect, either upon himself or upon his reader, by the substitution of one word for another. The first is an act of necessity, the second of free choice. The first is at most but half cognizant of the end attained; the second has this end explicitly in mind and works deliberately toward it. This, I say, is the broad distinction between the two as it is com- monly professed. We have seen that Max Muller's discussion of the radical metaphor, which we may take as typical of all existing treatments of the subject, makes it, in the last analysis, a self-conscious and some- what formal thing, by assuming that its two constituents exist separately in the speaker's mind and are welded together by his perception of a resemblance between them. But in spite of this illogical departure from his initial declaration, the idea seems to have persisted, not only in Muller's own mind, but in that of succeeding rhetoricians, ^ that radi- cal metaphor does essentially differ from the poetic variety by its greater naivete and unconsciousness of its end. The poetic metaphor is the radical metaphor grown into a consciousness of itself. While the radical met- aphor is unsophisticated, crude, forced upon its maker by the necessi- ties of his undeveloped vocabulary and his straining impulse to expression, poetic metaphor is "the premeditated act of a poet,"' "a conscious transference of a word from one object to another." This distinction between the two chief varieties of metaphor rests, then, upon the conceptions of each which are commonly entertained among philologists and rhetoricians. Whether or not the present inquiry will sustain it, depends upon the definition of the poetic meta- phor which we finally accept. So far as our study of the radical meta- * See Ch. I., p. 2. *See Ch. I, p. 3. 'See Ch. I, p. 3, and Bauer, Appendix B. Note also Gummere's statement: "In like manner we may render abstract by concrete. This is unconsciously done when- ever we speak of abstract ideas, for they can be expressed only by concrete words. Such a case is the word attention, which passes as abstract, but really means a stretching towards. Or we may do it half -consciously, as in the expressions 'deep thought,' 'cool determination.' But in poetry we do it consciously, as in the following: 'The very head and front of my offending.' — Othello." [The italics are quoted.] — Handbook of Poetics, p. 93. — 20 — phor has taken us, the distinction is upheld. The conclusion has been reached that radical metaphor is, as the philologists have said, but in a sense more thorough-going than theirs, a naive, an untutored, an uncon- scious act. So far from the primitive man's setting out to make a metaphor, he in reality did not make one at all. Society made it for him, by passing beyond the point at which the situation he saw could seem one. The radical metaphor was, then, an act thoroughly straight- forward and undesigned. If now, we shall find that poetic metaphor is an act thoroughly self-conscious and predetermined, we shall be justified in confirming the judgment already passed by the rhetoricians as to the essential separation between these two species of the figure. Let us, then, inquire seriously whether the poetic metaphor is, in truth, "the premeditated act" of its maker. Undoubtedly the origin of this idea lies in the fact that the making of poetic metaphor could not be regarded as, in the practical sense, at all necessary. Clearly the radical metaphor was needed; for, according to the old explanation, a man who wished to designate an object yet unnamed, was powerless to do so unless the name of another related object could be extended over it. But for him who makes a poetic metaphor no such need exists. Both objects concerned in his metaphor have long been fitted with proper designations. If, therefore, he chooses to use the name of one in place of the name of the other, it must be that he does so with some purpose less simple and obvious than that of relieving the immediate necessities of his speech. His act cannot be the instinctive one of seizing what he needs, as the hungiry man snatches food. It must b^ far more complex than this. He could have used simply the proper name of the object of which he thought. Why did he not do so.-* Why did he take the trouble to substitute for this accredited name that of an object resem- bling the one of which he wished to speak.^ What did he expect to gain by such substitution.? What could have been his motive."* As was noted in Chapter P the answers to this question are many, but all may be reduced to two. The writer wishes either to beautify or td^energize his speech. But this is not a final answer. Why should he wish to do either of these things.? What does it profit him to make his speech either more beautiful or more forcible.? What is his ultimate end.? And here also we have a dual answer. If the writer wishes to make his expression either ornate or strong, it is for the sake of its effect upon one or the other of the two parties to the process of communica- tion by language — himself and upon the reader. To state the case more specifically^'' the writer wishes to give pleasure to himself by indulging his natural love of adornment, in speech as elsewhere',^> or to please* the reader.* *p. 2. ' Including under this term the production of any desired effect. »Such explanations as that the maker of a poetic metaphor substitutes one name for another because a resemblance exists between the objects represented by each, is, of course, no explanation at all The fact of resemblance may, indeed, make such an exchange of names possible, but it will not necessarily bring it about. We could not speak of the sun as a golden — 21 — The idea that the maker of a poetic metaphor desires to please himself by substituting the name of one object for that of another resembling it, is not, to my knowledge, explicitly affirmed by any writer; though Cicero may suggest it in his comparison of metaphor' to bodily ornament. If we may suppose that ornament is adopted, in part at least, for the wearer's pleasure in it, we are entitled by the terms of the analogy to conclude that the use of metaphor results to some degree from the desire for the pleasure it gives the writer himself. We need not delay long upon this theory, which is treated chiefly for the sake of completeness. We may, however, note in passing the psychological difficulties in which it immediately involves itself. The "hedonistic fallacy" of supposing that the idea of pleasure can serve as a motive to action has been so clearly exposed, both by James and by Uewey, that direct quotation from their discussions of this subject will be suffi- cient to disprove the hypothesis. "We feel an impulse," says James, "no matter whence derived; we proceed to act; if hindered, we feel displeasure; and if successful, relief. Action in the line of the present impulse is always for the time being the pleasant course; and the ordinary hedonist expresses this fact by saying that we act for the sake of the pleasantness involved. But who does not see that for this sort of pleasure to be possible, the impulse must be there already as an indepe7ident fact? The pleasure of successful performance is the result of the impulse, not its cause." ^ And later, he affirms that "the pleasure-philosophers," who hold that we act always for the sake of the pleasure to be gained from the action, might as well "suppose, because no steamer can go to sea without incidentally consuming coal, .... that therefore no steamer can go to sea for any other motive than that of coal consumption."- In a foot note upon the same page is quoted Hume's assertion:^ "Though the satisfaction of these passions gives us enjoyment, yet the prospect of this enjoyment is not the cause of the passions, but, on the contrary, the passion is antecedent to the enjoy- ment, and without the former the latter could never exist.*" ball if a likeness in shape did not exist between the sun and a ball. But the mere fact of this likeness is powerless to induce us to substitute the word ball for the word sun. Why should we take the trouble to call the sun a golden ball, even if the two objects are alike? What do we gain by it? What is it that we wish to do? There must be some motive for the transfer, to make the fact of resemblance more than the inert possibility of a metaphor. "As if analogy were somehow a force!" remarks Dr. John Dewey, in quite another connection. (Psych. Rev., ■Vol. I, p. 557-) 1 Psychology, Vol. II, Ch. XXVI, p. 5^7. "^Psychology, Vol. II, Ch. XXVI, p. 558. 'From his Essay on the Different Species of Philosophy, § i, note near end. *In the "Briefer Course," James declares that the ideas of pleasure or pain are not the source of our primitive actions, those which are instinctive or emotional. "Who smiles for the pleasure of smiling, or frowns for the pleasure of the frown? Who blushes to escape the dis- comfort of not blushing ? Or who in anger, grief or fear, is actuated to the movements he makes by the pleasures which they yield? In all these cases the movements are discharged fatally by the vis a tergo which the stimulus exerts upon a nervous system framed to respond in just that way." He denies, also, that pleasure or pain cause our habitual acts. ''All the daily routine of life, our dressing and undressing, the coming and going from our work, or carrying through of its various operations is utterly without mental reference to pleasure and pain, except under rarely realized conditions As I do not breathe for the pleasure of breathing, but ' OF THB t -^ UNIVERSITY 3 — 22 — Dewey makes a clear-cut statement of the case: "We do not desire the object because it gives us pleasure; but it gives us pleasure because it satisfies the impulse which, in connection with the idea of the object, constitutes the desire. The child desires the apple, for he has the idea of the apple as satisfying his impulse. Only for this reason does he conceive it as pleasure-giving. Pleasure follows after the desire, rather than determines it.'" " It is true," he admits in another passage, "that good (happiness) is the satisfaction, evil the thwarting of desire. This measures or defines happiness in terms of desire: desire is the primary fact, happiness its fulfilling, its completion. Hedonism sees the con- nection, but reverses its direction. It takes happiness as a fixed fact, and then tries to define desire in terms of happiness — as that which aims at it. It is true that happiness is found in the satisfaction of any desire, particularly in the degree of its dominance: happiness is this satisfaction of desire. But hedonism transforms this fact into the notion that somehow pleasure is there as an ideal, and its contemplation arouses desire. As Green says {^Prolegomena, p. i68), the hedonists make the 'mistake of supposing that a desire can be excited by the anticipation of its own satisfaction.'"^ Dewey's own explanation is made in a succeed- ing passage: "Instead of the image of pleasure exciting the action, the activity already going on sets up a pleasure by calling into consciousness the conditions (the object) of its satisfaction. There is no image of a past pleasure once experienced or of a future pleasure to be attained; there is a present pleasurable experience."^ Green remarks, trench- antly, that pleasure "cannot be the exciting cause of the desire any more than the pleasure of satisfying hunger can be the exciting cause of hunger."* This is perhaps sufficient philosophic authority for the statement that the poetic metaphor can hardly have arisen from the desire of the maker to please himself by ornamenting his thought. The metaphor, once made, doubtless does give pleasure to its maker because it fulfills a deep-lying impulse or instinct of his. He follows this instinct and gains pleasure therefrom. Because the fulfillment of the impulse has been pleasurable, that is, the activity has moved on unthwarted, it has grown stronger and, when next it rises, will demand fulfillment with greater insistence — not at all because the maker of the metaphor thinks of and desires the pleasure which is likely to accompany the fulfillment of the impulse; but simply because this pleasure, which is the subjective read- simply find that I am breathing, so I do not write for the pleasure of the writing, but simply because I have begun, and being in a state of intellectual excitement which keeps venting itself in that way, find that I «/« writing still. Who will pretend that when he idly fingers his knife handle at the table, it is for the sake of any pleasure which it gives him, or pain which he thereby avoids? We do all these things because at the moment we cannot help it; our nervous systems are so shaped that they overflow in just that way. — Psych., Briefer Course, Ch. XXIII, pp. 44e^7' 1 Psychology, pp. 361-2. * Study of Ethics, p. 45. 1 ' Study of Ethics, pp. 47-8. ^ Proldgomena, \i. iti\. See also pp. 163-177. — 23 — ing of a condition of unchecked activity, has tended by the law of habit, to reinforce the original impulse.* cr To put the whole matter into a sentence: One does not make a poetic metaphor because he desires the pleasure which it will give him, but he makes it because he has to, gaining pleasure therefrom, though he has not directly aimed at it. Just why he " has to" make the meta- phor, is a question which we shall discuss in connection with the second hypothesis. Since most rhetoricians of the old school believed the process of . discourse to be essentially persuasive, they tended to emphasize always the hearer above the speaker. Their precepts were all directed to pro- ducing a certain effect upon the hearer. , Do this in order that you may conciliate him; do that to win his confidence; attempt by a third device to make him think well of himself. Seldom was it recommended "Say this because you feel it to be true; only express yourself."^ From this fact we should expect to find the rhetorical conception of metaphor strongly tinctured with the prepossession that a desire to produce some effect upon the reader or hearer has been its inciting cause. And this expectation is confirmed by a survey of the principal treatises upon rhetoric. Aristotle leans to the theory that the maker of a poetic metaphor > desires to give pleasure to his reader. One uses a metaphor because this figure is best adapted to impart instruction without requiring any effort on the part of the recipient, and "learning without trouble," we are assured, ' ' is naturally agreeable to everybody. " ^ Cicero seems to have vaguely in mind the same answer. His first "t^ statement is that the metaphorical use of words, originating in pov- erty of language, "became common from the delight which it afforded."* From this assertion we are not sure whether the delight is afforded to the writer himself or to the reader, or, possibly, to both. Later state- ments, however, seem both to account in part for the delight and to assign it, though without definiteness, to the reader. Metaphors of the more conscious sort are said to "bring some accession of splendor to our language."^ "Such words should be metaphorically used as may 1 The argument runs so far afield into psychology and aesthetics, that the proportions of this chapter will hardly allow its detailed exposition. The following references contain the data for the conclusions reached: Dewey, Psychology, ch. XVIII; Study of Ethics, ch. V; Outlines of Ethics, pp. 13-42; The Theory of Emotion, Psych. Rev., Nov., 1S94 and Jan., 1895. James, Psychology, Vol, II, pp. 549-559, Briefer Course, pp. 444-445. Green, Proleg. Bk. Ill, ch. I. 2 In admitting this general fact, one does not necessarily deprecate it. Doubtless it has been responsible for much of the discredit which even now attaches to the term rhetoric; but doubtless also it was necessary to over-emphasize the hearer in the two-sided process of dis- course, lest in the early development of the art, he be altogether overwhelmed under the natu- rally greater weight placed by the speaker upon his own ideas and feelings. 3 Rhetoric, Bk. Ill, ch. X. * De Oratore, Bk. Ill, XXXVIII. See Ch. II, p. 21. * De Oratore, Bk. Ill, XXXVIII. — 24 — tnake the subject clearer,"' Metaphor is employed "that the descrip- tion may be heightened,"' or "that the whole nature of any action or ■design may be more significantly expressed."^ "Sometimes also brevity is the object attained by metaphor."' These scattering suggestions come to a focus if we venture to formulate roughly Cicero's explanation of the motive for using metaphor as follows: Metaphor is employed for the sake of giving delight to the reader by heightening the writer's natural expression in language, making it more splendid or significant, clearer or briefer. Just how this heightening of the effect of language by means of metaphor becomes pleasurable he explains at length; but we shall postpone to the following chapter a consideration of this point. We are at present concerned only to know Cicero's belief that the names of two objects are interchanged because the writer knows that such inter- change will produce a certain effect upon his reader. ^ Quintilian thinks that we make the change from a "proper" to a metaphorical word ' ' either because it is necessary, or because it adds to significance, or because it is more ornamental."^ "We say," he asserts, "that a man is inflamed with anger, burning with desire, and has fallen into error, with a view to significance or force of expres- sion The expressions luminousness of language, illustrious ,K '' J birth, . . , thunderbolts of eloquence, are used merely for ornament." ^ '"'' And further in the same chapter, he declares that "metaphor has been invented for the purpose of exciting the mind, giving a character to things, and setting them before the eye."* Hitherto Quintilian had not told us definitely why the speaker , should wish to make his language forcible and ornate. But from the last passage cited we may conclude that the ultimate end is its effect upon the mind of the reader. Quintilian, then, joins Cicero and Aris- totle in affirming the writer's anticipation of the effect of his figure upon the reader as the motive of his employment of it.* X Campbell^ treats metaphor as a means to vivacity of language, that is, to producing a strong impression upon the reader. Whately's idea of the cause of metaphor is sufficiently evident from his famous recommendation that metaphor be used rather than simile wherever possible, ' ' because all men are more gratified at catching the resemblance for themselves, than at having it pointed out to them."' The writer, it is assumed, desires to produce upon the reader a pleasur- iDe Oratore, Bk. Ill, XXXIX. ^Institutes, Bk. VIII, Ch. VI, §4. »/«j/i7m/^^, Bk. VIII, Ch. VI, § 7. ♦§I9- 'Cf. Puttenham's three reasons for using metaphor: Arte of English Poesie, Lib. Ill, XVI: 1. " Necessitie, or want of a better word." 2. *' For pleasure and ornament of our speech, as thus in an epitaph of our own making •Whom virtue rerde, envy hath overthrown And lodged full low, under this marble stone.' 3. "To enforce a sense and make the word more significative." ^Philos. ofRket., Bk. Ill, Ch. I. ^Elements 0/ Rhetoric, Pt. Ill, Ch. II, § 3. f UNIVERSITY — 25 — able effect. This being granted, it is necessary only to inform him that metaphor is more likely to produce this effect than simile, to induce his choice of the former, "wherever possible," this saving clause perhaps meaning wherever fidelity to the writer's own self-expression does not forbid. Adams says that figurative language which originated from necessity was "afterward greatly multiplied by the charm which the discovery and display of ... . analogies possess over the minds of men,'" not distinguishing between the pleasure of the writer and that of the reader. Mr. Arlo Bates declares explicitly: "The object of using figures is to add clearness, or force or elegance — or all of these — to the presenta- tion of an idea. Constantly it happens that, by declaring that an unknown thing is like some known thing, the writer enables the reader to form an idea of it as it is Figures are used to increase the lucidity of style. "^ And E. E. Hale, Jr., after having stated that the simple conduces vi- to clearness and the metaphor to an effect of brilliancy and dash, exhorts the writer to use these figures with his eyes open. " You should be able to know," he says, "when you are writing, what you are trying to ac- complish. If you are writing some scholarly paper, you want to be clear If you are telling a story, you want to be brilliant, lively, vivacious. You must be able to tell for yourself what is the chief aim you have in mind, and you should now have some idea i of one of the means. by which it may be attained."^ Mr. G. R. Car- penter's statement is no less decisive: "A metaphor, as you no doubt know, is calling one thing by the name of another, for the sake of leav- ing on the reader's mind a more vivid or picturesque impression."* The utterances last cited imply with some distinctness the theory that a writer makes metaphor for the sake of producing a certain effect upon the reader. Comparatively few, however, of the later rhetoricians make any definite statements as to the motive in the writer's mind which incites to the making of a metaphor. We shall find, nevertheless, that almost all of them justify their practical precepts, (such as that in a met- aphor the resemblance should be neither too obvious, nor too obscure, that no metaphor should be pushed too far and that none should be mixed with plain statement), by the fact that conformity with these prin- ciples gives the reader pleasure, or, more generally, that it produces upon him the effect desired by the writer.* ^ Lectures on Rhetoric and Oratory, Lect. XXX. 2 Talks on Writing English, Ch. Vn. ' Constructive Rhetoric, p. 267. ^Exercises in Rhetoric, Adv. Course, 2d Ed., Ch. XII, p. Igg. 5 A. S. Hill remarks of tropes in general, that they, " should be fresh enough to give the reader a pleasant surprise, but not so strange as to shock him." — Prins. of Rhet., Ed. 1896, p. 131. DeMille objects to hackneyed metaphors, because, ''when the resem- blance is too familiar, the image has no effect;'" to strained metaphors because "when the resemblance is remote, it is not readily understood or appreciated." — Elements of Rhetoric, pp. 115-116. See also Genung, Pract. Rhet., ed. 1894, pp. 92-93; Gummere, Handbook of Poetics, p. 951; J. G. R. McElroy, Structure of English Prose, pp. 243-246; J. M. Hart, Hand- hook of English Composition, pp. 188-191; Barrett Wendell, English Composition, pp. 250-261. — 26 — It may, perhaps, be urged that the effect of the metaphor upon the hearer is regarded by the rhetoricians rather as a test for the metaphor, after it has been made, than as a statement of the end aimed at in mak- ing it. But it must be remembered that a test is meaningless except it refer, directly or indirectly, to the end for which the thing tested has been designed. Ultimate judgment must always be based upon the answer to the question, "Does this thing fulfill the purpose of its crea- tion.-'" Such a test the rhetoricians are accustomed to apply to the metaphor when they say, "If it fails to produce a certain effect upon the reader, it is worthless. Discard it." The metaphor is designed, so these practical precepts may fairly be interpreted, to make the reader see ,the point more clearly than he could otherwise do, to give him a more vivid or forcible conception of the object presented, to stimulate his mind, — in brief, to produce upon him a certain effect. We may, then, conclude, both from the stated theories of the writers upon rhe- toric, and from the implications lying in their practical precepts, that they, as a class, are fairly committed to the theory that a metaphor is made for the sake of inducing a certain effect upon the mind of the reader. * - ' We are asked to suppose that any writer who uses a figure of speech does so with a definite end in view, that of making the expression of his ideas more pleasurable^ to the reader than it could be if plainly ' One apparent exception to this statement should be noted. Mr. L. A. Sherman, in his Analytics of Literature^ defines a metaphor as "seeing one thing spiritually identical with another thing," (notes, p. 399) and states explicitly that the writer in a particular case "saw mentally this same identity, and said or wrote the metaphor because he experienced it thus vividly in his mind." (p. .62.) Having made this most interesting statement, however, he at once proceeds to discredit it, by referring casually to metaphor as "the assignment of two objects to a new class by using it (resemblance) as the basis of classification" (p. 62) and hinting pretty broadly that the fact that this operation is "especially agreeable to the ego" (p. 62) bears a casual relation to the production of the metaphor. Here are certainly two widely differ ent statements. The one shows the writer as having united two perfectly distinct objects into a single class because he wished the pleasure that the process would secure to him. The other makes the writer express directly his own vision of the identity of two objects. Sherman first adopts the latter hypothesis. According to this, the writer sies, at the height of the metaphor-process, not two objects at all, but one. Yet the one either had been two before this supreme moment of the writer's vision, or it became two after that moment had passed. Which belief does Sherman hold? His treatment of the subject clearly depends upon his faith in the first. "In metaphor it (the mind) perceives two objects, each with equal vividness, spiritually identified." (p. 399.) There were two objects. They have been fused into one. The process by which this fusion takes place now becomes the difficulty, and forces Sherman into a tacit acceptance of the hypothesis which regards two objects as being united in a single class by virtue of a resemblance existing between them, and, for the sake of giving pleasure to the author of the union. If Sherman had seen that psychologically the writer's vision must be single before it can be dual, he would not have predicated the existence of the two objects in consciousness as antecedent to the presence of the one, or their identity, and thus would not have been compelled to close up by artificial means a gap that did not actually exist. It may, then, be affirmed, that while Mr. Sherman refuses, at first sight, to be classified with the rhetoricians in his understanding of the metaphor-process, in last analysis he is at one with them. *I include, for the present, under this term all the cognate expressions, "forcible," "viva- cious," "economical," "stimulating," etc. The argument will attempt to disprove the theory that a metaphor is made for the sake of producing an effect upon the reader, pleasure standing as the representative of all these effects. — 27 — expressed. Suppose, for instance, that the writer wishes to speak of the sun. He might say "sun" without further ado, but that he desires to gain for his readers the pleasant sensation which he knows will result from the simple device of calling one thing by the name of another. Therefore he decides to call the sun something else — but what else.^ Here, as in the case of the radical metaphor, the fact of resemblance is the deus ex machina. The writer may be supposed to follow out in every direction from the object "sun" the radiating lines of its qualities or characteristics until one of these shall haply lead him to an object possessing the same quality or characteristic and thus connected with the original object, sun. Say, for instance, that he has in this way fol- lowed the line of the quality of roundness out from the round sun until he reached the round object ball. He can now call the sun a golden ball with perfect assurance that the reader cannot fail to receive there- from the pleasure which has been designed for him. It is difficult to avoid the ironical tone in discussing this explana- tion. It makes the act of metaphor so mechanical, so crude, so essen- tially cheap and tawdry that the sensitive reader of literature can hardly suffer serious consideration of its truthfulness. He revolts instinctively from the notion that Shakespeare, for instance, deliberately set about comparing the storm-whitened waves of the ocean to a culinary com- pound made foamy in the process of "rising," in order that he might tickle his reader's fancy by the phrase "the yesty waves. " ' Such an explanation virtually requires that Shakespeare, having one element in the metaphor — the storm-whitened sea — distinctly in mind, should con- sciously lay about him for an object which might serve as the other element. One must imagine him like the pedantic King Richard II, always studying how he might compare one thing to another. Or one must fancy the virile dramatist in the situation of a practiced conceit- monger, Thomas Sheridan, who thus owns to his poetic method: "I often tryed in vain to find A simile for womankind, A simile I mean to fit 'em, In every circumstance to hit 'em. Through every beast and bird I went, I ransacked every element; And after peeping through all nature. To find so whimsical a creature, A cloud presented to my view, And strait this parable I drew."' We are not surprised to learn that certain metaphors in our less naive and spontaneous poetry had their rise in such a process as this. Their flavor suggests an origin no less mechanical. And few are the happy teachers of English composition who have not thus explained to 1 Macbeth^ IV, 53. "^New Simile for the Ladies. The Poetical Works of Jonathan Swift, V6l. Ill, Aldine Ed.. — 28 — themselves the existence of many a metaphor in student essays.* The boy who speaks of Hawthorne as "the queen-bee in American Htera- ture," and the girl who characterizes reading as "the indispensable nectar of existence " present us no insoluble problem as to the metaphori- cal process which has gone on in their minds. Clearly it was somewhat like this: A figure is necessary to any well-regulated composition. Therefore let us have a figure. Since we are writing about Hawthorne, we may as well say that Hawthorne was something besides what he really was. Hawthorne was Hawthorne was What was Hawthorne, anyhow.^ He was awfully important in American liter- ature, the teacher said. Well, what else is important to something.? Perhaps a bee wings across the field of vision just at the moment of despair and is frantically clutched at by the despairing mind. Oh I guess Hawthorne was the queen-bee in American literature. And the successful author beams with satisfaction over the way that expression will "hit" the teacher. Not only, however, does the struggling student of composition thus manufacture his figures; but often the newspaper writer and compounder of dilute fiction seem to be conversant with the metaphor-recipe. Mr. A. S. Hill's incomparable illustrations,^ "hair shot through with sunset spikes of yellow light," "and lips with musical curves" are in point. Under the class of manufactured articles must also fall the follow- ing from Am^lie Rives: "The stars had looked like great drops of trem- bling quicksilver, just ready to splash from the inverted pewter spoon of the sky."* We may be quite as certain as internal evidence can render us that the writer of this astonishing figure, moved by a keen desire to make a metaphor, looked wildly about her for some like object to which she could compare the stars and the sky of which she wished to speak. At this particular hour in the morning the stars resembled silver in color; but this comparison was far too commonplace, and quicksilver was accordingly substituted as an equivalent. The sky was a dull, hard- looking gray. The use of quicksilver for the stars suggested pewter for the sky: and the shape of the aerial dome furnished another analogy, that of a pewter spoon inverted. Nothing could be neater than this process; and nothing surer invariably to furnish a metaphor on demand. 1 Such efforts as the following troop at once to mind: Margaret and Luke are reported as having at their first meeting seen "the juncture of two unfathomable revelations .... descend upon them like the downward flight of a soaring thing. Margaret felt the mist of a sudden indistinctness hurl her through infinite space, and a great gong seemed to beat a muffled distance of time and space across Luke's forehead." •'Liberty, powerful, just and equal, stands- as the beacon light for this country, whose foundations were laid and cemented on the blood of the patriots who died in defense of it." "Since that moment life has been a cesspool of malignity; an empty dream; a hollow mockery and a sham, totally nnknown to the smiling face of pleasure, and only goaded along its labyrinthian highway to its everlasting doom, by the most despicable spirit of yengeance, despondency and immutable woe." * Our English^ pp. 122-3. • The Story of a Heart , Am^lie Rives. Cosmopolitan, July, 1897, p. 331. — 29 — Swinburne has evidently employed such a device in his stanza: " Now the morning, faintlier risen, Seems no God come forth of prison, But a bird of plume-plucked wing Pale with thoughts of evening."' Such a metaphor as this can hardly have been produced otherwise than by holding before the writer's consciousness the thought of the pale dawn and casting about for an object which might in some particular resemble it. The traditional God of day is discarded because too robust. The morning has been shorn of its usual glories, so Swinburne says it seems a plucked bird. But this comparison does not convey a suffi- ciently vivid idea of the pallor of the dawn, so that the statement must be added that the bird is " pale with thoughts of evening." The figures cited are perhaps sufficiently absurd to demonstrate vj conclusively the essential difference between " those metaphors which rise glowing from the heart, and those cold conceits which are engendered in the fancy. "^ We feel acutely that there is a real division here; that those metaphors which we can conceive to have been manufactured by a con- scious effort of the mind directed to the recognized end of making a figure, for the sake, ultimately, of pleasing the reader,' inevitably fall into a class quite distinct from those fresh and vital figures which need no external witness to their spontaneous origin. For the former class of metaphors we should not hesitate to adopt -f the hypothesis that they have come into being as the result of the writer's effort to produce a certain effect upon the reader. But we shrink instinctively from this t-heory as applied to the latter species — the genuine poetic metaphor. We feel not only that it cannot account for these metaphors, but that it ought not to do so. Our half-conscious theories of literary art attribute to it a quality far less artificial — not to say meretricious — than that which characterizes the process of manu- facturing a metaphor in the manner described. We feel assured that to substitute one word for another in order to produce a pleasant titillation of the reader's fancy is a shabby gallery-play, beneath the dignity of the real artist. It is true that this is only a feeling, though a feeling so universal among people of literary sensibility that it might almost upon that ground be allowed in evidence. But if we trace this intuitional judg- if- ment to its source, we shall perhaps find that it roots in a philosophy of \ ^ the literary process not the less true because so often unconscious. For y the widespread feeling that metaphors made to produce a certain effect / 1 Pastiche, Poems and Ballads, Second Series. 2 Oliver Goldsmith, Essay on the Use of Metaphors. Cf . Emerson's distinction between imagination and fancy in the essay Poetry and Imagination. "Fancy is a willful, imagina- tion a spontaneous action." * This process will be farther analyzed under the head of the Pathology of Metaphor, Ch. V. — 30 — upon the reader fall below the level of real art, we find both source and justification in that philosophy of the literary process which regards it as having equal reference to two factors, the writer and the reader. This theory is that known as the communication theory of discourse, which has, in the later rhetorical systems, largely superseded the one-sided theories of discourse as persuasion and. as self-expression. The theory that discourse is self-expression has reference only to the speaker; the hypothesis that it is persuasion makes the hearer all-important. When discourse is regarded as communication the two factors in the process are equally emphasized. In order that the writer's vision be communi- cated to the reader's eye, the one factor is as necessary as the other. Not the reader alone is concerned, nor yet the writer, but both equally have a part in the literary process. , According to this .standard a piece of writing which seeks only to lay bare the writer's thought, with no reference at all to the capacity or interests of the reader, is condemned as bad art; and no less is the work found wanting which looks only to its effect upon the reader, little caring to be true to the vision of him who writes. And of this last sort must be the metaphor which is made for the sake of pleasing the reader, if no real sight of the writer lies behind. We find justification, then, in this conception of the normal literary process, for the instinctive feeling that such mechanical construction of the metaphor as is implied in the rhetorical explanation of its origin is bad art, because it leaves out of account one of the prime factors in the process of discourse. The dilemma then confronts us. Either we must accept the rhetorical explanation of the origin of metaphor, and bow to the conclusion that this figure lies outside the field of legitimate art; or we must deny the rhetorical explanation and hold to our instinctive faith in the artistic justification for a good metaphor. The first of these alternatives has found the widest acceptance. Acquiescing in the theory of rhetoric that the metaphor is an expression not necessarily of the speaker's own vision of things, but of his desire to make other people see them in a certain way, the "practical man" is straightway seized with a distrust of the figure, amounting almost to fear. He regards metaphors much as the old saints regarded women — as charmjng snares, in which he may too easily be entangled. Tell a jury that your opponent's most telling argument is " only a beautiful metaphor," and you have at once wholly discredited it. You have by this means conveyed to the jury with more or less distinctness the idea that your opponent has been trying to cheat them; that, without seeing the matter in that light himself, he has deliberately set out to make them see it as he wished; in a word, to produce a certain effect upon them for his own ends. Now this metaphoraphobia, if the term may be allowed, is only the logical consequence of the faith that metaphor arises from the desire of the writer to produce a certain effect upon the reader. We have a right to suppose that language will convey to us the speaker's thought, modi- — 31 — fied unconsciously perhaps by the speaker's knowledge of the hearer's capacities and predictions, but not wholly determined by them. When, however, we have learned to suspect any form of speech of occasionally mis-representing the thought it assumes to represent, we are bound to fear it as treacherous and misleading, and when we have finally become assured that this form of speech has no necessary reference to the speaker's thought, we can afford to scorn it. Having once determined its falsity, it can never deceive us again. We can safely snap our fingers at it and devote ourselves to warning other people against its deceptions. This last attitude is, as might be expected, that of the more vigorous and independent of straight- forward minds. Philosophers, logicians and scientific men, it has often been observed, exhibit a healthy scorn for metaphor, evidently regard- ing it in the light of a meretricious device, a crude overreaching artifice, unworthy of the simplicity of truth. Locke seems to voice the opinions of W-- gfp\ this class, when he says: " If we would speak of things as they are, we ^^^ (^ must allow that all the art of rhetoric, besides order and clearness, all J^ A<-^^^ the artificial and figurative application of words eloquence hath invented, rrJiMf^ , are for nothing else but to insinuate wrong ideas, move the passions, „Vt^*^ ) and thereby mislead the judgment, and so indeed are perfect cheats, and , ^^^ ^ therefore, however laudable or allowable oratory may render them in » -^ ^tf harangues and popular addresses, they are certainly, in all discourses '''*^'^^ that pretend to inform or instruct, wholly to be avoided; and where '^ ,, A truth and knowledge are concerned, cannot but be thought a great fault, '^■^*' ,^,, * either of the language or person that makes use of them."* -r^'^'li ^^ This is a point of view wholly self-consistent, if one hold with tfie i^*^^ ^^ faith that metaphor arises from the writer's desire to affect his reader in \^ ^^ a certain predetermined fashion. One who shrinks from this doctrine*'^ \ ^^^* may, however, seek another explanation of the genesis of metaphor. lyM*'^,^}^ And this search shall be our next task. Before entering upon it, we f^ may, however, note a further reason for discrediting the accepted rhe- torical theory, in that it makes no provision for the Arst poetic meta- phor. This surely cannot be explained as the product of a desire for the reader's pleasure; for the writer had then no assurance, either from his own experience or observation, that pleasure would result to the reader from the device of substituting one word for another. We are com- pelle'd to one of two explanations. Perhaps the writer stumbled upon the device. He "just happened" by a lucky accident to substitute one word for another; and the pleasure resulting from it to the reader either encouraged the writer to make another metaphor on purpose to confer this pleasure, or incited the reader to make one himself for another per- son's delight. But granting this to be a possible explanation, the "ac- cident " has still to be explained. There are psychological laws govern- ing the metathesis of letters in rapid writing; there is always some rea- ^Essay on the Human Understanding^ Bk. Ill, Ch. X, § 34: "Figurative speech also an abuse of language." I find in Blackwood 18: 719 the statement, " It is said to have been a b'>ast of Swift, or his friends, that ' he had hardly a metaphor in all his works.' " «a2— son why one mis-speaks.* ?i. ' -. .'did it happen that this original maker substituted one word for another ? The fact that a certain resemblance or analogy existed between the two objects whose names were con- cerned may have made the substitution possible. But we cannot sup- pose an explicit recognition of this resemblance on the part of the maker of the metaphor; for, as soon as we do this, we are forced to assign a motive that led to the transfer; and we have promised that no motive existed. The maker of the metaphor "just happened" to substitute one name for another. In his haste he mis-spoke. As in the case of the radical metaphor, the resemblance existed but was not explicitly perceived by the speaker. Could it then have been a factor in the sub- stitution which took place ? Undoubtedly it could, precisely as with the radical metaphor, where that which is later perceived as resemblance or analogy is, at the time of making the metaphor, seen but vaguely as complete identity between the two objects concerned in the figure. We can conceive that out of a primitive sensation of something fluffy white might spring into separate existence the two objects ' ' cloud " and ' ' snow. " At the moment when the two were disentangling themselves from the first homogeneity, our metaphor-maker might express the starting differ- entiation by saying " snow " as he pointed to the clouds. When he had done so, doubtless he would feel a certain pleasure in the unity of which he was dimly conscious under the variety, and the impulse to a second metaphor-making would thus be strengthened. And from this point the story is easily told. But this explanation practically identifies the source of poetic with that of radical metaphor — in the speaker's homo- geneous consciousness of a primitive sensation, out of which its con- stituent elements later separate themselves. So much for the first explanation of the ultimate origin of poetic metaphor. The second, we shall see, brings us to this same point. It is that this first maker of the poetic metaphor was led to the making because he had noticed the pleasurable effect consequent upon his per- ception of the radical metaphor and resolved to create enjoyment for himself or for another person by the same method, consciously employed. Now it is quite unlikely, to say the least, that the maker of the first poetic metaphor had already arrived at a stage of reflection so advanced that he could recognize his pleasure in the radical metaphor, analyze its sources, and determine to use the same means to attain a like enjoy- ment. "Poetic metaphor" appeared in ages far too unsophisticate for such subtle ratiocination. But whether we grant or deny the possibility of this explanation, we are with equal promptness brought to the source of the radical metaphor as the ultimate spring also of the poetic. We are, then, confronted at once by the question, "If the ultimate, why may not also the proximate source of the poetic metaphor be iden- ' There is a recent German treatise on this subject: Versprechen und Verlesen, by Ru- dolf Meringer and Karl Mayer, Stuttgart, 1895. Itica with that of the radical metaphoi ? 1 5 it not possible that the same psychological process of growing differentiation in perception may lie at the root of both phenomena ?" Let us face this question fairly. We have seen that the radical metaphor is the single expression of an undeveloped perception which later divides into its elements, the division bringing to light an incongruity between the dual character of the devel- oped perception and the simplicity of the phrase which once represented it. Now this process by which perception develops from a vague unity into a more clearly defined duality or complexity is asserted by psycholo- gists to be typical and universal. It must, then, go on in the modern as well as in the primitive consciousness. Civilization can only shorten the process, not do away with it. The savage who spoke of a nation that came "from so far off as the sun slept 'V perhaps never in his life-time learned to distinguish clearly between the action of a man in withdrawing to sleep, and that of the sun in disappearing for the night. The child in a civilized nation, however, completes the process in a far shorter time. The lamp globe is a moon to him at first. He sees, both when he looks at the moon and when he looks at the globe, exactly the same thing, so far as his undeveloped perceptive powers can tell him. But not many months pass before the two have successfully separated themselves, and the one can no longer be mistaken for the other. In the case of the adult civilized man, the process is still more rapid. Although for the flash of a second he may see a curled stick as a snake, he cannot do so for long. The perception rapidly differentiates until the two elements, the stick and the snake, become perfectly distinct in his consciousness. The whole development of the perception from homogeneity to heterogeneity, to use the scientific phrase, may, in this instance, occupy but the frac- tion of a second, instead of the years or ages needed for the slower- moving mind of the savage, and the months required by the undeveloped intelligence of the child. But in all these cases the process is the same. The sophisticated modern, when he gives utterance to perception before it has developed out of the homogeneous stage, is making a radical met- aphor just as truly as does the savage or the child. No two things are concerned in his thought, but only one. There is, in the ordinary sense of the word, no metaphor. The speaker has simply represented in words his own undifferentiate consciousness. But so rapid is the process of differentiation, that often utterance takes place when the two elements in the perception are just emerging from the primitive mass, and both, therefore, appear in the figure. Emily Dickinson in her poem called "The Snake" represents exactly 1 Compare the following observation: "En eSet, quand I'ecrivain suivant le tour de sa pensee, exprime les choses de la fa9on particuliere dont il les sent, ou las voit, il ne fait qu'obeir aux memes lois de I'esprit que le peuple. II n'y a point de difference entre les figures du style d'un ecrivain et celles de la langue populaire, sauf que chez I'ecrivain ce sont des hardiesses individuelles, tandis que chez le peuple, si ces hardiesses sont individuelles a I'origine, elles ont ete adoptees par tous, consacrees par I'usuage, et sont devenues habitudes de langage. — Arsene Darmsteter, La Vie des Mots, pp. 45-46. * Barrett Wendell, Eng. Comp., p. 249. K r ; — 34 — the development of a dim perception into definiteness. She says that she saw "a. whip-lash Unbraiding in the sun, — When, stooping to secure it, It wrinkled, and was gone. " Her hazy impression of a something long, brown, slender and convolute had already separated out of its mass the idea of "whip-lash", with others, such as that of a snake, just stirring into consciousness, when, "stoop- ing to secure it", of a sudden "it wrinkled and was gone", so that the dormant idea of "snake" sprang at once into full view and the figure was complete in which a snake is, according to the rhetorical dictum, "compared" to a whip-lash. Similar readings of poetic metaphors might be multiplied indefi- nitely. When Tennyson says that "fear chalked her face",* we can see that the thought has leaped to expression just at the instant when the first vague impression of a surface rapidly growing white was separating into the two yet half-conscious images of a surface being overlaid with chalk, and a face paling under the influence of fear. When he says again in the same poem, < ' I stole from court, cat-footed through the town", we are able to recognize the first nebulous perception as that of a stealthy, noiseless manner of walking, out of which were just emerg- ing into consciousness the constituent image of the cautious tread of a cat and the hero's own sneaking footsteps. When Milton talks about "low-browed rocks, "^ it is evident that the menacing effect of some- thing dark and overhanging is but just resolving itself into the beetling rocks and the low brows of a human face. When Keats spoke of the ' ' wings " of sweet peas, the vague perception of a rounded, half-pear- shaped outline was in the act of dividing itself into the form of the sweet- pea petals and the conventional shape of wings. Henry James says that "The gondolier's cry, carried over the quiet water, makes a kind of splash in the stillness^* thus showing the inchoate sensation of an inter- ruption, a sharp difference in the ordinary course of things, as branch- ing doubly into the two images of the sound produced by a cry breaking the stillness and that of a splash made in water. Such an explanation as this for the origin of metaphor has two advantages over that propounded by the rhetoricians. It is psychologi- cally defensible; and it avoids the artificiality of the rhetorical hypo- thesis. These two are of course, in the last resort, one, since the fact that metaphor is attributed to a normal and universal process of mind, which seeks only its natural expression in language, destroys at once the theory of an artificial origin. We may say, then,//in summary, that poetic metaphor, like radical, * 7'Ae Princess. *V Allegro. » Venice, The Century, Vol. XXV, p. 13. — 35 — is a straightforward attempt to communicate to another person the maker's vision of an object as it appeared to him at the moment of expression, not at all to carry out a dark design of persuading the reader that this object is something which the writer knows it is not. The writer is simply taking a snap-shot at his own process of perception in one of its intermediate stages. This stage may be one not reached by the maker of a radical metaphor. To him, the constituents of his vague perception had not yet disclosed themselves; while often the poetic metaphor expresses this primitive perception in the act of differentiation, two or more images appearing side by side in the figure. The mistake of the rhetoricians lies in their failure to go back of the simultaneous presence of these two images in the metaphor, to the earlier stage of perception in which the two were seen as one. Failing thus to trace back the duality of the metaphor to a primitive unity of consciousness, they were forced to account in some way for the presence of the two images side by side in the figure. The resemblance which existed be- tween them was accordingly made the connecting link, and the writer was conceived as making use of it for some purpose of his own to unite the two. Thus the whole artificial explanation of the rhetorical meta- phor has been built up. The occasion for it, together with the hypo- thesis itself, tumbles to the ground when we touch the question of"''^ psychological genesis. The old rhetorical hypothesis may be sharply contrasted with the psychological by saying that the former started with two objects and hitched them together to make the figure, while the latter begins with a single object or situation, out of which develop the two elements in the metaphor. The one explanation conceives of metaphor as a mechani- cal product, like a box, whose parts, gathered from different sources are put together to make the whole. The other regards it as the result of a vital process, more like a plant or an animal, whose members grow from the same source, out of a homogeneous mass into a clearly differ- entiated structure. The one represents the biological, the other the mechanical conception of the metaphor. Metaphor, from this point of view, is vital. It is not compounded like a prescription with intent to produce a certain effect upon the per- son who swallows it; but it springs spontaneously out of a genuine thought-process and represents with exactness a certain stage of a grow- ing perception. It is no artificial, manufactured product, but a real organism, living, growing and dying. We shall trace its further progress in the following chapter. CHAPTER III. THE EVOLUTION INTO PLAIN STATEMENT. We have traced the metaphor from its origin in a nebulous and un- developed perception of a situation to its first tentative outbranchings, from a state of perfect homogeneity to one of beginning differentiation. But our task is not yet ended. The growing perception does not at this point cease to develop. There is a sequel to the history thus far outlined. And this sequel is our present interest. What becomes of the developing perception of a situation after it has passed the metaphorical stage.' For instance, when the child says "moon" as he points to the lamp-globe, we know that "moon" repre- sents to him a single physical sensation which he refers vaguely to any large, white, softly glowing object which is present to his consciousness. The moon and the globe are one for his undiscriminating mind. But little by little out of this hazy sense of something large, round and softly glowing, emerge into half-distinctness the two objects lamp-globe and moon, probably first differentiated by the dawning sense of distance. When the two had become sufficiently separated in perception, they would doubtless be distinguished in speech. The lamp-globe might, for instance, become the "near moon" or "baby's moon." Later its proper name would be discovered and used, and the two objects would have become perfectly distinct in the consciousness, so that the common ori- gin would as such be quite forgotten. The two objects would stand in the mind of the speaker almost as completely separated as if they had always been so. Not quite, however, for between the two exists a con- nection, recognized by the speaker, the vestige of their former identity. This connection might find expression in such a statement as the follow- ing: "That lamp-globe looks like a full moon." It no longer is the moon, nor is it the "near moon" or "baby's moon." It is not a moon at all, but only looks like one. The two elements in the metaphor have separated so widely that each is seen as distinct from the other. So far then we have noted three stages of metaphor growth which may be briefly characterized as follows: The first represents that stage of perception in which the figure is still homogeneous. Teeth are pearls. The one name stands for a single sensation produced by a row of vaguely- perceived small objects, white, glistening and all but translucent. The second is that representing the stage of perception at which it has begun to differentiate into two main constituents. This is expressed by saying "pearly teeth." The third represents a later stage of perception in which the two objects, just beginning to draw apart from one another in the second period, have separated so far that a connection is visible be- tween them, this connection being commonly expressed in language by — 37 — the words "as" or "like." At this stage of the developing perception one would say "Her teeth are like pearls." The development of a metaphor is strikingly like the process known as fission in the case of the lowest forms of life. The amoeba, for in- stance, at one moment apparently a homogeneous jelly-like splash of protoplasm, shows an hour later a slight elongation and a constriction near its middle. It is becoming dual. Two sections of its body are be- ginning to show themselves where before was but one. The constriction narrows little by little, the two parts of the once single-celled animal be- come more distinct. Finally there is but a thread connecting them. It slowly parts, and there are two amoebae where one was before. In the metaphor process we have reached the stage at which the constricted portion has become visible as a connection between the two dividing sides of the figure. These two sides were at first indiscernible, united in the single homogeneous structure of the original perception. But as the perception developed it divided, and a connection of resem- blance was visible where complete identity had been. Gummere alone, of writers upon the metaphor, has recognized this ^ order of development. Metaphor, he argues,^ must have preceded simile in point of time. The one object was the other before it was seen to be like the other. " In genuine heathen poetry," he says, "there is no space for similes, since the gap between the literal and figurative, ab- stract and concrete, is so narrow that a metaphor leaps it almost uncon- sciously. There is no need for the simile bridge. I suggest, therefore, the following as a probable order of development in poetical style: i. That form of metaphor now known as allegory, entirely figurative, in- cluding, of course, personification. 2. Metaphor proper, where the lit- eral peeps through; as, in classique immittit habenas. 3. The meta- phor grown entirely conscious of itself, conscious of the gap between reality and figure; divided into (a) Implied simile: 'he is a lion' and (b) Stated simile: ' he ?.f ///^,■-..... Without going into the subject of allegory it may be noted that sxv-w,^^ Gummere's statement finds its explanation and justification in such a ,„ view of the origin of metaphor as has been presented in the foregoing.^^^^^^ chapters. His ' ' metaphor proper " corresponds to radical metaphor^ where the single name represents a single impression, yet undifferenti- ated. Here "the literal peeps through," but only from our own point of view, not from that of the speaker, who was unconscious of any du- plicity. Gummere's "conscious" metaphor is divided into "implied simile" and " stated simile."^ The implied simile corresponds to the "poetic metaphor," as we have termed it, in which the perception is just separating into its constituents: the stated simile to the simile as commonly understood, in which the constituent elements of the original perception have so far separated that the connection between them is recognizable as resemblance. 1 Mod. Lang. Notes, i :83, Metaphor and Poetry, 2This classification obtains also in Gummere's Handbook of Poetics^ Bk. II, Style, pp. 90-110. « ^Mt-a^tx^ ■ > — 38 — This last stage of development might, perhaps, with a degree of propriety, be termed ' ' conscious. " The resemblance or analogy between the two elements in the metaphor, which at first existed only as a hazy sense of identity, has now, in psychological phrase, "come to conscious- ness." It is recognized as resemblance. When Swinburne says:* "And fruit and leaf were as gold and fire," we know that the fruit and the gold are no longer included by him in the one large undifferentiated class of things that make upon him the sensation later identified as yel- low; that the leaf and the fire have become distinct from one another in his consciousness. They have so far separated that he sees them clearly as two, though recognizing that somehow they resemble each other. The fruit is in some unspecified respect like gold and the leaf like fire. So, in The Princess to draw again from this store house of metaphor, when the Prince, ' ' wrote In such a hand as when a field of corn Bows all its ears before the roaring East." the original vague impression of serried ranks bent all in one direction has divided into the sloping chirography and the field of corn, swept by a violent wind. These two elements have grown apart, yet a resem- blance is recognized between them. And when, in the same poem, we are told that " the women sang Between the rougher voices of the men Like linnets in the pauses of the wind." we recognize a separation between the two series of sounds, but a perception of the fact that these two have, after all, some common quality. 2 In all these cases, however, this common quality, the exact point of resemblance is not specified. The perception has not reached a stage of development sufficiently advanced for such specification. The analogy is felt rather than thought. It grows explicit, however, when the writer or speaker comes to perceive that not only two things are alike, but in what respect they are alike. This step is only the logical sequence of those which have preceded. The series of which we have hitherto spoken, beginning with the name "pearls" as applied to all small, white, glistening, translucent objects, and proceeding with "pearly teeth" and "teeth like pearls," is continued by the phrase "teeth white ^Atalanta in Calydon, First Chorus. 2 This stage of metaphor development is further exemplied by Shelley when he compares the skylark successively to four different things, "a poet hidden in the light of thought," ''a high-born maiden in a palace tovyer," " a glow-worm golden in a dell of dew," and "a rose embowered in its own green leaves." The exact point of likeness is nowhere specified, but it is implied, as Moyse says {^Poetry as a Fine Art, p. 51) in the idea of "invisibility or deep seclu- sion running through all." This statement of the implicit likeness is perhaps not quite adequate. The invisibility or deep seclusion of a fruitful, richly-diffusive presence seems to me to ap- proach the whole idea of the poet in its first unanalyzed state. — '69 — as pearls." Here the resemblance is particularized. That which at first constituted, in part at least, the identity of impression received from the two objects is now clearly perceived as a quality common to both. The hair of the Blessed Damosel, "yellow as ripe corn," has reached this stage of perception. It is not vaguely assigned to the com- prehensive class of things, making an impression of rich yellow color on the mind of the writer. Out of this class have already separated two distinct objects, the hair of the Blessed Damosel and ripe corn. They have drawn so far apart from each other that the writer cannot now say, though he might have done so an instant before, ' ' her hair was ripened corn." He sees now that the one is not the other but only resembles it. He might say that it is "like ripe corn" simply; but his thought-process has moved a step beyond this, and he sees in what respect the hair is like the corn. They are alike in color; both are yellow; so he says her " hair was yellow like ripe corn." The statement ' ' Her eyes were deeper than the depths Of waters stilled at even," reveals the same stage of metaphor-development. ' ' I wandered lonely as a cloud " shows not only that the two objects are separated from one another, and that a connection is observed between them, but that this connection is recognized as being a resemblance in one particular, that of isolation, not whiteness, not transitoriness, nor any other quality whatsoever. Browning's characterization of a "fruit-shaped, perfect chin" ' belongs to this family, as does Tennyson's "But bland the smile that like a wrinkling wind On glassy water drove his cheek in lines."* The shape constitutes the point of likeness in the one case; in the other the effect produced by each element in the figure unites them. We may, then, classify the developing metaphor as follows: 1. Radical metaphor, in which objects later recognized as two are represented by a single word or phrase which is equivalent to neither object, nor to the quality which they have in common, but to the one vague sensation or impression made by both objects upon the mind of the speaker, not yet defined or differentiated. Illustrations of this metaphor would be the term "cola"^ as applied by M. Taine's infant daughter to all sweetmeats, the word "moon," as including lamp-globe, the word "house" as used to designate a bird's nest,* etc. 2. Poetic metaphor, in which two objects or images are just be- ginning to disentangle themselves from this homogeneous sensation. Illustrations of this class are found in such examples as ' ' the hoarse wind,"* "a sorrow-clouded eye,"* "Sorrow darkens hamlet and hall,"* etc. ' A Face. 2 The Princess. »SeeCh. I. * Matthew Arnold, The Forsaken Merman. * Tennyson, Ode on the Death of the Duke of Weliington. — 40 — 3. Simile, including all cases in which the two constituents of the metaphor have so far separated themselves from the original homogene- ous sensation and from each other that they are recognized as two ob- jects slightly cohering by some resemblance or analogy. Of this there may be two species: (a) That simile in which the connection between the two objects is recognized by the writer only as a resemblance, the particular point of resemblance not being specified. "That face, like a silver wedge 'Mid the yellow wealth,"^ is an illustration, (b) That simile in which the resemblance between the two objects is limited to a particular quality or characteristic com- mon to both. This species may be illustrated by Keats's "jellies soother than the creamy curd" and "upon his knees he sank, pale as smooth-sculptured stone."^ V^^iOur conclusions, then, as to the relationship between metaphor and simile must be quite contrary to those commonly held by the rhet- oricians. The simile is not the earlier figure, transformed into metaphor by the simple device of cutting out the connective "as" or "like":'* but it is a stage later than metaphor in the process of developing a vague sensuous impression into the clear-cut judgment upon a given situation. The relation between these figures is more than merely verbal. It is a fundamental relationship of thought. Simile is a half-way house for the metaphor-process on its way to plain statement.* 1 Browning, Gold Hair. ^ Eve of St. Agnes. 'Precisely this statement is made by the following writers: ''A metaphor is a simile with the words like or as left out." — J. M. D. Meiklejohn, The English Language.^ p. 190. "The simile and the metaphor are essentially alike: and a metaphor can be made from any simile by omitting the word like or aj." — W. E. Mead, El. Comp. and Khec, p. 46. " A met- aphor differs from a simile in form only, not in substance." — Kames, Els. of Crit., Chap. XX, § 6. " The metaphor is a shorter simile." — Goldsmith, Essay (Unacknowledged) On the Use of Metaphors. The metaphor is an abridged simile." — D. J. Hill, Elements of Rhet- oric, p. 83. Metaphor is "an abridged comparison." — Blair. " Metaphora brevior est simili- tude." — Quintilian, Institutes, Bk. VIII, Ch. VI, § 8. " Eine abgekiirzte Vergleichung." — Brinkmann, Die Metaphern, p. 25. Metaphor is " no other in effect than a comparison in epi,- tome." — G. Campbell, Philos. of Rhet., Bk. Ill, Ch. II, Pt. I, "A metaphor is a simile in one word." — Lord Monboddo, Of the Origin and Progress of Language, Vol. VI, Bk. II, Ch. i, p. lOl. " A metaphor is a brief similitude contracted into a single word." — Cicero, De Oratore, Bk. Ill, Ch. XXXIX. * The foregoing account of the evolution from metaphor to plain statement does not con- flict necessarily with Mr. L. A. Sherman's observations as to the growing concentration of figures in our literature. Without endorsing his conclusions one can readily admit their consonance in all essentials with the theory advanced in these chapters. We have seen that metaphor, or, ac- cording to Mr. Sherman, the more " concentrated" form of the figure, represents a stage of per- ception antecedent to that expressed in the simile; Mr. Sherman urges that '' allegory or parable was first and succeeded by running metaphor: next clause metaphors, which were reduced to phrases, and phrases finally to compound or single terms." {Analytics of Literature, p. 79). These opinions seem at first quite incongruous; but they escape direct antagonism in two ways. First, Mr. Sherman's "concentration " is not always, though it is sometimes, a logical concentra- tion. The metaphor represents a thought less expanded, less developed than the parable, but a clause metaphor is not of necessity more expanded in thought than a phrase metaphor. In so far as Mr. Sherman's idea of concentration concerns expression only, we have nothing to do with it. In so far, however, as it is a matter of thought, it insists on being reckoned with. Upon this question it is somewhat difficult to come to terms with Mr. Sherman, inasmuch as the "con- —41 — We have now reached a point beyond which the development of the metaphor cannot go. When the two constituents of the fig- ure have been completely separated, their resemblance recognized as such and narrowly defined, no further separation can take place without a complete rending in twain, a disintegration of the metaphor. Let us suppose, for instance, that the series which we have traced as far as the expression "teeth white as pearl" be carried still further. The vague feeling that two sensations were identical has branched into a sense of two objects or images connected by resemblance, and this resemblance has been identified as a quality or characteristic common to both objects, — the quality of whiteness. This quality, now brought into clear con- sciousness, has freed the one object from the other. When one sees clearly why teeth are like pearls, he is not compelled to say they are like pearls in order to express the idea that they are white. He can say directly that they are white, a fact he did not conceive before. He knew centration'' of a metaphor on the thought side, evidently seems to him a matter of compression from some larger estate. His idea of the evolution of metaphor, considered as representative of thought, appears to be that metaphor, at first a large, well-defined, explicit comparison of two objects, became little by little reduced in bulk and obscured in details, limited to a comparison in one salient feature only, and that rather suggestive than overt. The coqception of this pro- cess which has been involved in the present discussion is far different. Metaphor in its first stage represents a vague, unorganized nucleus of a perception, which only slowly expands and defines itself into detailed comparison. Under this conception, the term " concentration " has little meaning. We may, p)erhaps, apply it to the first estate of metaphor as indicating its embryonic, comparatively undeveloped, unexpanded state; but doing so, we are brought face to face with the apparent contradiction between the two theories. Mr. Sherman would say that the development of the figure has been from parable (or simile) to metaphor. We must hold the logical order of de- velopment in a single mind to be from metaphor to parable (or simile). This brings us to our second means of escajje from a direct opposition to Mr. Sherman. The biologic theory of metaphor development urges that each metaphor in the mind of an indi- vidual passes through the stages indicated, becoming eventually a plain statement of fact. It does not, however, assert that for one hundred years after metaphor was invented, all metaphors passed through the first stage only; that they all reached the second stage in the second period of one hundred years, and that in the third period all metaphors were found in the third stage. It does not imply that radical metaphors were predominant in the world for a certain number of years, that then poetic metaphor became conspicuous, and that simile only was used after that time. This hypothesis is, perhaps, suflBciently discredited in the last pages of the present chap- ter. It will be necessary here only to state that the question of the variety of metaphor most in vogue at any given time is a question, not of the existence of the stage of thought-develop- ment which that variety represents, but of the rime of the thought's expression. In every per- ception the first embryonic stage exists, which, if then expressed, would be a radical metaphor; a more developed stage also exists, whicli must vent itself, if at all, in the poetic metaphor: the simile-stage and the plain-statement-stage also exist. The question is one simply of the time at which expression takes place. Now, inasmuch as in the most primitive times, fewer perceptions, in proportion to the whole number, have gone beyond the first, unorganized stage, the chances are greater for expression in radical mecaphor. Later, the more developed perceptions increase in number and the evolution from one stage to another has grown so rapid that the first stage may often be passed over uncon- sciously. Poetic metaphor might then be used and the earlier stage of thought disregarded, not coming to expression. So the simile might be used predominantly in writers so comparatively sophisticate as Homer and Vergil. So also is plain statement affected by many scientists and philosophers whose scorn for the figure would seem to indicate a belief on their part that they liad never seen in figurative guise these same things which they express literally. No doubt they believe they have never done so: but, as a matter of fact, every plain statement has passed through its metaphorical stage in their minds, though so swiftly, perhaps, that they are uncon- scious of the fact. A little introspection is necessary, in minds where the cerebral processes are practised and therefore somewhat rapid, to bring the earlier stages to light, to delay the evolution for a moment. — 42 — first only that somehow the two things made the same impression on him. Later he felt that they were alike in some way; but it is only when their alikeness has defined itself in his mind as the quality white- ness that he can be independent of pearls and say simply that certain teeth are white. The metaphor process is, then, a mode of getting at abstract ideas, and only when it is complete can one use "plain state- ments. '" Those savage tribes who have no words for ' ' round, " ' ' hard, " and such abstract qualities, but must say "like the moon," " Hke a stone," etc. ,^ are still in the midst of the process, as are those children who instead of calling a silver dollar large and a dime small, call the one " mamma" and the other " baby."^ It is only through the meta- phor-process that abstract ideas come into existence.* Only across this or to call it back, that one may become conscious of its several steps. This habit of introspec- tion the man of science lacks, being intent more upon results than upon processes and usually uninterested in psychologic analysis. He has also become accustomed to balance considerations long and carefully, to weigh all his facts, and, in fact, to think his thought clear to the end, before expressing it, so that the state of plain statement, representing a perfectly developed and finished perception, is the one most commonly represented in his speech or writing. All antecedent stages are partly unconscious, partly unvalued. The modern poet, however, is both introspective and accustomed to the expression of an unfinished thought. He is interested in perception for its own sake, not alone as leading to cer- tain conclusions regarding the object he sees. He is thus accustomed to note his own impressions with peculiar vividness, to call them up over and over, until he is familiar with every shifting phase of them, so that he would naturally remark an early and obscure stage in their develop- ment, and would often arrest the process at that point for expression. More and more this ten- dency to express a growing perception in its earlier stages would increase, as the taste of an age turned to suggestion rather than definition, to stimulus rather than to exact formulation in liter- ature. And thus would become possible the modern tendency which Mr. Sherman alleges toward the predominant use of the metaphor in modern poetry. I may, perhaps, say that my own read- ing would confirm Mr. Sherman's observation. I have left the point open not through unbelief, but because its acceptance demands exact proofs in the nature of numerical computations which are not in my possession. One further consideration of Mr. Sherman's interesting statement may be added. If it be true, in general, that metaphor is the figure characteristic of the romantic school in literature, as simile of the classical (using these terms in the widest sense as indicating the two perennially opposing forces in all art), one source of this affinity may be found in the fact that romanticism is subjective, introspective and self-conscious as against the objective straightforwardness of classi- cism. Thus the former tendency would seem to favor metaphor in so far as that depends upon a detention in the consciousness of early stages in perception likely otherwise to slip rapidly through the mind without coming to expression. * " Wie im Schreiben Bilderschrift f ruber war, als Buchstabenschrift, so war im Sprachen. die Metapher das friihere Wort, welches sich erst allmahlich zum eigentlichen Ausdruck ent- farbte." — J. P. Richter, Vorschule der Aesthetik. " Die Ursprache ist noch unentfarbte Bildlichkeit. In dieser Zeit gibt es noch keine Prosa, weil jedes Wort schon durch seine Wurzel und seine Zusammensetzung einen poetischen Ein- druck erregt, weil jede Anschauung schon ein Gedanke, jede Bezeichnung ein Versuchzudichten ist." — Bauer, Das Bild in der Sprache, I, p. 9. 2 See note i, Ch. I. * Cf. Sully, Studies of Childhood, pp. 163-4. * George Eliot's observation, made often by others, that rigidly plain statement is all but impossible, tends to confirm this view, as does also the universally noted phenomenon that meta- phorical speech is far commoner than plain, abstract language among people of undeveloped in- telligence, such as savages and children. " O, Aristotle! if you had the advantage of being ' the freshest modern instead of the greatest ancient,' would you not have mingled your praise of metaphorical speech as a sign of high intelligence, with a lamentation that intelligence so rarely shows itself in speech without metaphor — that we can so seldom declare what a thing is, except by saying it is something else ?" — Mill on the Floss, Bk. II, Ch. I. — 43 — bridge do we arrive at a plain statement, that is an abstract statement, of fact. ' Having once completed this process and defined the abstract idea of white or yellow out of the hazy impression of white or yellow objects, one is thenceforth free to use those terms, directly, in the case of these objects. He can say "Her teeth are white" or "very white." The hair of the Blessed Damosel may be affirmed to be of a rich yellow color, her eyes to be exceedingly deep. It is not now possible, in the case of these perceptions, to express them in their half-differentiate form, as was done before the common quality of whiteness and yellowness became ex- plicit in the consciousness of the speaker.^ Notice I said "in the case of these perceptions," it is not possible to express them metaphorically after the common characteristic has defined itself. It would be mani- festly absurd to say that as soon as yellow had come to consciousness as the common characteristic between hair and ripe corn, the cause of their making the same impression on the senses, it could not again be born into the mind through the same metaphorical process, though in a differ- ent situation, in connection with different objects. Thi> process would doubtless be greatly shortened the second time, still further abridged the third, and the fourth, but up to the nth instance would certainly occur. More and more subconscious it must also become, so that the practical, non-introspective intelligence often passes it quite unawares. But a close examination of one's mental processes brings this one to light. I make this statement to forestall the objection which is altogether likely to arise in many minds, tenacious of the dignity of metaphor and of the poetry of which it is the core, — the objection that, if metaphor represents a less highly developed, a less mature stage of perception, it must be less honorable than plain statement, and, further, it must cease from the earth after a time, when the intellectual faculties of men have fully ripened. As to the first half of this objection, it may be answered that a question of honor is not involved, but a question of function. If the function of the pioneer be less honorable than that of the citizen in a long established community, the original investigator in science less hon- orable than that of the teacher of accredited theories, the creator in lit- erature less honorable than the critic, then the function of the metaphor may be considered less honorable than that of the plain statement. But 1 When a metaphor has been transformed into plain statement, it is commonly said to have died. This term is perhaps appropriate in so far as death is taken to be only a change of form, but it is inapt so far as it implies the negation of life, of growth. Metaphor ceases to be metaphor, it is true, but only because it becomes something else, because it has reached a fur- ther stage of development. Whenever in this chapter, therefore, the "death " of metaphor is spoken of, the term must be understood as carrying with it an idea of fulfillment and progression rather than that of finality or decay. * It may be, perhaps, that the older form of speech will continue to be used, but in a sense perfectly literal. "Corn-colored hair" may, for instance, come to mean little more than yellow hair; the term "rosy cheeks," in our common speech seldom bears any reference either open or concealed to a red rose; "neat as wax," "sweet as a pink," "brown as a berry," " warm as toast '•' are but traditional equivalents for a sup>erlative degree of neatness, sweetness, brown col- oring and warmth. — 44 — it is useless to dispute over the relative values of function. Each is as essential, as dignified and as honorable as any other. The metaphor has perhaps the harder and the rougher task. It is no dainty, useless "ornament" of speech, as was long taught among us, but the first crude, indispensable stage of the growing thought. Plain speech, as a later development of thought, may show a finer finish, but cannot be more powerful or more serviceable. The notion that metaphor must ultimately be superseded by plain statement springs from a hmited view of the resources of nature and of the human mind. A new situation is necessary for metaphor, a sensa- tion as yet undifferentiated; but new situations and new sensations are presented to every individual every day of his life. He may look at the sea for the hundredth time, but at this hundredth time the situation is essentially different from the occasion of the ninety-ninth seeing. Doubtless some quality of light upon the waves, some feeling in the air, some fineness in the foam, some cadence in the roar of the waves is different to-day from what it has ever been before. The observer him- self is different. His senses are sharper from the experience of the ninety-nine times he has stood here before. His attention is arrested by a certain feature in the scene because some previous event has forced it upon his notice. The situation is one never presented to him before. If his senses are but half aroused, it may seem to him like the same old situation. But the alert, receptive, impressionable mind of a poet per- ceives it to be essentially different. Witness, for instance, the various epithets applied by Shakespeare to a single object, the sea. Once it is the "silver sea,"^ again "the furrowed sea,"^ "the deep mouthed sea,"' "the pretty-vaulting sea,"* "the imperious sea,"* "the vexed sea,"' "the wayward seas,"' "the multitudinous seas,"^ "a wilderness of sea."» And still Shakespeare views it under aspects other than these. He speaks of "the rude sea's surged and foamy mouth,"'" and of "the very hem o' the sea."" Bolingbroke and Norfolk are "in rage deaf as the sea,"'^ and when Orpheus played ' ' Even the billows of the sea Hung their heads."'* All these are essentially new situations for the mind of Shakes- peare. The object itself may be familiar, but the man's view of it is fresh. 1 Richard IT, Act II, Scene I, 1. 46. * Henry F, III, Prol. 12. ' Henry V, Prol. 11. * 2 Henry VI, Act III, Sc. II, 1. 94. " Cymbeline, Act IV, Sc. II, 1. 35. « Lear, Act IV, Sc. IV, 1. 2. T Pericles, Act IV, Sc. IV, 1. 10. 8 Macbeth, Act 11, Sc. II, 1. 62. * Titus Andronicus, Act III, Sc. I, 1. 68. »« Twelfth Night, Act V, Sc. I, 1. 81. " Timon of Athens, Act V, Sc. IV, 1. 66. i« Richard II, Act I, Sc. I, 1. 19. " Henry VIII, Act III, Sc. I. 10. — 45 — We need not, then, from the statement that metaphor depends upon a new situation, conclude that all situations will in time become so far analyzed and defined that vital, natural figures will be impossible. Such a view indicates limited imagination and narrow faith. It is psychologically true that the analysis of any situation deepens and changes not only that situation but its observer. His eyes are opened to see new objects, or old ones in a new light. He is a different man and his world is new. There is thus no limit to the new situations of which our expanding universe and our expanding selves are capable. Metaphor, while a stage in the perceptive process which must always be superseded by plain statement, must as certainly recur in a new perceptive process. Though one metaphor may die into abstract speech, another rises out of the very extension and complication of experience which the former process of growth and death has afforded. To paraphrase Swinburne's assertion, "Metaphors perish, but metaphor shall endure." 1 'Men perish, but man shall endure; lives die, but the life is not dead." — Hymn of Man CHAPTER IV. THE ESTHETICS OF METAPHOR. Hitherto we have discussed the metaphor mainly from the point of view of its maker. In this chapter we shall consider its effect upon the hearer or reader. The effect of metaphor upon the reader has by no means passed unnoticed among rhetoricians. In fact, as we have seen in a preceding chapter/ this effect is so conspicuous that the desire to produce it has been assigned by writers upon the subject as the all- sufficient cause of metaphor. Metaphor takes its rise, these writers have said, from the desire of the speaker to produce a certain effect upon the hearer. This theory we have discarded,' but with it we have not laid aside the fact that metaphor does produce a very definite reac- tion upon the mind of the hearer or reader. This fact is, indeed, quite indubitable. Rhetoricians may differ more or less widely as to its exact nature, but they all agree that an effect of some sort is produced. What, then, is the effect of metaphor upon the mind of the reader.? The answers of the rhetoricians can, I think, be grouped into three classes. Metaphor gives the reader pleasure;^ it economizes,' or it stimulates,* his mental energies. These three typical statements cover, so far as I am aware, all the answers made by the writers of rhetorics to the question — "What effect has metaphor upon the reader.?" Let us see, in the first place, whether this list of effects produced by the metaphor cannot be still further reduced. In Mr. A. S. Hill's statement of his principle of stimulus we have a suggestion toward this end. Spencer would persuade us that, in order to grasp the idea in- volved, metaphor demands less work from the mind than does plain statement. In other words, metaphor conduces to clearness or sim- plicity of expression so that it is easier than plain statement to understand. This idea of the metaphor as a means to clearness is one which, since the classical writers, has been practically abandoned by rhetori- cians.^ Mr. E. E. Hale, Jr., states in its extreme form the idea that metaphor effects not clearness but vigor of statement. "Metaphor," he says, "is in its very nature a confusion. . . . On the whole the metaphor is opposed to clear thinking. . . . But .... if metaphor ' Ch. II. 2 Aristotle, Rhetoric, Bk. Ill, ch. II; Cicero, De Oratore, Bk. Ill, Chs. XXXVIII and XXXIX; Kames, Els. of Crit., Ch. XX, § 6. ' Spencer, Philos. of Style, Scott's ed., § IV. Also D. J. Hill, Science of Rhetoric, p. 217. * A. S. Hill, Prtns. of Rhet., ed. 1895, p. 119. ^ I notice only two after Aristotle who treat of metaphor as primarily a means to clearness. These are J. G. R. McElroy [Structure of English Prose, p. 245) and Arlo Bates {Talks on Writing English, p. loi). Bain, Tompkins and De Mille regard metaphor as making for both clearness and force, Campbell, Whately, Hart, Cairns, Day, Wendell, A. S. Hill, Gummere, as contributing wholly to force. —47— does not tend toward clearness, it does give us liveliness, brilliancy, energy. . . . " ' This conception of the metaphor has come to prevail among rhetoricians, that it does not serve to make an idea so plain and clear that the mind has no difficulty in grasping it; but rather that it pre- sents certain difficulties to comprehension, but so suggestively or forcibly that the mind is enabled to over-ride them. This is precisely the view of Mr. A. S. Hill. "The metaphor," he says, "though shorter than the simile, does not achieve a 'great economy' in mental effort. It usually demands more mental effort, but it enables us to make the effort with greater ease. We are * gratified ' but we are also stimulated."' In short, metaphor does not render mental exertion less necessary than does plain statement. It rather stimulates the mind to do more work with less conscious effort, and thus decreases, not the whole amount of energy put forth by the mind, but such part of that energy as may be conscious, voluntar)' or forced. In such case it may indeed seem to the reader that he has put forth less energy, because all that he has exerted has gone out spontaneously; as in the oft-quoted case of the small boy whose fatigue is too great to admit of an errand to town but does not for a moment stand in the way of a ball game. In each instance, that of the ball game as that of the metaphor, there is truly economy of energy, but only of that energy which is forced, conscious, frictioning. Of the energy that springs forth spontaneously in answer to adequate stimulus there is extreme prodigality. So, at least, would Mr, Hill's dictum imply. If we adopt his hypothesis, that metaphor stimulates the mind of the reader, we need not hesitate to grant that the conscious, deliberate mental activity required to grasp an idea will be considerably diminished by it. This is simply the re-statement of an admitted principle of psy- chology-, that to produce a given action, the weaker the stimulus, the greater, in general, must be the conscious effort of the actor; the stronger the stimulus, the smaller the voluntary, self-determined element in his response.' We react to a sharp stimulus unconsciously. We need not try to see a flash of lightning, but a page of fine print in a waning light strains our eyes. We can readily understand how, if the metaphor stimulates the mind strongly it decreases the conscious effort demanded to grasp the idea involved, and thus, in a sense more limited than Spencer saw, economizes mental energy. We have, then, reduced to two our statements of the effect pro- duced by metaphor upon the reader. His mind is stimulated, or it is gratified. It is quite possible that eventually we shall find a connection between these two remaining effects so intimate that we are justified in considering them as aspects of the same general phenomenon. For 1 Constructive Rhetoric, p. 265. Cf. Marie Bashkirtseff: "If I make nse of a few figures jjl of speech, do not think it is for ornament; oh, no! it is simply for the purpose of describing as ^ nearly as possible the confusion of my thoughts." — -Journal, July 18, 1877. » Principles of Rhetoric, ed. 1895, p. 119. » Sec James, Psychology, Briefer Course, Ch. Xm, Voluntary Attention, p. 224. — 48 — the present, however, we shall set about investigating each effect by itself. First let us inquire as to the fact. Does metaphor stimulate the mind of the reader.? In order to answer this question adequately, it is necessary to trace somewhat in detail the actual operation of the meta- phor as read. Let us, then, see what takes place in the mind of a per- son who hears or reads a metaphor. The radical metaphor, before it has been recognized as metaphor, of course falls outside of our inquiry. Two savages, or two children of the same age, who say to one another "atta"^ meaning disappearance, when a person leaves the room, when the sun sets and when a light is extinguished, at each time of saying it simply set up in the mind of the hearer a hazy consciousness of some situation familiar as a whole which at this time varies from its usual form in having dropped out one of its elements. The situation, then, is one vaguely felt as disappearance or loss. The element which has dropped out of its usual environment is not sharply defined, nor is the way in which it has disappeared clearly evident. Any disappearance of anything in any way may constitute the image expressed by the one speaker by the word "atta" and called up in the mind of the other. To hearer as to speaker, "atta" means the situation when the sun sets as well as when a person leaves the room. The two perceptions are not distinguished either by hearer or by speaker. But suppose one of the children, or one of the savages, to discrim- inate between the manner in which a companion disappears from the room and that in which the sun sinks from sight. Then, when "atta" is said in his hearing, the vague, little-defined image of a disappearance will first rise before his eyes, this breaking rapidly into two images, one of a person walking out of a room, the other of the sun setting. These images are the elements once latent in the undefined image of a disap- pearance. Now they have sprung out of it and the mind of the hearer wavers between them for an instant. But suppose the speaker in this case meant the setting of the sun. The hearer following the speaker's eyes, would meet the situation in which the sun was ordinarily an ele- ment, but from which it had vanished. Instantly the image of the per- son walking from the room would take a subordinate place in the con- sciousness of the hearer, and the image of the disappearing sun receive the ictus of attention. After it had frequently happened that, the word ' ' atta " called up in the mind of the hearer two somewhat distinct images, the speaker's meaning being sometimes seen to be one, sometimes the other, the hearer would finally come to see that for the speaker "atta" meant neither the one image nor the other, but the vaguer, less-defined situa- tion of disappearance or loss, out of which the two more specific images had developed. The relation of these two specific images to the 1 See Preyer, The Mind of the Child, Pt. II, p. 150. — 49 — original vaguer image and thus to each other would be recognized, at first dimly, then more clearly, until it might even appear as resem- blance. When this recognition had taken place, the hearer could see how the speaker came to use the "radical metaphor," could retrace the steps in his own mental process, back to the undifferentiate image which still persisted in the speaker's mind, could see it branch in- to two distinct elements as his perception clarified, and could even go further, to the explicit recognition of the relation of resemblance between the two images. The person hearing the radical metaphor, having himself already emerged from the homogeneous stage of perception for the situations in- volved, has, it is plain, experienced the metaphor presented to him, working back -to its first undifferentiated state and then following its development into plain statement. So much for the reception of the radical metaphor. The next period of development for the speaker's thought is that in which the situation has begun to divide. The child Ilda says ' ' Mamma- ba," pointing to the sheep in the picture, "Ilda-ba," indicating the lambs.* The aboriginal, undeveloped sense of "large" or "grown-up" has divided into "Mamma" on the one side and sheep, or "ba," on the other; that of "small" or "baby" into Ilda on the one hand and sheep on the other. Now what process is set up in the mind of the hearer by this metaphor.'' At first, it is evident, the two images must be flashed almost simultaneously into the hearer's consciousness. They struggle for an instant; then the two images are quickly traced back to their root and made one in the vague original sense of "large" or "grown-up." Then it may be that the speaker's mental process is rap- idly reproduced in the hearer's mind. The vague original feeling of "large" or "grown-up" swiftly falls apart into the two branches " Mamma" and "sheep," and unites itself again in a clear perception of the likeness between them." One's perception of a metaphor chanced upon in reading will be found to follow this same general programme. We see on the page before us Lowell's "When grass-blades stiffen with red battledew."^ The line runs easily enough until the words "red battle-dew" are reached. Something red that stiffens the grass blades in time of battle is plainly enough blood. But no sooner has this image darted into the mind than it is opposed by another, that of ' • dew " which the last word in the line has aroused. What have these two to do with one an- other.? What is there in common between blood and dew.? From one image to another the mind vibrates, unable to rest in either or to resolve 1 Romanes, Mental Evolution in Man, p. 283. 2 It need not be insisted upon that always this reconstruction of the speaker's mental process carries itself in the mind of the hearer to a completion unattained by the speaker; but always if the idea in the speaker's mind be grasped by the hearer, the two elements named by the speaker must be traced back to their source in the speaker's vague perception of an undefined situation, in this case the nebulous sense of " large " or " small." Washers of the Shroud. ^^X^^ ■"l^p! T^ OF THH r >. :/ — 50- the two into one. But this conflict is not for long. The two elements are rapidly traced back to the original undefined sensation of some liquid dabbling the grass, thence differentiated into the red stiffening blood and the white limpid dew, and unified once again, but more completely, in the explicit perception that the blood of a battle-field is like a red dew spattering the grass. For a second illustration of the same process let us take Shelley's "The moon rains out her beams.'" The image of down-pelting rain is swiftly superseded by that of softly diffused moonbeams. At first the two oppose each other sharply, but the next instant they have merged themselves in a cloudy vision of something descending from the skies in a great flooding volume, this situation separating itself again into the rain drops and the moonbeams and becoming once more united in a clear recognition of the likeness between the two. Swinburne's "all their green-haired waters" ^ jars sharply at first, until the dissonant elements, hair and grass, have united in the obscure perception of fine long filaments, out of which they originally sprung. Then this perception may divide and come together as it will, each return to the fundamental unity establishing in the reader's mind a sharper, more definite idea of the relation between the two elements concerned. It is impossible to adduce direct evidence in support of the state- ment that this process, or one similar to it, takes place in the mind of the person who reads or hears a metaphor. One's own subjective expe- riences must constitute for each reader the chief witnesses; and these unfortunately are often impeachable. Indirect testimony, however, ap- pears in the account given by all rhetoricians of the process of metaphor- making as one of taking two objects or images and finding a connection between them. Why should all the philosophizers upon this subject have hit upon just this explanation, if it be not true that their theory describes the mental process of which each has been dimly conscious in reading a metaphor.? Nothing could be more natural than to conclude that because the reader must establish connection between two elements in the metaphor, the writer had originally to do the same thing. And the fact that the rhetoricians almost without exception affirm such a process of the metaphor-maker seems to indicate at least a half-con- scious recognition of the same, process carried on b}' themselves as readers. With this indirect evidence, and the directer, though more uncer- tain, testimony of introspection, we must consider the point established that the reader of a metaphor first vibrates between the two disconnected images aroused by the figure, next perceives vaguely the primeval unity out of which they both had emerged and, having secured this, may re- peat the original process by which the metaphor grew in the mind of the writer. Having granted the description, however, we have still to » 7'Ag Skylark. * Atalanta in Calydon. — 51 — answer the question: Does metaphor in any peculiar sense stimulate the mind of the reader? That is, specifically, does it stimulate his mind to a greater degree than does plain statement? If the foregoing description be accepted as a true account of the i process set up in the mind of any person who reads a metaphor, we I must grant to the figure a considerable stimulating power. The focus f\ of attention is first driven about helplessly from one image to the other. ■! It cannot rest upon either and the mind is consequently forced somehow 'i to unify the two, to trace back the opposing images to their common 'i\ source, to reproduce the writer's original perception, in all its teeming ij vagueness. The shock of the two incongruous images in such close juxtaposition acts as a powerful stimulus upon the mind. It is at least dimly conscious that these images were not incongruous, not unrelated, to the writer who put them side by side. To his mind there was a meaning in their association. But what was it? There is the zest of an enigma or a conundrum in the inquiry.' The puzzle must be solved. As often as the reader recurs to it, he is fascinated, unsatisfied. Each side of the paradox in turn holds possession of his mind. Then, with a flash of insight, the solution bursts upon him. The two incompatible images blend into one and the mind is at rest. Such is the story, when, for instance, one reads of Swinburne's Atalanta that she was " a maiden clean, Pure iron, fashioned for a sword. "^ If she had been only a stern, chaste, unswerving, unyielding, single- purposed maiden, one could read her languidly, grasping the idea of her character with little effort and with as little vividness. Each of these characteristics, we must conceive, had long ago gained a relatively inde- pendent existence in our consciousness. Each had been born of meta- phor, though now forgetful of it.^ The metaphor-process had taken place so often, the resolution to plain statement been made so frequently, that at last the dualistic stage had come to be passed over rapidly and all but unconsciously. The idea "chaste" had, for instance, grown out of the first dim perception of something spotless, unsullied. A sheet of new-fallen snow, a block of flawless ice, the character of a pure maiden, made vaguely the same impression upon our minds. This inchoate per- ception, however, soon branched doubly into, say, the ice on the one side, the maiden on the other. And finally, after the first hazy sensa- tion had many times separated itself into these two constituents, the connection of the two became evident as the point of resemblance, which we were thereupon able to name purity or chastity. Now it is probable that the word "chaste," whenever heard, sets up in the consciousness some such process as that by which it originally 1 Cf . Aristotle's statement that style is "a riddle if it consists of -metaphors." — Poeticst Ch. XXI. * Atalanta in Calydon. » S«e Ch. III> — 52 — came into being. This process, however, has become so famihar, so automatic, almost, that it takes place subconsciously. It is as when we transfer sensations from the field of sight into that of touch and say, seeing a certain fabric, — "Its surface is rough." The complex process by which we translate the sensations of sight into the judgment that the surface is rough goes on in our minds unconsciously, so accustomed is it. We do not note its stages. The result is all that claims our attention. And so it is with the metaphor-process which underlies our perceptive realization of each "plain statement" presented to us. It goes on un- noticed because so rapid and so familiar. But it is nevertheless a real experience continually passing through our minds. We have seen how it may be excited by the literal statement corre- sponding to Swinburne's figurative description of Atalanta. And we have also noted the mental process set up by the metaphor, ' ' pure iron, fashioned for a sword." We are ready now to inquire "What is the essential difference between the two.-*" Just this. The plain statement started a reaction that has become habitual in the mind, leading to the perception of a certain abstract j^ characteristic, say that of high severity. The metaphor conducts us to the same goal precisely, but over the road on which the writer had started, a road very possibly unfamiliar to the reader and certainly not chosen by him. Plain statement asks the reader only to arrive, by his own preferred route, at a certain idea. Metaphor demands more — that the reader come to the writer's completed idea and in the writer's own way. Since we can measure a stimulus only by the force of its reaction, we must adopt this method for determining the stimulating power exerted by the metaphor in comparison with plain statement. It is evident, from the foregoing description, that the force brought by metaphor to bear upon the reader's mind is considerably greater than that put forth by plain statement. While the latter avails only to propel the reader's thought along an accustomed and prefeirred channel, metaphor forces it to fall in with that of the writer, to trace the writer's branching idea back to its source and then to follow its ramifications beyond the point of actual expression, to traverse a road that may be wholly new, a coun- try hitherto unseen. We must grant, then, that metaphor exerts an influence upon the mind of the reader far more powerfully stimulating than that of plain statement. But we have still to pursue our second inquiry as to the pleasure-giving effect of metaphor. Let us carry on this inquiry quite independently, returning in due time to the point just established. To the question ' ' Is metaphor pleasurable to the reader?" but one answer is possible. Upon this point all testimony agrees. Whether the effect of metaphor be considered primarily to economize or to stim- ulate mental energy, every writer upon the subject, either implicitly or explicitly maintains its ultimate result to be that of pleasure-giving. Upon the fact, then, that metaphor is pleasurable, we need not pause, — 53 — since it is acknowledged by all writers, and attested by universal experience. So firmly, in fact, has the idea of metaphor as a pleasure -giving element in literature, taken root in our common con- sciousness, that we make this characteristic a standard for judging the individual metaphor. We know of no other way of pronouncing for or against a metaphor than by saying that it is pleasing or displeasing to us. If it be displeasing we are certain that there is something abnor- mal about it, something pathological, in fact. If pleasing there is noth- ing more to be said. The metaphor is a true metaphor. It has justified itself. So strong as this is our conviction that an essential characteristic of metaphor is its pleasure-giving effect. There is, we have said, among writers a general agreement upon the fact that metaphor gives pleasure to the reader. But there is no such agreement as to the reason why it does so. In fact there is no as- pect of the whole subject concerning which such diversity of opinion obtains. Cicero alone has four different hypotheses: i. "Because it is some manifestation of wit to jump over such expressions as lie before you, and catch at others from a greater distance." 2. "Because he who listens is led another way in thought and yet does not wander from the subject." 3. " Because a subject and entire comparison is despatched in a single word." 4. "Because every metaphor .... is directed immediately to our senses and principally to the sense of sight."' These by no means exhaust the list, but they may serve as a starting point for our discussion, being fairly typical of all the reasons assigned. We shall take them up in order of importance, beginning with that es- sentially superficial statement which traces the pleasurable effect of metaphor to its brevity.^ It is quite evident that we shall gain here no thorough-going solution of our problem. In the first place, brevity is not an invariable nor an essential characteristic of metaphor. It is true that metaphor is, as Spencer indicates, usually shorter than simile; but even this is not always true; and it may very often employ a larger number of words than does the corresponding literal statement. Thus even if it were possible to account for the pleasurable effect of some metaphors by their concise form of expression, this explanation would not cover all cases. And, in the second place, brevity as brevity is neither pleasing nor displeasing. A brief form of expression may sometimes be pleasurable, but it is not necessarily so. One could hardly maintain that "No," " Fire !" "Stand off !" although concise statements of an idea, are neces- sarily pleasing to the hearer. It is not brevity as such which in any case gives pleasure, but some deeper-lying characteristic, which Cicero made no attempt to discover. Spencer, however, seems at least dimly conscious that a small number of words alone (though formally he makes this the condition of pleasure), is not sufficient to account for the effect produced by meta- » De Oratore, Bk. HI, Ch. XL. ^Cicero's third reason above. See also Spencer, Philosophy of Style, Scott's ed., p. 25. — 54 — phor/i ' '^"i e*^pression of the same idea, he would say, is more pleasur- able iV'^ A t^an if expanded. In other words, the compression of the idea g*P S pleasure. That this is often true one would not wish to deny. Whether it is always or inherently true is a question not so easily set- tled. Spencer, however, assuming the fact, sets about its explanation. The agreeable effect produced by the conciseness of metaphor he refers to the economy of ment; effort involved. This explanation at once discredits itself. As a m? er of fact, economy of mental energy is not always subserved by'^ d number of words. A few more words, on the contrary, may (. it .Icar up a statement so that less effort is de- manded for its c(^n{\) sion. We cannot, then, admit Spencer's law that the fewer t^'-'^v used in expressing a given idea the easier that idea is to be und^-% 'and thus the greater pleasure is secured to the reader. We maf/ v^ever, admit to consideration such a modified statement as this, .i^t a conciseness of expression in the metaphor which does conduce to the economy of forced mental effort on the part of the reader may be pleasurable. But this statement, substituting as in effect it does, the principle of stimulus for that of economy ' brings us to a different series of questions. Does the conciseness of the metaphor conduce to its stimulating effect? And is this stimulating effect the source of the pleasure derived.-* The first inquiry is readily answered. The brevity of the meta- phor is stimulating, but not as brevity. ^ Reverting to the foregoing account of the process set up by a metaphor in the reader's mind, we see that this process is induced not by the shortness of the metaphorical expression but by its incompleteness, its fragmentary and suggestive character. It takes no fewer words to speak of ' ' beaded bubbles " as "winking at the brim'"* of the beaker than to say that they were burst- ing there, or breaking or collapsing. But in the second instance, the thought is expressed in its completed state. The rapid extinction of a globular, glistening object was the situation at first presented to the writers mind. This differentiated itself rapidly into the sudden disap- pearance of a bubble by bursting and that of an eye by the descent of the lid over it as in winking. Just at this stage of the thought's develop- ment expression took place and the bubbles were consequently said to be "winking at the brim." The reader into whose mind has been flashed abruptly these two at first irreconcilable images, bursting bub- bles and winking eyes, is for an instant torn between them. What did the writer mean.'* What was in his mind.-* The answer dawns upon him in the composite photograph made by the rapid alternate presentations of the two images — the shadowy picture of a globular glistening object suddenly disappearing. The reader was compelled to this resolution of the two images into their primeval substance, not because the phrase 1 We have seen in preceding paragraphs that this understanding of the term economy is alone capable of sustaining Spencer's theory. 'Cf. Sherman, Analytics of Literature, p. 67. * Keats, Oiie to a Nightingali;, — 55 — was short, but because it represented a state of mind iny, n\ sible unless the antecedent state were known. The idea pre; en as so fragmentary that it remained meaningless until traced to its soi e, yet so pregnantly suggestive of meaning that the reader was all but co.iipelled to trace it. The incomplete, the fragmentary, the -j-^nigmatic character of the metaphor, then, not its brevity J>er se, is '. ;^ cause of its stimulating effect. The second question, whether the st ^ulus given by this figure is of necessity pleasurable, we shall discuss lai jr .vronnmtion \vith some other explanations which lead us to the sam^e. ^ v.^i/ The implication in Cicero's fourth hypotbf yj-Q-jt the pleasure of metaphor arises largely or wholly from the f aci - ^^ ges are thereb)' aroused in the mind has been made as well by m jj /slern writers.' It >l^ seems to be assumed by these rhetoricians ttiy^,, ■< mind can think either in images or in abstractions, and that, for sq e reason, it prefers the images, so that metaphor, which induces it to think in images is more pleasurable than plain statement. The fallacy of this assumption has, I think, become evident in our previous consideration of the meta- phor-process. It is impossible to choose between thinking in images and thinking in abstractions. One must always do both — the one as an earlier, the other as a later stage in the development of the same idea. And that the one stage is inherently any more pleasing to the mind than is the other may well be questioned. To have an idea presented to the mind at the earlier stage of its development may, however, be more pleasing than to receive it later. And this possibly is what the writers who held to the theory of the mind's pleasure in images really mean. In this case, we are brought back once more for explanation to our description of the effect of metaphor on the reader. The essential difference, as we have seen, between the presentation of an idea in the metaphorical and in the prosaic or literal stage, lies in the fact that in the first case the images presented to the mind are more vivid than when plain statement is offered, and the process of their development is likely to be somewhat unfamiliar. In other words the mind is led to act more vigorously than usual and to deal with new mate- rial. Thus we find ourselves once more at the theory of stimulation as the effect produced by the metaphor, and our way is again blocked by the question — Is this stimulus of the mind on the part of metaphor neces- sarily pleasurable.? Let us turn, however, to Cicero's first theory. As stated, it seems to imply that the novelty or remoteness of the images aroused by the metaphor accounts for the pleasure derived. It is more than probable that this conception was obtained from Aristotle, who notes " an air of strangeness " ^ as conveyed by metaphor, having previously stated that 1 Campbell, P/ii/os. of Rhet., Bk. Ill, Ch. I, § r. Bascom and Morgan, Philos. of Rhet., p. 244. » Rhetoric, Bk. Ill, Ch. II. — 56 — "it is proper to invest the language with a foreign air, as we all admire anything which is out of the way, and there is a certain pleasure in the object of wonder." This theory has not found favor with the modern rhetoricians; but it deserves recognition, as bearing some relation to the hypothesis last discussed. The air of strangeness, observed by Aristotle, is not, as Cicero seems to think, derived from the remoteness of the images concerned, but rather from the novelty of their origin and relationship. This is but saying once more that the mind traverses new country under the spur of metaphor. But just why this should be pleasurable belongs to the in- quiry which has been deferred. Cicero's second theory, which alone is left for our consideration, seems to hint at a fundamental aesthetic principle — that of unity in var- iety. This principle has been implied in more than one of the later explanations of the pleasure derived from metaphor. We find it in Whately's notion, borrowed, as he says, from Dr. A. Smith, that "the more remote and unlike in themselves any two objects are, the more is the mind .... gratified by the perception of some point in which they agree."' Kames seems to imply the same idea in his comparison of a figure which has two constituents to "concordant sounds in music. "^ Bain says that an " emotional similitude," under which head falls meta- phor, imparts to the reader ' ' a shock of agreeable surprise " when the comparison possesses both novelty and remoteness.^ "In the assimila- ting operation of the intellect," he says, "whereby comparisons are brought from very remote sources, there often results a feeling of unex- pectedness, which is in itself an agreeable effect.* In these statements, Bain emphasizes the idea of the metaphor as stimulating the mind, while retaining the conception of unity in variety as the source of the pleasure afforded. He might perhaps agree to the following as a comprehensive statement of his position: Metaphor gives pleasure to the reader by stimulating him to see a unity in apparent diversity. At any rate, whether such a statement as this represents Bain's po- sition or not, we are driven to consider it as the only possible interpre- tation of the unity- in-variety hypothesis. Unity in variety is not pleasurable unless it acts upon the observer, unless it affects him posi- tively. If the metaphor makes the reader see or experience unity in variety it is pleasurable. Otherwise it may possess all the unified variety in the world, but remain inert; incapable of producing either pleasure or pain. We are now ready to consider in detail the relation of the stimulus given by the metaphor to the pleasure afforded by this figure. We have shown that the pleasure-giving effect of metaphor is not due to its brev- ity, as some have thought, nor to the fact that it brings concrete images i 1 /i/s. of Khet., Pt. Ill, Ch. II, § 3. a Els. of Crit., Ch. XX, § 7. •'' Jing. Comp. and K/ift., Tt. I, p. 146. * Eng. Comp. and Rhet., Pt. I, p. 145. -57 — before the mind, nor directly to its leading the mind over a new track. These circumstances may indeed conduce to the stimulating effect of the metaphor, but not necessarily to its pleasing quality, unless, indeed, stimulus is of necessity pleasurable. Such a proposition as this cannot, of course, be for an instant allowed. Stimulus is quite as likely to be disagreeable as pleasant. If the first stage in a normal, harmonious, life-fulfilling activity, it will be felt as pleasurable; if initiating an activ- ity unsymmetrical, abnormal, destructive of the organism, it will be painful.' This last statement may perhaps stand as a loose compound of the tests for a pleasure-giving activity formulated by writers upon aesthetics. The theory upon which the following discussion is based demands a fuller explanation and exacter phrasing. This rough summary of previous writers serves, however, to indicate the necessity of examining the whole activity, not merely that initial stage of it termed popularly the stimulus.' And this whole activity in whose character inheres the pleas- ing or unpleasing effect of metaphor is that which we have discussed in this chapter as set up in the mind of the person who reads a metaphor. Why, we may ask, should such an activity be felt by the reader as pleasurable? It will be necessary in entering upon this inquiry to define some of the terms used. What, for instance, is meant by "pleasurable" in the aesthetic sense? Feeling in general we shall define' as the subjective reaction of a physical attitude. An instinctive movement to strike an assailant, itself half-conscious and wholly unintentional, the survival of a long series of ancestral habits, reacts at once upon the striker first as a vague feeling of anger, then, perhaps, as the distincter thought, "I'll fix him," or "He had no business to." Just so the feeling of pleasure is the indi- vidual's sense of a certain attitude of body, one which involves the har- monious interplay of muscles or the perfect poise of physical functions.* The ' ' lines of composition " in a picture give us pleasure because the eyes in following them move harmoniously with one another, the instinctive action of the muscles set up by them is varied, yet with a certain unity or law, the breathing is symmetrical, and the whole body is in a state of nicely poised activity.* Pleasure is, then, our feeling of harmonious physical action. This ^See Grant Allen, Physiological Aesthetics, Ch. II; H. R. Marshall, Pain, Pleasure and Aesthetics, Ch. n, and Aesthetic Principles, Ch. 11; Wm. James, Principles of Psyehelogy, Ch. XXV; John Dewey, The Theory of Emotion, Psych. Rev., Vol. I, pp. 553-569 and Vol. 11, pp. 13-32; Vernon Lee and C. Anstruther-Thomson, Beauty and Ugliness, Contemp. Pev., Vol. LXXn, pp. 544-569 and 667-688. *See Dewey, The Reflex-Arc Concept in Pyscliology^ Psych. Rev., Vol. m, pp. 357-370. ' Following Dewey's modification of James's theory of emotion. Psych. Rev., Vol. I, pp. 553-569. and n, 13-32. * In justice to Dr. Dewey, it should be stated that he nowhere makes, and perhaps would not sanction, this application of his theory to the problems of aesthetics. * Vernon Lee and C. Anstruther-Thomson, Beauty and Ugliness, Contemp. Rev., Vol. LXXn, pp. 544-569 and 667-688. — 58 — brings us very close to Grant Allen's doctrine of pleasure as the normal functioning of the body ' and Edward Rowland Sill's statement that beauty has a life-giving power .^ But to apply this theory to the metaphor. If an action felt as pleasurable is one which is symmetrical or harmonious, that set up in the reader by metaphor must be symmetrical or harmonious if it is to give pleasure. And so, indeed, it is. We have many times traced it I before, in its intellectual aspects. It consists of a sudden tension in the mind between the two incompatible images introduced, a resolution of that tension in the perception of the single impression or sensation out of which the two images had emerged, the subsequent division of the original impression and its return to a distincter unity. This means, of course, in the technical phrase, an intellectual experience of unity in variety. But this activity set up by the metaphor is not purely cerebral. Other parts of the body participate in it. The sudden tension brought about in the mind by the almost simultaneous introduction of the two elements in the metaphor means physically a sudden catching of the breath, a sense as if the lungs were being forced wide apart. Anyone who notes his sensations while reading a metaphor, will, I doubt not, discover something at least analogous to this experience. The breath cannot be released until the unity has been reached. Then, with a sigh or a laugh, it is let go, the lungs fall together and one feels the pleasure that always accompanies a lifting tension, if it has not been unduly prolonged.' This physical experience can be, I think, easily accounted for in terms of Dr. Dewey's theory, that all feeling is the subjective reaction of a bodily attitude once useful to the organism. It is a familiar fact that, as Dr. Dewey puts it, "all expectancy, waiting, suspended effort, etc., is accompanied, for obvious teleological reasons, with taking in and holding a full breath, and the maintenance of the whole system in a state of considerable tension. Now let the end suddenly 'break,' 'dawn,* let one see the 'point,' and the energy discharges. . . . This sudden relaxation of strain, so far as occurring through the medium of the breathing and vocal apparatus is laughter." And later the statement is made that "The laugh is thus a phenomenon of the same general kind as the sigh of relief. " * Either phenomenon may take place as the reader vividly experiences a metaphor. I should say, from my own experience 1 Physiological Aesthetics, Ch. II. ^ Principles of Criticism, Atlantic Monthly, Nov., 1885. ' Evidence of the existence of this tension and of the pleasurable effect of its breaking was furnished, outside my own experiences, in the course of some experiments made by me in the laboratory of psychology at the University of Michigan in 1895. In reading to several hun- dred students, taken singly, metaphors selected for the purpose, I noticed the grave, intent, ex- pectant look which was visible on the face of the individual while the first words were being read, break into a sudden smile or even a sub-vocal laugh, as the meaning of the figure flashed upon his mind, — that is, when he had traced back the two conflicting images to their root in the speaker's mind and was thus enabled to "see how he got it." ^ Psych. Rev., Vol. I, p. 559. — 59 — and observation, that one is almost certain to occur. At any rate, the sense of strain and of relief, whether recognized in its physical manifesta- tions or not, will be recalled by most sensitive readers as concomitant with their enjoyment of a metaphor. But further, as soon as the embryo of the metaphor comes to con- sciousness, it begins straightway to branch and divide. Having found its beginning, the same process occurs in us that took place in the mind of the writer. The metaphor grows and branches into two main trunks. We follow this division with a symmetrical expansion of the two lungs, a bracing of the two feet on the floor, a sense of balance ever returning to a single center and ever distributing itself anew. We experience the metaphor, not only mentally but physically as well. Its reading brings about in our physical organisms harmonious action which is felt by us as distinctly pleasurable. The pleasure derived from the reading of a metaphor may, then, be accounted for by several aesthetic formulae. It may be subsumed under the general principle that the perception of unity in variety affords grat- ification to the mind; for the metaphor, as we have shown, impels the mind to seek a pre-existing unity as at once the source and the justifica- tion for the incongruous variety of images presented. The successful discovery of this unity doubtless gives pleasure to the reader. Or, accepting Marshall's doctrine ' and that implied by Dewey, that relief from physical tension, not prolonged past the point of elasticity, is pleasurable, the effect of metaphor becomes readily comprehensible, since it induces a physical strain which is speedily relieved as the unity of the two images in the figure flashes into consciousness. Or, finally, if a normal, harmonious activity of the body be pleasur- able,^ metaphor is so, for it sets up certain physical activities which are both normal and symmetrical, both unified and various. The effect of metaphor upon the reader is, then, agreeable, be- cause metaphor stimulates him to actions, both of mind and of body, which fulfil the law of unity -in variety, which offer an outlet for pent energies, which establish a symmetrical exercise, a moving poise for the physical functions, and which consequently are felt as pleasure. 1 The doctrine of relief from tension falls under Marshall's principle that the outlet of energy repressed for a limited time is pleasurable. The breath is held for an instant, the mus- cles are rigid. The lungs and the muscles are then allowed to act and the feeling of pleasure succeeds. See Pain^ Pleasure and Aesthetics, pp. 171, 209-215. ^ Beauty and Ugliness, Contemp. Rev., Vol. LXXII, pp. 544-569, 667-688. CHAPTER V. PATHOLOGICAL FORMS OF METAPHOR. The growth of the normal metaphor has been traced as it takes place both in the mind of the writer and in that of the reader. There are, however, certain pathological forms of metaphor which lie outside ot the field already covered. These we shall discuss briefly in the light of the principles disclosed in preceding chapters. Bad metaphor is of two varieties^) The first we may call conceits or inverted metaphors ;^e second mixed or inconsistent metaphors. Conceits are what the old rhetoricians called "frigid metaphors." They may appear either as artificial throughout, or as beginning in a nat- ural and spontaneous process but perverted later into conceit. That is, the metaphor may be wholly, or only partially, "frigid." The first species of conceit, that which is artificial from the begin- ning, we have already touched upon in the previous consideration of poetic metaphor.^ It is that figure which results when its maker, in- stead of following the typical metaphor-process from a dimly per- ceived situation to its clearly discriminated parts, inverts this order and begins with the two images which he welds into one by virtue of some resemblance between them. The outlines of this process have been presented elsewhere^ and only a few additional details will here be necessary. The maker of a conceit, having noted that he in reading a metaphor establishes between its two constituent images a certain connection by means of resemblance, or analogy of situation, concludes that the meta- phor was created by the same process, in fact that all metaphor comes into being by tacking two images together by means of resemblance. He sets out to make a metaphor in this way and succeeds in placing side by side two images between which exist a certain connection. So far as verbal structure is concerned, he has achieved a metaphor; but he has done so by a process directly the converse of that taking place in the poet's mind. This process of conceit-making has perhaps never been more exactly represented than in the famous passage from Richard H: " I have been studying how I may compare This prison where I live unto the world; And for because the world is populous And here is not a creature but myself, I cannot do it; yet I'll hammer it out. My brain I'll prove the female to my soul, My soul the father; and these two beget A generation of still-breeding thoughts, And these same thoughts people this little world, In humors like the people of this world, For no thought is contented." > Chapter IL » Chapter IL — 61 — Each thought is then described as a person, and the conceit becomes more and more artificial with each added detail. These are the elaborate absurdities which might easily be expected from a man who was continually studying how he might compare one thing to another. He would always succeed in doing so, but the result must bear to the legitimate metaphor such a relation as the small boy's walking on his hands does to the normal method of locomotion. Un- doubtedly one could walk on his hands, but why should he.^ Such a question recurs continually in reading the attenuated Eliza- bethan conceits. There is in them so much ado about nothing, such meaningless filagree of images, that patience fails to read continuously a long succession of them. The following verses will perhaps serve as a sufficient illustration of this type of conceit: " Lady, the melting crystal of your eye Like frozen drops upon your cheeks did lie; Mine eye was dancing on them with delight, And saw love's flames within them burning bright. Which did mine eye entice To play with burning ice; But O, my heart, thus sporting with desire. My careless eye did set my heart on fire. that a drop from such a sweet fount fljring Should flame like fire and leave my heart a-dying! 1 bum, my tears can never drench it TiU in your eyes I bathe my heart and quench it: But there, alas, love with his fire lies sleeping. And all conspire to bum my heart with weeping." * It is possible to disclose under this mass of syllabub the plain idea that undoubtedly stood in the mind of the writer. He wished to say something like this: ' ' You seemed at first cold, but later I saw a gleam of love in your eyes that awakened my love for you. This love has grown so intense that it cannot be assuaged even by possession of you — nay, that will only increase my passion." But of course it would not do in poetry for him to present his idea in this bald, prosaic form; so he hunts about for comparisons. They lie easy to his hand. Anything cold is of course likened to ice; and love, as hot and consuming, to flames To be set on fire had even at this early period been long accred- ited as the synonym for being inspired with love and ' ' burning ice " for the chills and fever of the tender passion. Hence the conceit, which might be diagrammatically represented as a series of paired ideas, these paired ideas constituting, as the writer fondly supposes, a met- aphor, but in reality only representing a manufactured substitute for it. So much for that form of conceit which is artificial alike in its gen- * From Thomas Greares's S