UC-NRLjF B 4 1=^5 Blti ^LidiL^JiiLt A STUDY IN ENGLISH METRICS BY ADELAIDE CRAPSEY VERSE A STUDY IN ENGLISH METRICS A STUDY IN ENGLISH METRICS By ADELAIDE CRAPSEY New York ALFRED A. KNOPF MCMXVIII COPYRIGHT, 1918, BY ESTHER LOWENTHAL PBiirrED ra the united states of amebioa !^ f^ AN INTRODUCTORY NOTE It is hoped that the study here presented may prove of value not only to professional students of metrics but to all who are seriously interested in poetry. It forms only one part, in itself com- plete, of an investigation of certain problems in verse structure the full carrying out of which was prevented by Miss Crapsey's death in the autumn of 1914. An indication of what remained to do may be found in the note appended to the essay. The tables printed as Additional Analysis pre- sent a portion of the data which had been col- lected for a second part of the study. To any one who reads to the end, it must be evident, that Miss Crapsey regarded the use of exact measurement and analysis — which prob- ably is denoted as the application of phonetics to metrics? — as essential to the "finer and righter" appreciation of poetry ; that she considered a full awareness of technique the necessary equipment of one who would understand fully the subtle and delicate beauty of verse. This is a position that 6 AN INTRODUCTORY NOTE has long been won in regard to music; no one would attempt to be a critic of musical composi- tion who was unaware of the technical problems of musical construction. Is it not likely that the criticism of poetry may become far more signifi- cant when our literary critics consider necessary a corresponding equipment? At least it may be stated that the laborious analysis was in Miss Crapsey's case dictated by an acute sense of the beauty of verse, by an aesthetic experience of unusual intensity. To one who knew her and watched her work Miss Crapsey's untimely death seems to have brought about a double frustration ; it prevented the completion of the study in pros- ody and the undertaking, which that study would so much have enriched, of a series of essays in criticism. That a scientific knowledge of the technique of verse may be a potent tool in the hand of the poet also will probably at once occur to the readers of Miss Crapsey's verse. E. L. SYNOPSIS MAIN THESIS That an important application of phonetics to metrical problems lies in the study of phonetic word-structure. SUB-THESIS UNDER PRESENT CONSIDERATION That systematic analysis of English poems indicates the existence of a distinct structural differentiation of vocabularies into three main types : I. A type of vocabulary purely, or mainly, mono-dissyllabic, i. e,, showing a characteristic occurrence of polysyllables running from to about 2%. II. A type of medium structural complexity, i, e,, showing a characteristic occurrence of poly- syllables running from about 3%- to about 5^/2% » with a tendency to drop towards 2% and to rise towards 6%. III. A type of extreme structural com- plexity, L e,, showing a characteristic occurrence 8 SYNOPSIS of polysyllables running from about 7% to about 8%%, with a tendency to drop towards 6% and to rise towards 914% (or 10% ?) . Note: The term "polysyllable" is used to in- clude all words over two syllables in length. The discussion of the sub-thesis falls into three main parts: A, Presentation of analysis 1. Derivation of scale of polysyllabic occur- rence for experimental testing from (a) 125 Nursery Rhymes (b) Milton (Table I) (c) Pope (Table II) 2. First testing of scale from the work of (a) Tennyson (Table III) (b) Swinburne (Table IV) (c) Francis Thompson (Table V) (d) Maurice Hewlett (Table VI) B, Summary of three important points in- volved in differentiation of vocabularies 1. Elementary word-forms entering into combination 2. Range of values in word-accent 3. Conditions of "weighting" SYNOPSIS 9 C, Importance of differentiation of vocabu- laries in study of Metrics indicated mith reference to 1. The problem as a whole 2. Tennyson and the development of the decasyllable 3. Swinburne and the development of triple rhythms A STUDY IN ENGLISH METRICS A STUDY IN ENGLISH METRICS It is my object in the * prei^ent^ disoii^^loil- to venture the suggestion that an important appli- cation of phonetics to metrical problems lies in the study of phonetic word-structure. I have given first (in tentative formulation, of course) a specific conclusion with supporting data, and second a brief indication of the reasons for main- taining the general position. To deal with a definite, if still limited, range of fact before ap- proaching the wider theoretical issues has seemed to me the better method, at least for the present. One offers thus as first evidence the results of systematic analysis and in so far as these possess, or seem to possess, a certain solidity and co- herence within themselves, they are in some sort a guarantee that the underlying theory is worthy of attention. May I say that the statement here given is to be regarded as nothing more formal or definitive than a first rough summary drawn up in order to 13 14 A STUDY IN ENGLISH METRICS open discussion and (if the conclusions indicated will hold at all) to serve as a basis for correction and further investigation? The position taken can be outlined quickly and, for the time being, with I think fair ex- plicitness in the following way. Scansion iso- lates, for the sake of analysis, the basic metrical units of verse — feet ; the same analytical scrutiny must, I believe, be given to the basic phonetic units of speech — L e,, phonetic word-forms — be- fore we can possess sufficient data for the study of one of the fundamental problems of verse as a whole, the relation of the word to the foot. The scope of the proposed analysis must evi- dently parallel within its own field that of scansion within the metrical field; that is, as the study of English scansion deals with the whole possible variety of metrical units in English verse and with the special occurrence of these in individual poems, so the study of phonetic word- structure must deal with the whole variety of word-forms existing in English and with the A STUDY IN ENGLISH METRICS 15 comparative occurrence of these in specific vocabularies. Obviously there can be given, at this time, in support of the position outlined but a limited amount of experimental analysis, and as obviously, one must select for this first examina- tion and presentation a group of facts which will yield results of main or central significance. Accordingly I submit for immediate considera- tion, as summarizing what seems to be the most important single issue involved, the following tentative formulation, namely — That the systematic analysis of English poems seems to indicate the existence of a tendency to- ward distinct structural differentiations of vocab- ulary, the main types being three in number: — I. A type of vocabulary purely, or mainly, mono-dissyllabic. II. A type showing medium structural com- plexity, L e,j containing a medium number of words of three syllables and over. III. A type showing extreme structural complexity, i, e., containing an extreme number of words of three syllables and over. Before proceeding to the discussion of this 16 A STUDY IN ENGLISH METRICS — ^^— ^^^— ^^—^^^i— — — — — "^—^ ^■^-^™^— — i»^»i— ^— ^■^ thesis, it is necessary to deal with certain ques- tions of detail — more especially with certain diffi- culties — which are bound to arise in the actual carrying out of the work. In the first place, even admitting it to be theoretically desirable, do we possess to-day a pronunciation sufficiently standardized to make possible the analysis of vocabularies on anything like the scale suggested? Variations in pro- nunciation are notorious. How can we be as- sured that a classification of the words in any given poem will represent the pronunciation of the poet who ^Tote? Is it not, rather, certain that the analysis will depend upon the pronun- ciation of the critic who dissects, and that the results of analysis will, consequently, vary with each new critic? And further, will not the diffi- culties be hopelessly increased when different his- toric periods are to be considered? No attempt is made to minimize these difficulties, nor, for the present, to meet them in detail. Two imme- diately practical considerations are, however, urged. First, as to uncertainties of pronunciation per se. Nothing it should be noted is under present A S TUDY IN ENGLISH METRICS 17 p— "^^^^ ^^^^^^^™^^^— ^^^^^— ^^— — ^^^'— ^^™^^^— — ^""^^^™— ^^ examination except syllabification (i. e., the number of syllables) ; possible differences of ac- centuation do not enter into the matter at all. Moreover, since the classification of vocabularies in question is based on the relative occurrence of words of three syllables and over, only two main groupings of words are dealt with: the mono- dissyllabic group regarded as a whole; the "polysyllabic" group regarded as a whole. As a practical matter of fact, therefore, as far as the present investigation is concerned, the cases of possible uncertainty narrow down to the par- ticular group of words where there is a question between two and three syllables. I do not think that, however classified, the number of words in this group is large enough to affect, in any seri- ous way, the general results obtained. Second, as to the question of changing stand- ards of pronunciation. Here again nothing is urged beyond the reduction to a minimum of the difficulties involved — in this case by selecting as far as possible work which allows the use of what may be roughly called the present standard of pronunciation. The single exception to this is the work of Milton. Here what may be roughly 18 A STUDY IN ENGLISH METRICS called the "Elizabethan" standard has been used. This statement, it is most hastily to be said, im- plies no absurd assertion that one has been able to reconstruct Elizabethan pronunciation as a whole. For the practical matter in hand, the main concern is simply with the fuller syllabifica- tion of a perfectly well-recognized class of words — the ion, ious, etc. class. In the analysis of Milton's vocabulary given below this fuller syl- labification has been kept as consistently as pos- sible — ocean, union, nation, for instance, being counted as trisyllables. It is to be noted, how- ever, that the point made just previously holds here also. The results, as stated, would be af- fected only by the classification of those words where there is the question of a change from three to two syllables. One somewhat smaller detail is still to be men- tioned — the classification of compounds. The rule followed has been to regard compounds as whole words, many-fountained, for instance, be- ing classed as a word of four syllables. In finer analysis it will, of course, become necessary to take into account the extent to which compounds are present in the whole polysyllabic group — as A STUDY IN ENGLISH METRICS 19 r it is also necessary to take into account the extent to which proper names are present. Turning now to my thesis, I have chosen the following poems for first analysis; (a) 125 Nurseiy Rhymes, (b) Paradise Lost and Sam- son Agonistes, (c) five of Pope's poems (see below). The reason for this selection is plain enough. If they exist at all, we have here, pretty clearly, examples of the three indicated types of vocabulary. It is difficult, at any rate, to imag- ine much doubt as to the facts that in Nursery Rhymes there are few "long" words, while Mil- ton's is the great example in English verse of a polysyllabic vocabulary, or probably, as to the fact that Pope's vocabulary would come some- where between these two extremes. The first business in hand is, therefore, to see whether sys- tematic analysis will bear out this impression of differentiation where it is strongest, and, if so, what exacter arithmetical values are to be given to the words in which the range of difference has so far been expressed. What, in English, do we more precisely mean by "few" or "more," or "many" polysyllables? 20 A STUDY IN ENGLISH METRICS My results for the 125 Nursery Rhymes are as follows: first, 59 of the Rhymes — very nearly half (47.2%) — are purely mono-dissyllabic; sec- ond, taking the Rhymes as a whole, of the total number of words used (6,928), 97.86% belong to the mono-dissyllabic group, 2.13% to the poly- syllabic group. At the lowest extreme may, therefore, be placed an occurrence of polysyllables running from zero to about 2%. The next section of analysis — that of Milton and Pope — can best be given in tabulated form. (Seep. 21.) Summarizing: the tables show a characteristic occurrence of polysyllables in Milton's poems running from about 7% to about 8l/^%, with a tendency to drop toward 6% and to rise to 9%, and a characteristic occurrence in the poems by Pope running from about 4% to about 5%%. These figures may be held, tentatively, to rep- resent the extreme and the medium occurrence of polysyllables. A STUDY IN ENGLISH METRICS 21 Table I. Milton Total No. Per cent Mono- Per cent of words dissyllabic Polysyllabic Paradise Lost I 5,960 91.67 8.33 II 7,917 92.24 7.75 Ill 5,56^ 92.07 7.92 IV 7,700 92.74 7.24 V...... 6,804 92.01 7.99 VI 6,773 90.95 9.03 VII 4,774 91.40 8.58 VIII 4,921 91.45 8.53 IX 9,010 93.01 6.98 X 8,370 91.74 8.24 XI 6,859 92.48 7.50 XII 4,930 91.78 92.03 8.21 Total 79,584 7.95 Samson Agonistes Dialogue 9,465 92.04 7.94 Choruses 3,427 90.92 91.75 9.08 Total 12,892 8.23 Table II. Pope Total No. Per cent Mono- Per cent of words dissyllabic Polysyllabic Essay on Criticism 5,744 94.91 5.08 The Rape of the Lock . . . 6,149 94.71 5.28 Elegy — Unfortunate Lady 652 95.86 4.14 Essay on Man I 2,288 94.32 5.68 II 2,251 94.32 5.68 Ill 2,481 94.43 5.56 IV 3,141 95.54 94.72 4.46 Total 10,161 5 27 Epistle to Dr. Arbuthnot 3,353 95.91 4.09 22 A STUDY IN ENGLISH METRICS Restating now the description of the types as at first given we have : I. A type of vocabulary purely, or mainly, mono-dissyllabic : i, e,, containing a characteristic occurrence of words of three syllables and over, running from to about 2%. II. A type of medium structural complex- ity : i. e,, containing a characteristic occurrence of words of tlii-ee syllables and over, running from about 4% to about 5^%, with, probably, a ten- dency to drop towards 3% and to rise toward 6%. III. A type of extreme structural complex- ity: L e.y showing a characteristic occurrence of words of three syllables and over, running from about 7% to about 8^%, with a tendency to drop towards 6% and to rise to 9%. The next step is to see how far the scheme thus roughly established can be applied with reference to the vocabularies of other poems. Here selec- tion has been made — always the selection of whole poems — from the work of Tennyson, Swinburne, Francis Thompson and Mr. Maurice Hewlett. The results of analysis are given in tabulated form. A STUDY IN ENGLISH METRICS 23 Table III. Tennyson Total No. of words Per cent Mono- dissyllabic Per cent Polysyllabic Oenone Ulysses Tithonus The Coming of Arthur, Merlin and Vivien . . . . Lancelot and Elaine . . . The Holy Grail Guinevere The Passing of Arthur, 1,988 556 599 4,256 7,896 11,799 7,474 5,671 3,855 94.31 96.94 96.33 96.54 95.88 95.83 96.48 95.68 96.59 5.68 3.05 3.67 3.45 4.10 15 50 30 39 Table IV. Swinburne Total No. of words Per cent Mono- dissyllabic Per cent Polysyllabic Chastelard I ...... . II Ill IV V Total Atalanta in Calydon Dialogue Choruses Total Hymn to Proserpine. . Hesperia The Forsaken Garden 4,712 3,975 3,703 7,061 6,001 25,452 12,832 5,536 18,368 1,003 1,283 671 98.59 98.56 98.65 98.30 98.18 98.42 95.84 96.83 96.14 97.50 97.03 98.80 1.40 1.43 1.35 1.68 1.81 1.57 14 17 3.85 2.49 2.96 1.19 24 A STUDY IN ENGLISH METRICS Table V. Francis Thompson Total No. of words Per cent Mono- dissyllabic Per cent Polysyllabic The Hound of Heaven . An Anthem of Earth . . Sister Songs I . II Total 1,205 2,798 2,658 5,457 8,115 92.61 90.59 92.02 92.11 92.06 7.38 9.39 7.97 7.89 7.92 Table VI. Maurice Hewlett Total No. of words Per cent Mono- dissyllabic Per cent Polysyllabic Minos, King of Crete .... Ariadne in Naxos Death of Hippolytus .... 7,882 8,325 8,080 96.41 95.92 95.96 3.59 4.07 4.03 It is clear that these vocabularies fall readily into the suggested classification. In the work of Tennyson the vocabulary used is of the "medium" type, but it is to be remarked that while in the poems of Pope under analysis the tendency is to rise from 4% toward 5%, there is here a tendency to drop from 4% toward S%. Also of the medium type, "Atalanta in Caly- don" shows the same tendency, while the *'Hymn to Proserpine" and "Hesperia" drop still fur- ther, from 3% towards 2%. In "Chastelard" A STUDY IN ENGLISH METRICS 25 and "The Forsaken Garden" the occurrence, under 2%, is that of the first type. The use of so markedly mono-dissyllabic a vocabulary is, of course, particularly to be noted in a poem of the length of "Chastelard." The work of Francis Thompson, without any analysis easily to be recognized as of the "ex- treme" polysyllabic type, shows, under analysis, in the three poems chosen, the characteristic oc- currences derived from Milton — 7% to 99^. Mr. Maurice Hewlett, on the other hand, uses in his trilogy, "The Agonists," the medium type of vocabulary, with, as in the case of Tennyson, the 3-4% occurrence rather than the 4-5% ex- emplified by Pope. Thus the only changes to be made as the result of this section of analysis are to give 2% as the lower limit of polysyllabic occurrence for the medium type of vocabulary and to indicate a slight tendency to rise above 9% in the extreme type, taking 9^% (or, perhaps, even 10%?) as the experimental upper limit. I can carry the direct demonstration of my thesis no farther at present. It remains in this 26 A STUDY IN ENGLISH METRICS direction but to continue analysis, working at the two main groupings of words here given and, at the resokition of these into their finer sub- divisions, establishing the percentages of sepa- rate occurrence for the main classes of words (monosyllables, dissyllables, trisyllables, etc.) and within each of these main classes as a whole, the percentage occurrence of their respective ac- centual sub-types. This detailed analysis will give (as that already submitted for the larger groupings) exacter information as to total range of occurrence and as to comparative occurrence in specific vocabularies. But continuance of analysis without pausing to consider the many delicate and controversial questions involved would be unprofitable and it would be, I think, unsatisfactory, even apart from these difficulties, to present further ac- cumulations of fact without meeting directly the cardinal issue of their prosodic application. To this issue I therefore turn in the second part of my exposition. May I however at this juncture make some brief and informal comment on points connected not with the prosodic application but with the A STUDY IN ENGLISH METRICS 27 purely linguistic interest of the matters just pre- sented? So far as I know the analysis covers new ground and I naturally find myself wonder- ing to what extent the facts revealed may seem in and of themselves important — whether, that is, the suggestion that such differentiations of vo- cabulary may exist will arrest attention as rais- ing a question of genuine significance in regard to English and whether it seems more useful to express the total and comparative occurrences of the various word-forms in definite arithmetical values rather than in the vague terms of few, more and many. It may be that such data as this would be valuable in tracing the historic develop- ment of English and in establishing comparisons between English and the "more" polysyllabic languages ? That the few polysyllables of English are lit- erally so few surprised me. I had expected (or since I was surprised I suppose I had expected) a generally higher percentage of occurrence. The marked stability of the various occurrences surprised me, too; I had expected rather wider fluctuations. II The task of formulating a single proposition and presenting facts in evidence now gives way to the more intricate business of showing this proposition not in isolation but as part of a con- sistent theory of English verse-structure as a whole. That I give the merest first sketch of such inclusive formulation has already been stated and is surely too obvious to need repeti- tion, but I should like to add that, acutely aware at once of the difficulties involved and of my own but slender competence, I mean to go no further in the unavoidable generalizations than is neces- sary in order to indicate the connections which I think it may be possible to establish. My dis- cussion moreover will be cast as little as possible in terms of pure theory and as much as possible in terms of technique. And, finally, it may be well to say at once that small emphasis is put upon any matter of novelty, that in fact, as will in a moment appear, it is for the most part the quite contrary point which is insisted upon. 29 30 A STUDY IN ENGLISH METRICS It will make probably for greater clearness and economy if I work again from the central point of a single sharply formulated thesis. I submit therefore the following, namely: — That no prosodic theory is adequate which fails explicitly to recognize within verse-struc- ture as a whole a complex of three inter-existent structures: 1, the verse- form proper, itself two- fold, consisting of (a) the rhythmic arrangement and (b) the syllabic aiTangement by means of which the rhythm is exteriorized; and 2, the sub-structural phonetic speech-arrangement. By the phonetic sub-structure I mean, finally, everything connected with the organized physi- cal material of language; but attention in the present discussion is sharply focused on the word, which, I would contend, is in its phonetic aspect the basic structural unit of language phys- ically considered as the foot is the basic struc- tural unit of the verse-form proper. The discussion will fall into two parts: first, as preliminary, a rapid consideration of the ex- tent to which English prosodic theory has so far tended to recognize, either as a whole or in part, the principles here stated; and, second, a state- A STUDY IN ENGLISH METRICS 31 merit of the more important points in technique which, if correctly observed, seem to force the acceptance of the thesis, again in the end, as a whole, but with a present emphasis on the valid- ity of the last term. (1) In considering the first point, I must more than ever deal in makeshift discussion since its adequate treatment would require, what I by no means possess, a thorough-going knowledge of the historical development of English prosodic theory. Rather however than leave (the only alternative) an absolute gap in my argument, I give for what it is worth the following roughly generalized statement of what seems to emerge as a coherent line of development. It will serve at any rate to recognize the fact that the ability to demonstrate the steady convergence of theo- retical speculation toward it as a common con- clusion will be an important factor in the estab- lishment of any final synthesis. At the beginning of the sequence, then, as I see it, are to be considered the prosodists who think of verse as a simple uncomplex whole. 32 A STUDY IN ENGLISH METRICS Among these are most of the earlier and a scat- tering of the later men. Professor Saintsbury is (I think) the main present exponent of a theory based upon any such unanalyzed reaction. It is also significant to note that we have here the average layman's view of the matter. In this stage of introspective observation conscious at- tention seems to be focused on the arrangement of syllables by means of which the rhythm is ex- teriorized (the b term of my thesis). Such simplification has, in the main, given way to conscious awareness of the differentiated rhythmic arrangement and I think it would probably be safe to say that many present-day metrists tend to recognize the twofold character of the verse-form proper (the co-existent a-b terms of my thesis). The development of this awareness of the rhythmic factor per se^ signal- ized (I suppose?) by the appearance of the ''mu- sical" and "temporal" scansionists, may be said to mark the second stage in introspective analy- sis, what had seemed to be an uncomplex whole now revealing itself as a complex entity contain- ing within itself two inter-existent entities. It is not to be expected, however, that the new A STUDY IN ENGLISH METRICS 33 theoretical formulation conditioned by the ad- vance in accuracy of observation should be im- mediately achieved. There is especially to be reckoned with, as one would expect, a tendency to throw very great emphasis on the newly ob- served factor, rhythm, and where this is carried to the extent of a decided over-emphasis an im- portant result follows — the failure, namely, properly to correlate the manifestation of rhythm under consideration with its specialized medium, in this case, language. Such failure is probably to be regarded as the vitiating flaw in the theories of the musical scansionists and, it follows, of their method which is based on an attempt to transfer the terms and notation de- veloped in relation to the manifestations of rhythm in one material to its manifestations in a different material. It is also to be noted that where over-emphasis on the rhythmic factor exists in any marked degree it results, naturally, in reversion toward the primitive view of verse as an uncomplex whole, but with conscious atten- tion now shifted to focus on the rhythmic arrangement (the a term of my thesis). The treatment which (so far as my knowledge goes) 34 A STUDY IN ENGLISH METRICS shows least any such deflection of emphasis is that of Mr. T. S. Omond. Mr. Omond, if I understand him, while differentiating sharply between the rhythmic arrangement and the manifesting syllabic structure, indicates with equal clearness their proper condition of inter- existence within the verse-form as a unit. Nor, further, is it to be expected that the more complex methods necessitated by the advance in theory should be immediately perfected. As a matter of fact the crucial change in method in- herent in the new theoretical standpoint would not be immediately apparent and one of the most important achievements of recent prosodic study has been, I should say, to discover and state this change. The desideratum is, clearly, a method allowing the close study of the rhythmic groups of verse, the "musical" scansion just mentioned being an attempt to meet this need. What has now become apparent is that we soon reach here the limits of possible analysis based on simple ob- servation "by ear" or by our "sense of" rhythm. The delicate and accurate study of the rhythmic groups of verse must, it is seen, be carried on by means of laboratory experiment. This issue as A STUDY IN ENGLISH METRICS 35 to method, the importance of which it seems to me impossible to exaggerate, has now been definitely raised, more especially, whatever one may think of his conclusions, by M. Paul Ver- rier. As a result it is no longer possible to dis- cuss except on the basis of relevant evidence gathered by genuinely scientific laboratory analysis such fundamental questions of verse- structure as the isochronism of the rhythmic groups or — the crux of the whole matter — the fundamental difference between prose and verse. It is to be noted that in thus correlating the prob- lem of rhythm in verse with the whole rhythmic problem there is necessitated, since a speech- group is concerned, making relevant questions of syllabic length and the nature and function of accent, a first application of experimental phonetics to prosodic study. Two stages in what may be at least reasonably suggested as a coherent sequence in the develop- ment of English prosodic theory have now been indicated. Awareness of verse-structure in terms of a naive simplification gives way to growing awareness of its real complexity. The increased accuracy of observation conditions a 36 A STUDY IN ENGLISH METRICS change in theoretical formulation and as obser- vation and theory grow more precise they neces- sitate commensurate advance in precision of method. It is now the logically sequent step in my argument to maintain that continued intro- spective analysis of our reaction to verse-struc- ture as a whole will next result in conscious awareness of the existence within it of the sub- structural speech-arrangement (the 2, of my thesis). I give in briefest sunmiary three points which may not unreasonably be regarded as indicating this to be in fact the case. (a) There is discernible in some recent dis- cussions of English verse a growing reluctance to admit any differentiated verse values as neces- sary in accounting for verse-structure. This atti- tude shows itself markedly in, for instance, the discussion of "accent" or "stress" where an im- portant controversial question is that relating to the use in verse of syllables carrying secondary word-accent and of certain grammatical classes of words, conjunctions, prepositions, etc. The ''Rules of Stress Rhythm" formulated by Mr. Robert Bridges give clear expression to the A STUDY IN ENGLISH METRICS 37 ^^^^''""""*'''^^™'''''"''"^™'™''^''^'''™"™™'"'"™'''''^'^*'^~"^^'™"'""^~™'*''''™'"™ opinion denying these any value in verse other than their "natural speech" value: — e, g.. Rule II "The stresses must all be true speech- stresses," and Rule IV "A stress has a peculiarly strong attraction for its own proclitics and en- clitics." {Milton s Prosody, p. 91, p. 93.) The same view is implicit in Mr. Maurice Hewlett's introductory note explaining the versification of The Agonists, "Now all I ask of mine [my hearers] is that the verse be read to them as prose, with the stresses where they would nat- urally fall, and full value given to the vowel sounds of ordinary speech. If this rule be ob- served, and the indicated pauses followed, the three plays ought to be revealed as verse" (The Agonists, pp. X-XI). The attitude expressed in these statements is, I would submit, normally to be expected at the present stage in the development of English prosodic theory. It signalizes growing aware- ness of the speech-arrangement per se, manifest- ing at the same time exactly the over-emphasis on this arrangement which would naturally ac- company its first conscious perception. The effect of such over-emphasis is seen in the in- 38 A STUDY IN ENGLISH METRICS clination either to reduce the whole verse-struc- ture to the speech-arrangement alone, thus re- verting towards the primitive view of verse as an uncomplex whole, or to suppress the middle term (any differentiated syllabic arrangement, that is) the speech-arrangement being regarded as itself externalizing the rhythm. (b) Other recent discussions, on the contrary, recognize with varying degrees of explicitness what is often called the "contrast" of speech- and verse-units, or their "non-coincidence." In anv statement of this kind recognition of the dif- ferentiated speech-arrangement is so clearly made that it is difficult to give it distincter ex- pression. When, for instance. Sir Walter Ra- leigh writes of Milton — "His chief study, it will be found, is to varv the word in relation to the foot, and the sentence in relation to the line" {Milton, p. 199) — he recognizes the co-existence within the whole of the verse-form proper and the substructural speech-arrangement and, not only this, but he recognizes specifically, in the first clause, what I have indicated as seeming to me the primary structural point, the relation of the foot to the word. In this connection I would A STUDY IN ENGLISH 3IETRICS 39 further, and especially, point out the importance of Professor Saintsbury's insistence on the "non- coincidence" of verse- and speech-units, particu- larly (again in specific recognition) of the foot and word. I have not made a detailed examina- tion of Professor Saintsbury's prosodic state- ment, but I am inclined at least to wonder whether when in its final evaluation this is done, his constantly increasing emphasis on just this point of the non-coincidence of foot- and word- division will not emerge as a feature of outstand- ing significance. The first direct statement of it (that I noticed at least) is in a foot-note — "I think it is a mistake to try to make foot- corre- spond with word-division: the best metre is often that which divides the words most" {History of English Prosody, I, p. 387). In the second and third volumes the point is increasingly insisted on, whether directly or by implication, reaching in the third volume the clear-cut assertion that ". . . in poetry except in so far as our abun- dant monosyllables prevent it we positively avoid, save for special reasons, coincidence of foot- and word-end" {ibid.. Vol. Ill, p. 456). And now in the recently issued History of Prose 40 A STUDY IN ENGLISH METRICS Rhythm Professor Saint sbury is almost ready to see in this fact of non-coincidence one of the fundamental differences between prose and verse, smnmarizing the view thus in his table of axioms and inferences: "There is no objection to the falling of a foot-end in the middle of a word. But it is less frequent in prose than in verse; and its comparative rarity perhaps fur- nishes one of the diiFerences between prose- and verse-rhythm" {History of English Prose Rhythm, 1^, 479, IF 9). In such recognitions of "contrast" or ''non- coincidence" as these quoted from Professor Raleigh and Professor Saintsbury there is, again I would submit, a variant attitude which would normally occur, along with the one previously noted, at the present stage of development in English prosodic theory. We have here a clear awareness of the co-existent speech- and verse- arrangements within verse as a whole, an aware- ness acute moreover at the primary point, the re- lation of the two basic units concerned, the foot and the word; but the full theoretical implication of the observed fact is not discerned and there is consequently no generalized theoretical state- A STUDY IN ENGLISH METRICS 41 merit of it. In this stage there is evidenced that curious condition of seeing and not seeing which, however difficult to describe or account for, is nevertheless as a psychological condition easily recognizable. (c) And, third, it is possible to give at least one instance (there may be more) of the formu- lation which would naturally follow the one just described: a formulation, that is, in which the fact of the inter-existence within verse as a whole of the verse- and speech-arrangements is not only observed but given its generalized theoretical value. I quote from Mr. Thomas Rudmose- Brown: "When M. Verrier retorts . . . . that 'Phonetics divides ... in accordance with what we hear' and that metrics cannot di- vide otherwise on pain of being but 'arbitrary dogmatism' he ignores what is fundamental in all metrical investigation, namely that a line of verse is a portion of speech-material with all its phonetic features (corresponding to its ethos as well as its logos ) adjusted, without violence, to a fixed and definite metrical scheme. The two en- tities, metrical scheme and portion of speech- material adjusted thereto, are distinct and the 42 A STUDY IN ENGLISH METRICS chief study of the metricist is the manner of ad- justment of the latter to the former, the way in which a suitable portion of phonetic liquid is chosen and poured into metrical bottles" {Eng- lish and French Metric, Modern Language Re- view, Jan., 1913, p. 104) . I do not know to what extent Mr. Rudmose-Brown develops the state- ment here made in his general treatment of metrics (the matter is complicated by his postu- lation of two different basic principles as opera- tive in English verse) but this is, as far as it goes, clear enough to serve fairly, I think, as evidence in the matter at issue. For these reasons, then, it seems to me that it may be possible to correlate present analyses of verse-structure with the sequence in development experimentally postulated for English prosodic theory as a whole. We find, to recapitulate, in considering these analyses in relation to the as- sumed next step in that development, i. e., grow- ing awareness of the sub-structural speech- arrangement as existent within verse-structure as a whole: — (a) statements conditioned by so keen an awareness of this arrangement, in its first conscious perception, that it obliterates or A STUDY IN ENGLISH METEICS 43 obscures conscious awareness of the other in- cluded arrangements: (b) statements condi- tioned by an awareness which extends to recog- nition of the fact of the inter-existence of the verse- and speech-arrangements but which stop short of any clearly generalized perception of the theoretical implications of the fact: (c) a state- ment which, conditioned by such generalized per- ception, gives generalized theoretical formula- tion. These statements, taken together, show, I would now submit, the third main stage in the postulated sequence to be in process of accom- plishment, and it is clear that a main justification of the present discussion would rest upon the assumption that it represents the next step in this process, i. e., the step necessitated by the fact that generalized theoretical formulation is not really complete until it is accompanied by a fully conscious awareness of the necessary implica- tions as to method. It is this issue as to method which is here definitely raised, and I would in- sist that if the position taken will hold at all, it holds as part of it that we have now reached in relation to the third main advance in correct ob- servation, as already in relation to the second, a 44 A STUDY IN ENGLISH METRICS place where the issue as to method becomes a crucial one. The desideratum in this case is a method allowing that close study of the sub- structural speech-arrangement which is neces- sary to a proper understanding of the co-existent verse-arrangement. Such methodical investiga- tion must, I would contend, begin at the basic point, the relation of the foot to the word and this brings me, in theoretical approach, to my initial, and central, suggestion — that an impor- tant application of phonetics to metrical prob- lems lies in the study of phonetic word-structure (see p. 1) since it is only by means of such study that we can obtain the requisite information as to the principles of phonetic word-structure, the varieties (in their inherent classification) of English phonetic word-forms and their compara- tive occurrences. I have in the analysis already submitted, put this suggestion into experimental operation using the method advocated, not to present a new point, but to test certain basic dif- ferences in vocabularies of which we seem to get clear and unmistakable first report ''by ear." The results obtained tend, it has been seen, to verify this impression, but while the first un- A STUDY IN ENGLISH METRICS 45 tested report gives us information in vague terms and is even within these limits generally satis- factory only where extremes are concerned — where we have to distinguish, that is, between an unmistakably mono-dissyllabic and an un- mistakably polysyllabic vocabulary — systematic analysis not only translates vague into determin- ate arithmetic values but allows the tracing of intermediate gradations with a precision and delicacy otherwise impossible. Is it necessary to say that one is of course all the time working "by ear" — but by a reasoned and tested hearing? (2) It remains to indicate how the study of word- structure can be considered to have direct bear- ing on specific problems of verse-technique. There has been, of course, throughout the whole discussion implicit general assumption that such direct bearing exists, for if we recognize "by ear" the varying of the word in relation to the foot, or structural differences in vocabularies, it is because these as basic conditions are perceived by us in terms of total final effect. The present 46 A STUDY IN ENGLISH METRICS infinitely difficult and elusive problem is to re- solve the whole generalized observation into its component detail. Again it is necessary to write selectively, choosing only what has main signifi- cance and presenting not full discussion but con- densed outline. Perhaps it may be as well to give first, with momentary disconnection from the central thesis, such part of the matter as can be considered with little or no reference to any particular theory. To begin with, then, it is to be noted that where a mono-dissyllabic vocabulary is used the completed arrangement is constructed, entirely or almost entirely, of combinations of two main phonetic word-forms, one of these the dissyllabic, existing in two sub-forms. With the introduc- tion of polysyllables the number of different ele- mentary word-forms increases, the completed ar- rangement being composed of combinations not of two, but of three, four, five or more main word-forms with the corresponding increase in the number of sub-forms. There is then to be considered in a polysyllabic as opposed to a mono-dissyllabic vocabulary an increase in the variety of elementary word-forms entering into A STUDY IN ENGLISH METRICS 47 combination to form the completed arrangement. A main distinction between the "shorter" and ''longer" word-forms — the distinction which justifies their division into two main groups — lies in the fact that the longer word necessitates for its construction the use of a second accentual value, secondary word-accent. With the in- creased variety in word-forms found in the poly- syllabic vocabulary there is, then, also to be con- sidered the introduction of this further value in word-accent. A third point in connection with the introduc- tion of polysyllables seems to me of great and perhaps not always fully realized importance; its effect, I mean, on the problem of what for lack of a better term may be called the problem of weighting. The English monosyllable is in general a "full" or "heavy" syllable; polysyl- lables contain usually one or more "light" syl- lables. When therefore a vocabulary contains a very high percentage of monosyllables it tends, roughly speaking, to be in a consistent condition of heavy weighting; with an increase, in dissyl- lables as well, but more especially of polysyl- lables, the number of light syllables increases and 48 A STUDY IN ENGLISH METRICS when these reach a sufficiently high occurrence the significant fact to notice is that the vocabu- lary containing, as always in English, a large number of heavy syllables, combines with these an appreciable number of syllables in the oppo- site condition of lightness. This I state with reference to the present pro- nunciation of English. The problem of weight- ing, I may note, seems to have been most ex- plicitly discussed in connection with the changed condition in the language resulting from the loss of the final e. That this meant not only a struc- tural change as altering in many words the num- ber of their syllables, but that it also, owing to the complete disappearance of so large a number of light syllables, brought about a general change in weighting throughout the language is commonly recognized, together with the special effect of this latter change on Milton's technical problem as compared with Chaucer's. I have been, to add a related comment, puzzled over the fact that while the final e and its vanishment have been, as all the world knows, so thoroughly dealt with, an analogous change from Eliza- bethan (or Tudor) pronunciation to our own has A STUDY IN ENGLISH METRICS 49 been left, naturally not unnoticed, but rather strikingly undiscussed, at any rate in its prosodic connection. This change, affecting words in ion, ious, etc. (see p. 18), resulted, as did the loss of the e, in the shortening of many words by one syllable and in the consequent total loss of a large number of light syllables; moreover, since these syllables occurred always within the word, the resultant change in its flexibility is even greater; and finally, what did not at all follow the loss of the e^ there results in this case a marked increase in the occurrence of certain con- sonantal sounds, notably sh. It needs but to mention as example the Cherub Contemplation. Keightley, noting his usage in regard to Hebrew and Classical proper names, says that Milton "abhorred sh" (Geightley, Life^ Opinions and Writings of John Milton, London, 1859, p. 439, p. 448). However detail must wait for later discussion. The main conditions involved, here not much more than enumerated, are ( 1 ) the total number of word-forms entering into combination to form the completed arrangement (2) the range of values in word-accent (3) general conditions of 50 A STUDY IN ENGLISH METRICS weighting. The question, irrespective of theory, is whether these conditions tend to force them- selves on one's observation as significant condi- tions within the medium. The discussion already given goes to show that there is an increasing tendency to note just such conditions and, as a matter of fact, it would, I suppose, be agreed without any discussion that the particular obser- vations enumerated present in themselves noth- ing new. The argument flows not from assump- tion of novelty but from the assumption that a general tendency to note certain conditions, with whatever degree of indirectness, is in itself evi- dence for the correctness and importance of the observations and at once furnishes and justifies the impulse to correlate all cognate observations as a means of dealing adequately with the ques- tion of their total significance. And with this final question of significance to be met, there must be a return to the sharply defined theoreti- cal connection. I make now curt experimental assumption first of the complexity of the whole verse-struc- ture as stated in my second thesis : second of the classification of vocabularies as stated in my first A STUDY IN ENGLISH METRICS 51 thesis : and, finally, I shall add to these, since it is not here under direct discussion, such assump- tions regarding the verse-form proper as are necessary to my argument. I must obviously, where the verse-form itself is concerned, work with reference to a primary classification according to rhythm. I assume, therefore, the existence in English verse of two main rhythms, duple and triple, each existing in two varieties: duple rhythm, rising and falling, and triple rhythm, rising and falling. In duple rhythm the "normal" syllabic unit, or foot, con- tains two syllables: in triple rhythm three syl- lables; the difference in cadence, whether rising or falling, is determined by the position of the "strong" or "accented" syllable. This is to fol- low more particularly Mr. Omond's suggested classification {A Study of Metre) and it is Mr. Omond's whole theory that I have at this point most in mind; however the terms are, I think, widely enough used to be generally intelligible without special explanation and their relation to the more frequently used classical terms suffi- ciently obvious. The correspondence of the two sets of terms barely needs stating; what is here 52 A STUDY IN ENGLISH METRICS called verse in duple rhythm, rising cadence, is identifiable with the English verse called ''iambic": duple rhythm, falling cadence, with "trochaic": triple rhythm, rising cadence, with "anapsestic" : triple rhythm, falling cadence, with "dactylic." As to the correctness of the de- scription implied in either set of terms there is much question; as to the verse to be identified (even with the rejection of both descriptions) very little. Given these assumptions, my first point is that an exact study of verse-technique requires spe- cialized reference to each of the main kinds of rhythm, in their two varieties, since with the changes in the structural unit appearing as norm, the conditions of the problem change. This, which would be true even were a simple structure concerned, is in a complex entity of essential importance since with a change in any one of the arrangements a whole new series of inter-relationships is established. And, further, for the same reason, specialized reference to the main kinds of rhythm is not alone sufficient, but there must be such reference to each kind in rhythm as correlated with each A STUDY IN ENGLISH METRICS 53 kind in vocabulary. The primary classification as to rhythm — the final aesthetic condition — must, that is, be correlated with the primary classification of structural phonetic conditions in language — the medium — before a complete statement of the technical problem is possible. We must in each particular case know whether we have the rhythm, duple or triple, operating in relation to a vocabulary of the mono-dissyllabic type, or of the type of medium or of extreme structural complexity. Thus what would be, if English vocabularies showed a uniform struc- tural character, two main specialized conditions becomes, with the assumed differentiations, six main specialized conditions — (or twelve, allow- ing for the two varieties in each rhythm) . This evidently gives, in effect, a general scheme for the study of English verse technique, necessarily resulting from the assumptions made as to the principles of verse-structure as a whole, a scheme which at once summarizes within itself its determining principles and furnishes the basis for their further analytical testing. First illustrative application of the method suggested may be rapidly outlined with refer- 54 A STUDY IN ENGLISH METRICS ence to the problem which arises in connection with syllables carrying secondary word-accent, where the question at issue is whether these syl- lables can be used as "accented" (or "long" or "strong") syllables in verse. This has already been noted (p. 36) as a controversial question and that it has become so may fairly be regarded as indicating its importance. I am here con- cerned with noting the connections which must be established as a condition of its adequate treatment, not with the statement or discussion of the problem within itself. In approaching, then, the systematic study of the problem of secondary word-accent there must be, in accordance with the scheme outlined, first reference to the structural character of the vo- cabulary. Where there is a purely mono-dissyl- labic vocabulary the problem obviously does not exist; on the other hand in the extreme polysyl- labic vocabulary, where secondary accent- syllables reach their highest occurrence, the problem is present in its most acute form. In one case, to leave it for the moment in terms of extremes, the poet would have worked, nat- A STUDY IN ENGLISH METRICS 55 urally, with no reference to a non-existent condi- tion; in the other, reference to an acutely exist- ing condition would have been a constantly de- termining factor in the development of the tech- nique as a whole. Statement in exact terms of the occurrence of the problem requires, next, correlation of the rhythmic condition with the structural units of the vocabulary. Here I shall limit myself to the barest indication of the central issue given with reference to verse in duple rhythm, rising ca- dence ("iambic") and in triple rhythm, rising cadence ("anapsestic"). The fuller and finer statement of the initial problem and the question of its solution — the way, that is, in which Eng- lish poets have actually dealt with it — must be left for later, separate discussion. Concisely given the issue is this. When the verse is in duple rhythm (rising) the occurrence of every word over two syllables in length except mid-stress trisyllables will, if the "normal" dis- syllabic foot is to be kept, force the occuiTence of a syllable carrying secondary word-accent in the verse-accent place : e, g. 56 A STUDY IN ENGLISH METRICS To whom thus ^lichael. Death thou hast seen P. L. XI. 466. IthuneZ and Zephon with winged speed P. L. IV. 788. If on the other hand there is non-occurrence of the secondary accent-syllable in this place condi- tions arise which immediately bring up the ques- tion of the admission of variant feet. The most usual condition of non-occurrence is illustrated in the following lines : Saw where the sword of ^lichael smote, and fell'd P. L. VI. 250. Him thus intent IthuWeZ with his spear P. L. IV. 810. or to show both occurrence and non-occurrence in the same line : A second T>amel ! a DanzeZ, Jew ! Merch. of Ven. In triple rhythm verse (rising) the points to be noted are that non-occurrence of the secondary accent-syllable results in establishing a "normal" trisyllabic foot: A sensitive plant in a garden grew And the hyacinth purple, and white and blue The Sensitive Plant. A STUDY IN ENGLISH METRICS 57 or in establishing the condition illustrated in the following lines where there is evidently at least the possibility of a four-syllable foot to be con- sidered : The leaves they were -withering and sere Our talk had been serious and sober Ulalume. while the occurrence of the secondary-accent syl- lable in the verse-accent place must result in es- tablishing a variant dissyllabic unit: The snow drop and then the violet The water-blooms under the rivulet The Sensitive Plant. Thus, in the main, there is an important reversal in the terms of the occurrence of the problem with the change of rhythm. Systematic scrutiny of the problem, then, shows its absence or presence (and in what de- gree) as resulting from the structural character of the vocabulary. It shows that given the prob- lem as a whole an important difference in its oc- currence results from changes in the rhythmic condition and in so doing makes clear the close inter-connection at this point of the question of 58 A STUDY IN ENGLISH METRICS secondary word-accent with the question of the admission of variant feet. Seen from this angle there is obviously established an important con- nection for the study of the central matter of in- terest as regards the verse-form proper — the syl- labic variation occurring within the condition of rhythmic uniformity. And there is a further connection which may be just indicated — the connection with the problem of weighting. It is to be noted that there is here question of making syllables not only of secondary value in word- accent, but often in themselves extremely light syllables, carry verse-accent; and this, if it seem to present a difficulty in duple rhythm where a "normal" foot is concerned, will probably be felt to present an increased difficulty in triple rhythm where it will give a light syllable as the strong, or active, member of a foot already, in terms of the rhythmic norm, short of its full amount of sound material. This, however stumblingly put, will serve to indicate in what way a scheme of correlated ob- servations logically resultant from a consistent view of verse-structure as a whole will allow the posing of a whole problem, its limits defined, its A STUDY IN ENGLISH METRICS 59 connections established. And it is a little diffi- cult to see how, short of such schematic treat- ment, whether this or a righter one, it will ever be possible really to get anj^where. A second and if less systematic perhaps more immediately vivid illustration of the method sug- gested may be indicated in terms of the poems already analyzed. And here, lest I seem guilty of evasion, I must dare, if ever so tentatively, foreshadowing answer to that question which is after all the final test of validity — but how far is all this necessary as a basis of right judgment? That I cannot, in the present stage of the work, go beyond the tentative and the foreshadowing (and in rather informal manner) is too clear to need emphasis. What the 19th century thought of itself, what it thought of the 18th century, what we are to think of the 19th century, its work and its judg- ments — these are, I suppose, for whom such things exist at all, questions most alluring. And the allurement, unless one is the more deceived, is the ringing allurement of challenge. There is a present extraordinaiy aliveness in the air, a sharp exhilaration. We take poetry seriously 60 A STUDY IN ENGLISH METRICS and feel ourselves confronted with immortal issues. The names of Tennyson and Swinburne are two of the latest to be given great praise and lat- terly there has been some sound of dispraise. In the case of Tennyson a recent discussion gives for the matter this statement : "The hosts of criticism, until lately subdued to a common domination_, are now divided and there is no denying that the younger generation is distinctly hostile. And this at- titude represents a corresponding hostility among a consid- erable (and that not the least intelligent) section of the pub- lic. It is idle to ignore this antipathy, idle to pretend that it has not some real justification. . . . What then is the gravamen of the charge against Tennyson.^ It is not surely any allegation of technical insufficiency. Indeed one suspects that recognition of the poet's commanding tech- nical power adds not a little to the bitterness of the hostile feeling. It is rather as a 'thinker' (vile phrase !) that Ten- nyson is attacked" (Tennyson and the Critics^ The Spec- tator, Feb. 22, 1913). It is worth raising as a first question whether there is not in all of this an odd over-statement. Is there really on the part of the younger gene- ration hostility, or is it indifference? Though in- deed it makes an instant retort that precisely in- difference is of all hostilities, if the stillest, there- fore the most deadly. But whatever name it be given, the difference of attitude, the greater cool- A STUDY IN ENGLISH METRICS 61 ness or the lesser warmth, is the significant fact; to account for it the problem. If I now take refuge in an increasing use of the pronoun of the first person it will be recognized, I hope, as being not the I of arrogance but the I of humility and caution. Just what may be in the composite mind of a total hungry generation I should scarce venture to say ; I can but make sober and faithful report of what part of it is known to me as mine. And for Tennyson my answer to any such statement as that which I have just quoted would be this. It is not a matter of bringing "charges" against him but of finding out what his work really is ; and while there is not exactly "allegation of technical insufficiency" the ques- tion of technique is nevertheless at issue. Throughout the whole discussion cited as well as in the few sentences which I have quoted, the terms used in speaking of Tennyson's technique seem to me to show very great exaggeration, an exaggeration which results from the assumption of an illegitimate comparison. I quote another sentence in further illustration: 62 A STUDY IN ENGLISH METRICS "Far and leisured was the journeying which enabled him to master the principles of rhythm and modulation estab- lished by his great precursors and elaborate them to a^ pitch of perfection and variety which no other poet has ever equalled." And I add as expressing, but with greater defi- niteness, the same way of thinking, a detached clause from a sentence in Professor White's in- troduction to his study of the verse of Greek comedy: — ". . . just as the English heroic line, passing from Shakespeare to Milton and from Milton on to Tennyson, became under his magic touch a new instrument of melody modu- lated to every theme" (White The Verse of Greek Comedy, p. VII). The assumption, im- plied in the discussion of The Spectator, directly expressed by Professor White, is that there is a literal technical advance from Milton to Tenny- son. Now if my analysis is correct, and if the whole chain of reasoning so far presented is valid, it means that no such advance can exist, at least for the particular poems analyzed, since in them Milton and Tennyson work with reference to differentiated technical problems. Milton deals with the problems that I have indicated as inherent in a vocabulary of extreme structural A STUDY IN ENGLISH METRICS 63 complexity; his greater variety of word-forms imposes upon him all the difficulties of their manipulation, problems of weighting, of the management of the delicate, and treacherous, secondary accent syllables, and with these, since it is verse in duple rhythm, the question of vari- ant feet. These things if present for Tennyson are far less acutely present and with the change in the basic condition of the vocabulary, the whole weighting and balance of the line change. Any literal technical comparison between Tennyson and Milton seems to me, then, to fall to the ground. The proper comparison, as I make it out, is between Tennyson and Pope. I am of course aware that there are between these two poets important secondary differences, es- pecially, as to interlinear connection, the differ- ence between blank and rhymed verse, and the different management of grammatical pause; but these do seem to me secondary differences and unquestionably important as they are, I can- not quite see how they can be allowed in any sound criticism to obscure the perception of a primary likeness. Of Swinburne's craftsmanship there has been 64 A STUDY I N ENGLISH METRICS much superlative speaking and again, with all deference, it seems to me an exaggerated speak- ing. Comparison has not been made between his and the Miltonic technique in the specific matter of the decasyllabic (where I think no one asserts for Swinburne any absolute supremacy) and the question shifts to another ground. This I can best present by quoting two judgments lately expressed by Professor Gilbert Murray in his address — What English Poetry May Still Learn from the Greek: "Professed imitations of Greek rhythm in English poetry- seem to me to have gone practically always on quite wrong lines. They ought to have been more intensely rhythmical than the average; as a matter of fact they think they are being Greek when they lose lyrical rhythm altogether. Swinburne, as usual, as far as metre is concerned, gets triumphantly to the heart of the matter: She is cold and her habit is lowly, Her temple of branches and sods; Most fruitful and virginal, holy, A mother of Gods. That has a strong clear rhythm, full of majesty and sweet- ness. . . . But if you take, let us say, the most admired lyrics in Samson Agonistes: God of our fathers, what is man? That thou towards him with a hand so various. Or might I say contrarious, Temper'st thy providence through his short course, Not evenly, as thou rul'st A STUDY IN ENGLISH METRICS 65 The angelic orders and inferior creatures, mute. Irrational, and brute; or, This, this is he: softly a while; Let us not break in upon him. ... Or do my eyes misrepresent.^ Can this be he. That heroic, that renowned Irresistible Samson, whom unarmed No strength of man or fiercest wild beast could withstand? Who tore the lion as the lion tears the kid . . . This may be poetry of the highest order; I can quite im- agine that those who know it by heart even enj oy the rhythm of it. But surely it is clear that the rhythm is exceedingly obscure and utterly unlyrical in quality?" Professor Murray is in all these matters of the elect, "a Roman of Rome and very well thought of in Heaven" ; yet may I, even with a little de- corous impishness, suggest that his words con- cerning Milton have familiar sound . . . "one of the poems upon which much praise has been bestowed ... of which the diction is harsh . . . the numbers unpleasing" . . . and thus, remembering that the great Doctor himself on occasions erred, gather courage for disagreement? Setting aside the question of Greek metres Notes — In the quotations from Samson Agonistes I have kept the punctuation given by Mr. Murray in his printed paper {Essays and Studies by Members of the Eng- lish Association, Vol. HI, p. 27) . It is not the punctuation given in Canon Beeching's reprint of the edition of 1671. The absurdity "with a hand" must be (I suppose) a misprint. 66 A STUDY IN ENGLISH METRICS (about which I, all woefully, know nothing) and holding to the question of English metres, in English poetry after all undeviatingly the main concern, the issue raised by any such statement seems to me to be this. The lines quoted by Professor Murray from Swinburne are in triple rhythm and it is Swinburne's handling of this rhythm that, at any rate "as far as metre is con- cerned," is oftenest given the unqualified en- thusiasm of which I have spoken. It leaps to the mind in instant question — am I then to suppose that Swinburne's technique in triple rhythm is held to be, in general, comparable to the Miltonic technique in duple rhythm? It is for me an effort to phrase or squarely to envisage a notion so bewildering. Yet Professor Murray seems to make the comparison^and to the discomfiture of Milton. If imperturbability is for a moment wind-blown by gusty amazements it will be I hope forgiven me. After all there is difficulty in remaining imperturbable when one whose schol- arship imposes an all wistful deference is heard saying that he ''can imagine" it possible that one may "even enjoy" the rhythm of the Samson Choruses. For this long while the mere phrase A STUDY IN ENGLISH METRICS 67 "he knows Greek" has seemed to me, one de- prived and in exile, almost, for its potency, a magic formula; yet here, and for not the first time, I meet what I must believe to be a pro- found and subtle initiation in the aesthetics of Greek literature coupled with what I cannot force my mind to see as other than an amazing unawareness of the subtleties of the English forms. It is not possible to live at peace in the company of a bewilderment of this sort and at least I have tried to find out what I mean. What they mean, those more blessed others who "know Greek," I can only, with every real and humblest questioning, continue to ask. And what, in the present instance, I mean, since to justify my words I must here make re- port, is this. The general point made for Tenny- son and Milton in the matter of the decasyllabic will hold, within the limits of the analysis pre- sented, in the case of Swinburne and Milton with reference to the decasyllabic, but also, and more importantly, with reference to the different rhythms. Swinburne, that is, in the poems analyzed, works with vocabularies of the simpler sort, either decidedly mono-dissyllabic or of the 68 A ST UDY IN ENGLISH METRICS medium type with the lower occurrence of poly- syllables; and the different structural condition in the vocabularies makes impossible, it seems to me, for the reasons already given, any direct technical comparison between his problem in triple rh5i;hm and Milton's in duple. Further, this, as I make it out, has an impor- tant bearing on our understanding of the his- torical development of triple rhytlims in English poetry. Such a comparison as Professor Mur- ray's seems at any rate to imply that we find in Swinburne a poet using all the resources of the language; in the analysis presented it is seen that Swinburne works within a very limited range. Unless he can be shown to have made exclusions for special technical purposes, this, if it holds for his work in general, means I think that we find in him not a highly developed but an early technique. He has not mastered all the resources of the language; he has not even di- vined their existence. Between Milton and Swinburne there would then be this tremendously significant difference, that one stands at the end of a long sequence in development, the other at the beginning of a A STUDY IN ENGLISH METRICS 69 sequence. And it scarce needs the saying that there is here that which gives special direction to all the alertness of one's mind and sense. Suppose we are moving in the great main line of development in English poetry towards a mas- tery of triple rhythms really comparable to the Samson Choruses? Isn't that, then, the essential clue ? And without it where are we ? The bearing of all this on the final question of exact discriminations is, I think, clear. If it be true, as I have suggested, that we approach, by intelligible stages, a completer understanding of the whole complexity of English verse-structure, it follows that our scrutiny will contain within it- self a whole series of re-actions which, as felt not at all, or felt in some degree short of complete awareness, did not enter into the making of nine- teenth century judgments. It presents itself as a not unreasonable contention that, given its ex- istence, such an increase in fully conscious per- ception would be, necessarily, an essential factor in determining any critical re-estimation of values that may lie before us. An opposite opinion is, I know, possible; but I confess that there is for me an inherent per- 70 A STUDY IN ENGLISH METRICS suasiveness in the notion that finer comprehen- sions reflect themselves in finer and righter judg- ments. It may be well to add brief safe-guard against possible misunderstanding at one point, where the types of vocabularies are concerned the sole intent has been to discover and state possible dif- ferences. There is no question whatever of translating these differences into absolute terms of better and worse. That English poetrj^ may, in its vocabularies, possess a richness and com- plexity of three instead of a limited one is now the first consideration. The next, which must be left for far fuller treatment, will concern the special conditions of achievement or failure within each type. Here but a single point may be glanced at as inti- mately connected with the work just discussed. How far is it likely, I wonder, that the question of the long poem will become in this connection an important one? Evidently dependence on a mono-dissyllabic vocabulary for the short lyric is one thing; its use in a poem of the length of Chastelard another. A STUDY IN ENGLISH METRICS 71 And, finally, in dealing with observations of this kind it must be remembered that the poet's own awareness of his work, the extent to which his activity is felt to be consciously directed, is always a factor to be reckoned with. Again, rea- soned exclusions are one thing; exclusions that are accidental and quite unrealized by the poet himself, are another. NOTE The work I wish to do next deals with the problem of secondary-accent (pp. 53-57). The analysis is already completed except for general re-testing. It was in fact the analysis from which the generalization as to the differentiated vocabularies was originally derived. It includes Gascoigne's "Steele Glas," ''Tamburlaine," I and II, "Lycidas," ''Paradise Lost," "Samson Agonistes," "An Essay on Criticism," "The Rape of the Lock" and "Hyperion." The most generally interesting points, I fancy, would be the comparisons in technique between "Paradise Lost" and "Samson Agonistes" and between "Paradise Lost" and "Hyperion." It seems to me that statement of the technical difference at this point between the blank verse of Milton and that of Keats throws a good deal of light on Tennyson's work. 73 ADDITIONAL ANALYSIS While it cannot now be systematically dis- cussed, tabulated analysis is here given showing (a) for all the poems previously examined, the separate percentages for the occurrence of monosyllables and dissyllables (Tables VII- XII) and (b), for some few of the poems, the percentage of occurrence within the dissyllabic group as a whole of dissyllables accented on the last syllable. ( Tables XIII-XV. ) These tables, that is, contain preliminary data for the closer study of the mono-dissyllabic group hitherto treated as a whole, giving the dis- tribution of the two main types included within it, and, further, of the two sub-types included within the whole dissyllabic group. The closer study of the polysyllabic group must be carried out in the same way. This detailed analysis will give (as the pre- vious analysis for the larger groupings) exacter information as to the whole range of occurrence of types and sub-types considered separately, 74 ' A STUDY IN ENGLISH METRICS 75 and as to their relative occurrence in specific vo- cabularies. For instance, as to range of occur- rence, that there are "many" monosyllables in English means, in verse, according to these re- sults, a per cent occurrence running from the 71.43% of "Oenone" or the 73.39% of "Samson Agonistes" to the 89.46% of "Chastelard"— (and, of course, to the 100% found sometimes in short poems). Roughly, that is, it is probable that of the whole number of words in any Eng- lish poem, at least 70% are monosyllables. Again the fact that, in English, dissyllables ac- cented on the first outnumber those accented on the last syllable means, as here recorded, a range of occurrence for the rarer sub-type (within the group as a whole) running from the 35.30% of "Samson Agonistes" down to the 11.11% of "The Forsaken Garden"— (and to the 0% found sometimes in shorter poems). Tables giving separate percentages for the occurrence of monosyllables and dissyllables fol- low on pages 76-78. 76 A STUDY IN ENGLISH METRICS Table VII. Milton Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic Paradise Lost I 91.67 72.06 19.61 II 92.24 73.13 19.11 Ill 92.07 73.30 18.77 IV 92.74 75.23 17.51 V 92.01 73.41 18.60 VI 90.95 72.31 18.64 VII 91.40 73.14 18.26 VIII 91.45 74.65 16.80 IX 93.01 74.52 18.49 X 91.74 74.22 17.52 XI 92.48 74.51 17.97 XII 91.78 73.81 17.97 Total.... 92.03 73.75 18 28 Samson Agonistes Dialogue . . . 92.04 73.98 18.06 Choruses . . 90.92 71.72 19.20 Total.... 91.75 73.39 18.36 Table VIII. Pope Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic Essay on Criticism The Rape of the Lock . . . Elegy — Unfortunate Lady Essay on Man I II Ill IV Total Epistle to Dr. Arbuthnot 94.91 94.71 95.86 94.32 94.32 94.43 95.54 94.72 95.91 74.83 74.59 77.15 75.26 73.39 75.57 76.98 75.45 77.66 20.09 20.12 18.71 19.06 20.93 18.86 18.56 19.27 18.25 A STUDY IN ENGLISH METRICS 77 Table IX. Tennyson Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic Oenone Ulysses Tithonus The Coming of Arthur. Merlin and Vivien .... Lancelot and Elaine . . . The Holy Grail Guinevere , . The Passing of Arthur. 94.31 96.94 96.33 96.54 95.88 95.83 96.48 95.68 96.59 71.43 79.32 78.14 81.08 79.69 81.46 80.39 81.28 81.42 22.88 17.62 18.19 15.46 16.19 14.37 16.09 14.40 15.17 Table X. Swinburne Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic Chastelard I 98.59 98.56 98.65 98.30 98.18 98.42 95.84 96.83 96.14 97.50 97.03 98.80 89.13 89.18 90.17 90.22 88.57 89.46 82.48 81.50 82.19 82.35 81.76 84.05 9.46 II Ill IV 9.38 8.48 8.08 V 9.61 Total 8.96 Atalanta in Calydon Dialogue 13.36 Choruses 15.33 Total 13.95 Hymn to Proserpine Hesperia 15.15 15.27 The Forsaken Garden . . . 14.75 78 A STUDY IN ENGLISH METRICS Table XI. Francis Thompson Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic The Hound of Heaven An Anthem of Earth . . Sister Songs I II Total 92.61 90.59 92.02 92.11 92.06 76.18 71.30 74.41 73.40 73.72 16.43 19.29 17.61 18.71 18.34 Table XII. Maurice Hewlett Total per cent Mono- dissyllabic Per cent Monosyllabic Per cent Dissyllabic Minos, King of Crete .... Ariadne in Naxos Death of Hippolytus .... 96.41 95.92 95.96 81.08 78.57 79.22 15.33 17.35 16.74 Tables giving the percentage of occurrence within the dissyllabic group as a whole of dissyl- lables accented on the last syllables (type L) follow on pages 79-80. A STUDY IN ENGLISH METRICS 79 Table XIII. Milton Per cent Dissyllables in whole vocabulary Per cent type !- in total no. of Dissyllables Paradise Lost I 19.61 26.17 II 19.11 32.91 Ill 18.77 31.96 IV 17.51 31.94 V 18.60 33.20 VI 18.64 34,75 VII 18.26 30.04 VIII 16.80 39.17 IX 18.49 37.27 X 17.52 37.28 XI 17.97 37.38 XII 17.97 35.44 Total 18.28 34.05 Samson Agonistes Dialogue 18.06 35.16 Choruses 19.20 35.61 Total 18.36 35.30 Table XIV. Pope Per cent Dissyllables in whole vocabulary Per cent type - in total no. of Dissyllables Essay on Criticism 20.09 31.62 The Rape of the Lock Elegy — Unfortunate Lady 20.12 18.71 26.75 26.22 Essay on Man I 19.06 26.14 II 20.93 30.14 Ill 18.86 32.69 IV 18.56 30.18 Total 19.27 18.25 29.96 Epistle to Dr. Arbuthnot . . 27.28 80 A STUDY IN ENGLISH METRICS Table XV. Per cent Dissyllables in whole vocabulary Per cent type in total no. of Dissyllables Tithonus The Passing of Arthur . Chastelard Atalanta in Calydon . . . The Forsaken Garden . . The Hound of Heaven . Sister Songs Minos, King of Crete . . 18.19 15.17 8.96 13.95 14.75 16.43 18.34 15.33 20.18 18.97 21.79 20.42 11.11 19.69 19.40 16.29 TO- si^Y 14 DAY USE I RETURN TO DESK FROM WHICH BORROWED LOAN DEPT. This book is due on the last date stamped below, or on the date to which renewed. ' ^|^Df>rcki h9<>ks aw ^UlJect £p immediate recall. W^ : w immediate recall. : 13^67 1?; AM LOAN -V '■ m 9. 6 1981 flf l\ 5 1963 a a RECCIR MAR 13 '81 f-,ri^ JUNll'60-7^C; ^ AUG 16 1982 AUG 1 4 1982 ^ ii 1988 RECElVEr T^^rm -^ Vhiem^^f^^SS^^^^ ... Q*7n r ff g^l 1^ , /S^S JAN 2 B 1989 r!]I!in!C JUL 12 1990 U.C. ^£m UNIVERSITY OF CALIFORNIA LIBRARY