r PP\3597 ■-5S UBBABV OF CONGRESS 1111 „ , 014 1535833 # I 3 5 1 ■■ 12 • 13 ■■ 15 ■ 17 ■ 18 21 22 24 27 28 29 \\ LIBRARY OF CONGRESS illllil 014 153 583 3 SYLLABIFICATION AND A^ PR 3597 i.B8 ' IN THE (Copy 1 Paradise Lost A DISSERTATION PRESENTED TO THE BOARD OF UNIVERSITY STUL '"^^- OF THE JOHNS HOPKINS UNIVERSITY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY. BY GEORGE DOBBIN BROWN. BALTIMORE JOHN MURPHY COMPANY igoi ^ ^ !902 )) -^^ ^^^' 1^"^ 9 CONTENTS INTEODUCTION. Page. Metrical views of Bridges 1 Dr. Johnson 3 Guest 5 Poe 10 Ellis 11 Masson 12 Abbott and Seeley 13 Lanier 15 Schipper 17 Mayor ., 18 Gummere 21 Symonds 22 Editors of the Arden Shakespeare 24 Konig 27 Kobertson 28 Lewis 29 PART I.— SYLLABIFICATION. A. Within a Word. a. Synizesis and dieresis. 1. a before or after tonic vowel 34 2. a 4- atonic vowel 35 3. e before or after tonic vowel 35 4. e + atonic vowel 36 5. i before or after tonic vowel 36 6. i-\- atonic vowel 37 7. before or after tonic vowel 38 8. + atonic vowel 38 9. u before or after tonic vowel 38 10. u + atonic vowel 38 11. Other tonic combinations 39 iii CONTENTS. lopment of syllabic nasals and liquids and consequent synizesis with con- tiguous syllables 40 ' The syllabic nasal or liquid undergoes synizesis with atonic vowel : bordering > bordr-ing > bordring 40 1 40 m 41 n 41 r 42 II a. Development of syllabic liquidor nasalfrom vowel-^-liquid or nasal, and consequent synizesis with tonic syllahle : fallen >falln >falln...AO, 43 1 44 m 44 n 44 r 45 II b. Development of syllabic r from r + vowel, and consequent synizesis with tonic vowel : flourishing >flou-r-shing >flourshing 40, 46 r 46 y. Syncopation 46 B. Syllabification between Woeds. Customary elisions 47 Less usual elisions 48 Apocope and apheresis 49 PART II.— NUMBEE AND POSITION OF ACCENTS. Professor Bright's metrical theory 51 Symbols and terms 54 1-4. Light and Heavy iambs. 1. Light iamb 56 2. H'6avy iamb 56 3. Lig^ + heavy iamb — Conflict of stresses 56 a. Conflict between two words 57 6. Conflict between two syllables of one word 57 List of words like against used in descending rhythm 57 c. The relation of light + heavy iamb to triple rhythm 58 4. Erroneous scansion of light -f- heavy iamb by anapaestic postponement of arsis 59 5. Trochaic substitution 60 Trochee -\- heavy iamb compared to E * 61 CONTENTS. V 6-8. Unstable iamb in normal environment. Page, 6. Final unstable iamb 62 7. The non-csesural unstable iamb formed by one word 62 List of unstable iambs 63 8. The unstable iamb formed by two words 65 9-11. The unstable iamb in combination with other feet than stable iambs. 9. Trochee -f- unstable iamb 66 10. Unstable iamb -j- heavy iamb 67 11. Unstable iamb 4- unstable iamb 68 12-13. Light iamb -|- unstable iamb. 12. Light iamb + unstable iamb + stable iamb 68 13. Light iamb + unstable iamb + heavy iamb 69 The accent of various words 70 Life. 75 INTRODUCTION. In a recent number of The Academy Mr. Le Gallienne regrets that Sidney Lanier spent his precious time and energy in the composition of The Science of English Verse, " I wonder," he says, "whom these learned treatises on metre benefit. Not the poets, I am thinking. I imagine that Mr. Stephen Phillips would have written as good blank verse though Mr. Robert Bridges's treatise on Miltonic blank verse had never seen that dim light of publicity vouchsafed to technical masterpieces." We are not concerned with that part of Mr. Le Gallienne's criticism which expresses contempt of learned treatises on metre. Such deliverances on the part of critics of the class to which Mr. Le Gallienne belongs upon the futility of the scientific study of aesthetic phenomena delight, of course, the art-amateur for whom they are especially composed. But the conscientious investigator against whom they are directed is to an even greater extent amused. And so no harm is done. The part of the criticism to which I would call attention is its estimate of Bridges's essay, for Mr. Le Gallienne expresses the general judgment which has been passed upon that work. Mr. Bridges's treatise, Milton's Prosody, Oxford, 1894, is con- sidered an excellent exposition of the underlying principles of Miltonic blank verse. What, then, it may be asked, is the need of another such study? Metricians are agreed upon the type-line in blank verse. Dramatic blank verse differs from this norm often very widely, and may therefore be termed free. With such variations we are not concerned. Non-dramatic blank verse, on the other hand, as in Milton, Wordsworth, Tennyson, adheres much more closely to the model and may be termed strict. Even in strict blank verse, however, variations from the norm are frequent. It is 1 2 SyUabification and Accent in the Paradise Lost. with the nature of these variations from the norm in Milton's blank verse (both epic aad dramatic) that the nucleus (pp. 7-45) of Mr. Bridges's paper deals, all the fundamental questions being touched upon in the part concerning Paradise Lost (pp. 7-23). On page 6 Mr. Bridges defines a regular blank verse as having (1) ten syllables, (2) and five verse-stresses, (3) which fall on the even syllables. This is a satisfactory definition of the typical line. But a number of variations are allowed, and it is with the nature and extent of these variations that the metrician is chiefly concerned. With regard to (1), Mr. Bridges decides that, with the exception of a final light eleventh syllable making a feminine ending, there are no supernumerary syllables, those which apparently occur being obviated by elision. He is thus enabled to scan the verses of Milton without recourse to trisyllabic feet. In this we fully agree with him. There is required, however, a somewhat fuller account of this matter, which we shall attempt in Part I of this study. With reference to Milton's usage under (2) and (3), we hold that Mr. Bridges is altogether wrong. Under (2) he thinks we may have an unstressed foot or pyrrhic ( u u ). Thus he gets lines of only four, and even of but three, accents. It is strange that he does not in like manner go to the other extreme of finding / / feet with two verse-accents, as in this new world, and thus getting a line of more than five verse-stresses. Of course we do fre- quently have two unstressed, as well as two stressed, syllables used as a verse-foot. But in each case there is one and only one ictus (or verse-stress). Just how this is managed is considered in Part II. In the matter of (3) the distribution of the five verse-stresses, we shall find that a variation in the position of an ictus results in the substitution of a trochee for an iamb. The restrictions to trochaic substitution are three : — (a) It may not occur in the final foot. (6) It may not occur in two successive feet. (c) It may occur only initially and after a caesura. Syllabification and Accent in the Paradise Lost. 3 In these several aspects of (3), also, Mr. Bridges errs. A retreat- ment of this subject is therefore desirable. This we have attempted in Part II. Before passing to a study of Miltonic blank verse, we shall briefly indicate the teachings of various other metricians, some of whom are concerned with Milton in particular, while others discuss the subject of blank verse in general. It will be necessary to state the opinions of each upon the three points mentioned above : — (1) The number of syllables in the line, — -false theories allowing trisyllabic feet; (2) the number of verse-stresses in the line, — false theories allowing the typical number of five to be increased or diminished ; (3) the disposition of the stresses. Here two kinds of erroneous scansion are possible : {a) improper use of trochaic substitution ; (^8) just the inverse of trochaic substitu- tion is the method of scansion which may be termed anapestic postponement of the ictus. It is illustrated by _ X X / /, Pursues as inclina\tion or \ sad choice. II 524. This kind of scansion is never to be tolerated. The line must be read . .. ,x / X /. Pursues an inGlina\tion or \ sad choice. In his four papers^ on Milton's versification in the Rambler for the year 1751, Dr. Johnson makes no very explicit statement as to (1) the number of syllables in the poet's epic line ; but from two passages we may gather that he considered a hyper metric syllable to be allowable only when final. In the first of these passages he recognizes (though in general censuring) the large rdle which elision plays in Milton's verse. Thus for example he scans IV 249, and VII 236, If true, here only and of delicious taste. And vital virtue infused, and vital warmth. '" I believe every reader will agree," he says, referring to these and other lines, " that in all those passages, though not equally in all, the music is -injured, and in some the meaning obscured. There are other lines in which the vowel is cut ofi", but is so 1 Nos. 86, 88, 90, 94. 4 Syllabification and Accent in the Paradise Lost. faintly pronounced in common speech that the loss of it i» scarcely perceived ; and, therefore, such compliance with the measure may be allowed." Among examples of the less undesir- able usage he cites They viewed the vast immensurable abyss. YII 211 . Yet even such contractions, he thinks, increase " the roughness of a language too rough already; and though in long poems they may be sometimes suffered, it can never be faulty to forbear them." " Milton," he continues, in the second passage above referred to, " frequently uses in his poems the hypermetrical or redundant line of eleven syllables I also err^d in overmuch admiring. IX 1 1 78." These passages would show that the extra-metric syllable is,, according to Johnson's view, always to be obviated by elision,, except when forming the feminine ending to a line. With regard to (2) the number of stresses. Dr. Johnson says " In some [lines] the accent is equally upon two syllables together^ and in both strong; as [the last two syllables in] Thus at their shady lodge arrived, both stood. TV 720. ... In others the accent is equally upon two syllables, but upon both weak, [as the first foot of] And when they seek, as now, thy gift of sleep. TV 735." These usages, which Johnson considers to some degree ' licentious,' are examples of what he calls the mixed measure, " in which some variation of the accents is allowed ; this, though it always injures the harmony of the line considered by itself, yet compen- sates the loss by relieving us from the continual tyranny of the same sound, and makes us more sensible of the harmony of the pure measure." He does not explain how in these and like examples the arsis is distinguished from the other equally accented syllable in the foot, — how, in other words, these lines are not lines of six and four arses respectively. (3) Dr. Johnson sets a limit to trochaic substitution. He says- " In the first pair of syllables the accent may deviate from the rigour of exactness, without any unpleasing diminution of har- Syllabification and Accent in the Paradise Lost. 5 -mony . . . But, excepting in the first pair of syllables, which may be considered as arbitrary, a poet . . . should seldom suffer more than one aberration from the rule in any single verse." This expression does not rule out the final trochee, successive trochees, or the non-csesural trochee. The terminology used in Dr. Guest's History of English Rhythms, 1838 (Skeat's edition, London, 1882), is too unusual to be introduced here ; moreover the principle upon which this terminology is based, namely that the section is the unit of verse, has not been found serviceable by English metricians. It is easy, however, to point out what his system allows as to the (1) number of syllables, and (2) the number and (3) position of accents. (1) The number of syllables. Dr. Guest often obviates extra- metric, syllables. " The synalsepha or coalition [or synclisis] of two vowels," he says, (Skeat's ed. p. 68 f.), " is now tolerable in very few instances. . . . Formerly this union of vowels was far more general." Milton pushed this liberty, "as every other license, to the utmost." In the matter of what he terms elision, or what is more commonly known as syncopation (which, however, we believe to be more properly a species of synizesis), Dr. Guest approves of the printing of the earlier editions, which indicated when a word was thus shortened. " Of late years, however, the fashionable opinion has been that in such cases the vowel may be pronounced without injury to the rhythm. Thelwall discovered in Milton *an appoggiatura, or syllable more than is counted in the bar,' and was of opinion that such syllables 'constitute an essential part of the expressive harmony of the best writers, and should never in typography or utterance be superseded by the barbarous expedient of elision.' He marks them with the short quantity, and reads the following verses one with twelve, and the other with thirteen syllables ! u _ u Covering the beach, and blackening all the strand. Dryden. u u u Ungrateful offering to the immoi^tal powers. Pope." (p. 175 f.) We thus see that Dr. Guest would generally do away with extra-metric syllables in the strict heroic line, and he accordingly 6 Syllabification and Accent in the Paradise Lost. properly considers ominous as dissyllabic in II 123, Om(i)nous conjecture on the whole success. Thus too he considers ethereal trisyllabic and elides the e of the in I 45, X X / HurVd headlong flaming from \ tK ethe\real sky | . We are therefore much surprised to find him carefully indicating trisyllabic feet in : X X / II 450, Me from attempting where\fore do 1 1 assume. X X / X 121, So dreadful to theef That \ thou art na\hed who. and in a number of other lines. Especially noteworthy is this misscansion, X X / YII 592, Now resting, blessed and hal\low'd the se\venth day. when we note the subordination of the numeral in the similar lines VII 260, 338, 386, 448, 550. It is true that in YII 275 the necessarily dissyllabic form of second brings its first syllable into the arsis ; but in VII 592 the numeral must undoubtedly be scanned seventh and sink to the thesis of the last foot. (2) With regard to the number of accents in the line. Guest expresses himself explicitly, saying (Skeat's ed. p. 87) " Milton wrote his poem in verses ot five'^ accents. This remark is made with reference to the proper scansion of a series of equally import- ant monosyllabic words. Concerning this matter he says " When two or more words of the same kind follow each other consecu- tively, they all take an equal accent. If they are monosyllables, a pause intervenes between every two." Thus he finds proper juxtaposition of stress in every line of Sydney's sonnet in Arcadia, Lib. Ill, beginning . / / / Virtue, beautie and speech did strike, wound, charm, III My heart, eyes, ears, with wonder, love, delight. And yet he considers the metrical merit of these verses but small, — a rather severe judgment, as they follow his own rule. The Elizabethan interest in metrical questions and experiments makes it possible, but not probable, I think, that Sydney did intend this sonnet to be read with six accents (some juxtaposed) to the line, and indeed the verses are represented as having been Syllabifieation and Accent in the Paradise Lost, 7 " with some art curiously written ; " but it is more likely that they are pentametres and that their curious art is to be found in the heavy theses which are so frequently used as to give the lines a marked rhythmic effect. Guest considers that the strictly proper scansion of equally important monosyllabic words as strike, wound, charm is fre- quently violated, and that one or more of the monosyllables must lose the accent ; thus, for example, the words marked A i^i A / / A / / / Fear, sickness, age, loss, labour, sorrow, strife. Faery Queen, 1, 9, 44. This violation of the rule "is probably ... on account of the great number of English monosyllables," which practically often necessitates the subordination of some of them to the thesis-position. " False accentuation " of this kind, he continues, " very often leads to ambiguity. In High-climbing rock, low sunless dale. Sea, desert, lohat do these avails Wordsworth, White Doe of Rylstone. 7, 14. there might be a question, whether the author did not mean sea- desert, the waste of ocean. "When the words are collected into groups, this law of [the equal accent of a] sequence [of equally important monosyllabic words] affects the groups only, and not the individual. Thus I think there would be no fair objection to the mode in which Byron accents the verse, / , / Young, old, — high, low, — at once the same division share. Child e Harold 1, 71. Nor to Milton's famous line / / / / / Rocks, caves, — lakes, fens, — bogs, dens,^ — and shades of death. This last verse has been variously accented. Mitford accents the first six words, thus making a verse of eight accents, though Milton wrote his poem in verses of five." The important point in all this is that Guest allows only five accents in the line, and in order to keep down to this number ^ Guest makes this note: Den means a low wood bottom, such as often marks a stream . . .; hence it is coupled with bog. 8 Syllabification and Accent in the Paradise Lost. insists that prominent words must at times suffer an accentual subordination, which with regard to sense he considers to be more or less faulty according to the grouping of the words. Concerning the method of obtaining five arses when the prose- accent would furnish less than five, he says "There is no word, however unimportant, which may not be accented when it lies adjacent only to unaccented syllables." Thus / A murderer, a revolter, and a villian. Samson Agonistes, 1180. And in this line might have been placed in the same category. We thus see that Dr. Guest demands five and only five metrical accents or arses in the pentameter. (3) It is in the matter of the disposition of accents that Dr. Guest goes most astray, though it may be said to his credit that it is usually in the ease of really " hard " lines. His chief sin is that he allows double or successive trochaic substitution ; thus / X / X With im\petuous | recoil and jan-ing sound. II 880 / X /x Shoots in\visi\ble virtue e!en to the deerp. Ill 586. Also in V 413, 760, 874, VI 33, 333, VII 527, 533, VIII 299, XI 79, 377; and in P. R. I 175, II 180, 243, IV 289, 597. It is hardly necessary to say that the substitution of a single trochaic foot is fully recognized by Guest. Anapsestic postponement of the accent— just the inverse of trochaic substitution — Guest does not find in Milton, fortunately letting the laws of the metre override the natural prose-accent rather than making such a faulty inversion. Lines where this kind of conflict between the verse-requirements and the prose- / / accent occurs he considers faulty. So the prime for the prime in She was not the prime cause, but I myself. Samson Agonistes, 234. And many others.^ In V 158, / Thy goodness beyond thought and powW divine. ^ Of this fault, he says, even Pope was guilty : In words as fashions the same rule will hold. Essays on Criticism, 333. SyUabification and Accent in the Paradise Lost. 9 Dr. Guest seems to think that the accent beyond may have been / . / / / . used even in prose, as also besides, because, before, against, etc., found in other lines. It is to be noted, however, that such a prose-accent is now impossible, and yet Tennyson writes A life of nothing, nothing worth I From that first nothing before birth To that last nothing under earth. It is strange therefore that in Byron's Manfred, 2, 5, Guest scans with two consecutive anapaestic postponements of the accent, thus making the impossible reading X X / X / / _ Look on 1 me, the \ grave hath | not changed \ thee more. The proper scansion of this halting line is, of course, // // Look on I me, the \ grave hath \ not changed thee more. Guest's book is very valuable as a storehouse of examples of metrical phenomena, but the metrical theory which the book expounds has been rightly rejected. It is interesting, in this general dearth of insight into matters of prosody, to find half a paragraph containing a really penetrating remark upon the roles played in rhythm by pitch and expiratory accent ; but of the important fact here observed he makes no further use in his book. It is a fact, however, which has a most significant bearing upon the fundamental problems of English metre. After noticing on p. 75 that " though an increase of loudness be the only thing essential to our English accent, yet it is in almost every instance accompanied by an increased sharpness of tone," he points out (p. 80) that when an " emphatic syllable adjoins upon one, which ought, according to the usual laws of construction, to be more strongly accented," ... we may " dis- tinguish the emphatic syllable by mere sharpness of tone, and leave the stress of the voice, or in other words the essential part of the accent, on the ordinary syllable. Thus in Spenser's line Flesh can impair, quoth she, but reason can repair. F. Q. 1, 7,41. both the rhythm, and the common laws of accentuation will have the last syllable of repair accented ; but purposes of contrast 10 Syllabification and Accent in the Paradise Lost. require that the first syllable should be emphatic. The stress therefore falls on the last syllable and the sharp tone on the first." So he marks, with the pitch-accent the emphasis on us in Samson Agonistes, 309, Who made our laws to bind us, not himself. It is true that pitch-accent as here used by Guest to mark a rhetorically emphatic syllable falling in the thesis, does not immediately offer a means of reconciliation in the conflict between a fully accented syllable falling in the thesis, and a weaker syllable appearing in the arsis, — a phenomenon illustrated by the last foot of the line Which of us who beholds the bright \ surface \ . We shall see later, however, in the first pages of Part II, that by a slight extension in the use of pitch-accent we may find a means of marking as arses syllables which do not receive the chief prose-stress. This important metrical principle Dr. Guest would seem to have nearly discovered. Edgar Allan Poe's essay The Rationale of Verse has some good points in regard to English metre. His treatment of classical metres, however, is altogether futile. How could he have enjoyed Horace, when scanned as he suggests ; turning this line, for example, Mcecen\as, ata]vis \ edite \ regibus. into an accentual dactylic line — uu — uu — uu — uu ! Moreover, he entirely misses the meaning of synclisis and syni- zesis, whereby two vowels in contact may blend into a (monosyl- labic) diphthong without loss of their phonetic integrity. With reference to successive trochaic substitutions he says, " Where vehemence is to be strongly expressed, I am not sure that we should be wrong in venturing on two consecutive equivalent feet, although I cannot say that I have ever known the adventure made except in the following passage, which occurs in 'Al Aaraaf/ a boyish poem written by myself when a boy. I am referring to the sudden and rapid advent of a star," which is described in the midst of a series of iambic pentameters, as coming — KJ — O Headlong \ hiiher\ward o'er the starry sea. Syllabifidation and Accent in the Paradise Lost. 11 Poe deserves praise for laying emphasis upon one important metrical fact throughout his essay. I refer to the temporal equiv- alence of feet. Thus he says, "No feet .... differing [in time] can ever be legitimately used in the same line. . . . The point of time is that point which, being the rudimental one, must never be tampered with." Further on he speaks of the necessity of the "whole time of a foot being unchanged" as having been already established in the essay. Again he objects to the current books on classical prosody which declare "that the first foot of this species of verse [the verse of Horace's first ode] may be a trochee, and seem to be gloriously unconscious that to put a trochee in opposition with a longer foot, is to violate the inviolable principle of all music^ time." Mr. Ellis's views, as stated in his Essentials of Phonetics, Lon- don, 1848, his Early English Pronunciation, 1869, and repeated by quotation in a paper read before the English Philological Society in June, 1876, and published in its Transactions, are not quite so free as Mr. Masson's, to be presently considered, and yet admit licenses we should forbid. It is to be noted that Ellis is speaking about blank verse in general, not, like Masson, only of Milton's. With regard to (1) the number of syllables in the line, he admits trisyllabic feet. He also admits (2) doubly stressed or unstressed feet, thus both raising and lowering the number of stresses from the number in the type-line. (3) The occurrence of trochaic substitution is limited by these two sentences of 1848 and 1869, which, as Mr. Mayor points out, do not wholly agree. 1848 : "It is necessary that there should be an accent on the last syllable, either of the third and fifth meas- ures, or of the second and fourth. If either of these requisites is complied with, other accents may be distributed almost at pleasure." "1869 : There must be a principal stress on the last syllable of the second and fourth measures ; or of the first and fourth ; or of the third and some other. If any one of these three conditions is satisfied, the verse so far as stress is concerned is complete." That is, Ellis considers iambs are necessary only as follows : — ' The italics are mine. 12 Syllabification and Accent in the Paradise Lost. 1st. 2nd. 3rd. 4th. 5th. 1848'^' u — u — u — u — u — A glance at this table shows that Ellis allows final and suc- cessive trochaic substitution, and (since he makes no mention of an antecedent verse-pause as a necessary condition) non-csesural inver- sion. About all he forbids, indeed, seems to be, first, a succession of more than two trochees, and, second, double trochaic substitu- tion twice in one line. Mr. Masson in the introduction to his Poetical Worlcs of John Milton, London, 1874, takes a very free view of (1), (2), and (3). His ideas may be indicated by a few quotations. After scanning the first twenty-six lines of Paradise Lost, he says : " In thirteen lines the iambus is absent from the second place and we have a pyrrhic, a trochee, a spondee, or even an anapest, or an amphibracchius instead." "A study of the facts puts all formally right by declaring that English blank verse admits a trochee, a spondee, or a pyrrhic for the iamb in almost any place of the line. " Less numerous than the' lines that escape the strict 5 X « [== 5 iamb] formula by the substitution of the trochee, the pyr- rhic, or the spondee, for the iamb, but still very frequent, are the lines which escape from the formula by the bolder substitution of one of the trisyllabic feet. This occasions even greater irregularity in appearance ; for wherever an anapest, a dactyl, or a tribrach or other trisyllabic foot, displaces an iambus, the line, of course, is lengthened to eleven syllables. Nevertheless the trisyllabic varia- tion consists with the genius of English blank verse, and imparts to it an additional power and freedom. . . . All these lines might be rectified into decasyllabics by supposing elisions, slurs, or con- tracted utterances ; and there are some who seem to favour such a practise. There could be no more absurd error. . . . Are there ^And in one other foot at the same time. Syllabication and Accent in the Paradise Lost. 13 any examples of two trisyllabic variations in one line ? There are, though exceedingly rare. " Though five beats or accents are the normal measure of blank verse, yet the number of accents, unless in a peculiar sense of accent not realized in actual pronunciation/ is also variable. In a good many of the lines [in a list he has just given] only four dis- tinct accents can be counted. In three lines, I can detect but three ; and, on the other hand, in a very few spondaic lines the number seems to mount to six, seven, or even eight." Thus we see that Masson (1) lets the number of syllables exceed ten, (2) varies widely the number of stresses in a line, and (3) admits successive, final, and non-csesural trochaic (and its variant dactylic) substitution. In Abbott and Seeley's English Lessons for English People, London, 1879, the authors, in a chapter on modern English metre in general, say with regard to (1), "As a general rule it may be stated that the modern blank verse is, for the most part, more strict than that of Milton, and Milton is more strict than Shakspere, in lim- iting himself to ten syllables in a line. Milton uses capital, popu- lous, as trisyllabic feet." Frequently however an apparently supernumerary syllable is obviated, for "a vowel termination before an initial vowel is often elided in Milton." (2) Defect of accent is not allowed, for a syllable unaccented in prose may take, when falling in arsis-position, a verse-stress, thus — Nor served \it to\ relax their serried files. Similarly, excess of accent is not allowed, for a syllable, heavily stressed in prose, takes, when standing in theses-position, no verse- stress, thus — With head, | hands, wings, \ or feet, pursues his way. where hands, wings, though equally and heavily stressed in prose, furnishes but one arsis. Further, as a consequence of the two licenses above, it may happen that a light arsis falls directly before a heavy thesis, as in the I love-tale \ . " This sequence," described ii^ the English Lessons as being "so common in our best poets," Mr. Abbott had fre- ^ This half-admission is to be noted. 14 Syllabification and Accent in the Paradise Lost. quently tried in his earlier book, A Shakespearian Grammar, 2nd ed., Loudon, 1870, to obviate in a very violent manner, by consid- ering the heavy thesis to have become " lengthened " or else dis- syllabic, and thus to fill at once the place of the preceding arsis and of itself, causing the syllable which might have stood as a light arsis to become a thesis ; thus in the Grammar we find not Your breath first kindled the dead, coal of war. but X / X X / (X) / Your breath \ first kindled | the dea — | — d coal \ of war. This is what we have already named the anapsestic postponement of the arsis, — a method of scansion which is always erroneous. It is well that Mr. Abbott has seen fit to give this principle up ; and yet, without it, and having announced in his English Lessons that metrical accent, " if it fall on any syllable of a word, must fall on the principal word-accent," it is hard to see how he would scan the marked off feet in such a line as X 805, His sen\tence be\yond dust \ and nature^s law. where it seems necessary either to let the verse-accent fall on be-, or to make the extraordinary kind of backward inversion (anapaestic postponement of arsis) described above. Each of these methods Abbott seems to forbid. The first is of course to be followed without hesitation. (3) As to the restriction of trochaic substitution, Abbott and Seeley say " It may be laid down as a rule that in iambic metre one trochee cannot follow another. It is usual to quote as an exception / . / / / / Universal reproach far worse to bear. Such a line would be a monstrosity, and it is far more likely that Milton pronounced the word Universal, perhaps influenced by the fact that the a is long in Latin." Earlier, in the Grammar, Abbott seemed to think two successive trochees probable, though not quite certain ; and indeed he there cites / . / Uni\versal I reproach far worse to bear. Trochaic substitution in the final foot I do not find mentioned in the English Lessons. It is allowed in the Grammar. Syllabifieation and Acoent in the Pai^adise Lost. 15 That at least a slight verse-pause is required before trochaic sub- stitution, is explicitly stated in the English Lessons. It is also implied in such a remark as " "We may lay it down as a rule that a trochee in the middle of a verse must not follow an unemphatic accent. Hence it is allowable to write / / X Be in their flowing cups freshly remembered. because the emphatic word cups, long in quantity as well as emphatic, necessitates a kind of pause after it which makes a break between the two accents. But we could not so well write / / Be in their happiness \ freshly \ remember' dJ^ -ness being too light for a following pause. " The following seem to be remarkable exceptions : — / / Burnt after them to the \ boUom\less pit. and / / Light from above, from the {fountain \ of light.''' Here the rule that Abbott and Seeley have already laid down, that the metrical accent, "if it fall on any syllable in a word, must fall on the principal word-accent," makes it impossible for the authors to let the syllables -tom and -tain figure as arses, as of course they must. The fundamental note of Sidney Lanier's The Science of English Verse (Charles Scribner's Sons, New York, 1880) is this (p. 65) : " Rhythm of any sort is impossible, except through the co5rdina- tion of time. Time is the essential basis of rhythm. 'Accent ' can effect nothing, except in arranging materials already rhythmical through some temporal recurrence. Possessing a series of sounds temporally equal or temporally proportionate, we can group them into various orders of larger and larger groups ... by means of accent ; but the primordial temporalness is always necessary." The value of Lanier's book consists in the constantly recurring emphasis and luminous illustration of this essential truth. Lanier was an accomplished musician as well as a poet, and his insight into metrical matters was much increased and clarified by his musical knowledge. But there are at least some principles of musical rhythm which may not be applied to the rhythm of verse, and by making a faulty application of this sort Lanier fell into one of the few errors of his book. He says that a syllable 16 Syllabification and Accent in the Paradise Lost. upon which a verse-accent would regularly fall according to the rhythmic type, may either (a) be moved from under the accent, or (6) omitted altogether, letting its place in either instance be supplied by a temporal rest. This is indeed true of music. It is true also of the end of a line in verse. Thus, letting 1 repre- sent the temporal rest, we have, for example, a common hymn- stanza : / / / It is not death to die 1 I _ I I To leave this weary road 1 / III And 'midst the brotherhood on high I I . I To be at home with God -L. The use of such a rest in the midst of a line, however, is not to be allowed. Thus Lanier's method of scanning / X X / X (a) Who would believe \mef -\0 peril\ous mouths. X X / and (6) Than the soft myr\tle ; 1 \ but man \ 'proud man. is certainly erroneous. In the case of (a), where a syllable has only been removed from under accent, correction is to be made by replacing this syllable in the arsis. Thus either X / / X X / Who would believe \ me f \ peril\ous mouths. with a pause after and following trochaic substitution ; or, preferably, omitting the pause and using the secondary stress of perilous as arsis, / // / Who would believe \mef \peril\ous mouths. It may here be remarked that Lanier does not recognize the use of the secondary stress as a structural element in verse. " Of course," he says, " no one would read The mistress which I serve quic/cens what's dead." If indeed the use of the secondary stress as arsis is not absolutely necessary in this line on account of a possible pause after serve which would permit quickens, it is in some instances unavoidable, as in // Which of us who beholds the bright sur/ace. In the case of (6), where a syllable has been altogether omitted, the only method of scansion is with a monosyllabic initial foot : -X-X_// X- X — Than \ the soft \ myrtle ; \ but man, I proud man SyllahificaUon and Accent in the Paradise Lost. 17 Schipper, in his Englische Metrih (Bonn, 1881-1888), holds (1) that there may occasionally be extra-metric syllables. Thus, for example, he finds trisyllabic feet in To adore the conqueror, who now beholds. I 323. / / Whom reason hath equalled, force hath made supreme. I 248. (2) Two light and two heavy prose-syllables when forming a verse-foot, he correctly treats as having one and only one verse-stress. (3) When prose-stress demands that verse-stress (belonging as it does on the even syllables) should, to coincide with it (the prose- stress), fall on the preceding odd syllable, or when in other words there is a tendency to trochaic substitution, there are two methods of treatment according to Schipper : A. We may have trochaic substitution. This, however, he lets occur at improper places ; as when no caesura precedes, / / X . Of mart's \ first dis\obedience and the fruit. I 1. or in two successive feet, / X / X With them from bliss | to the \ boiiom\less deep. P. B. I 361. He forbids, however, trochaic substitution in the final foot. B. We may at other times have what Schipper considers to be in all cases a more or less faulty usage, hovering stress. The phenomenon of hovering stress is this. The arsis does not, as in A, pass forward so as to coineide with the word-stress, but instead the stress hovers (which I take to be equivalent to rests practically level) on the two syllables, with, however, a slight distinguishing accent on the second syllable or arsis. Hovering stress is therefore best available when the two syllables have in prose almost level stress ; thus moonlight, with slightly out- weighing initial stress in prose, easily goes into the iambic line / as moon-light. Schipper, however, uses hovering stress even when there is more disparity between the two syllables, thus, / Burnt after them to the botto^/iless pit.^ ^Strangely enough, Schipper suggests a different scansion for P, M. I 361, With them from bliss | to the \ bottom\less deep. using double trochaic substitution. 18 Syllabijication and Accent in the Paradise Lost. There are thus, according to Schipper, two methods of reconcil- ing conflict between the accentual systems of prose and verse when the syllables showing the conflict fall in the same foot. When, however, the conflict is between a (light) arsis and a following (heavy) thesis — the thesis of the next foot, — we can have recourse only to hovering stress, as an inversion (liere anapaestic postpone- ment of the arsis — just the reverse of trochaic substitution) would be in this case impossible. An example of this is beyond in X 805, X / X / His sen\tenoe be\yond dust \ and Nature's law. In criticism of the doctrine of hovering stress as expounded by Schipper, we may say that it satisfactorily explains the double verse-forms man-kind and man-hind, etc. ;' and even compounds / / of the nature of unknown and unknown. For here the practically level prose-stress takes with the greatest ease, at the suggestion of the environment, the requisite ascending or descending rhythm. Hovering stress will not, however, explain how bottom, which, though indeed possessing a minor stress on -torn, is by no means levelly stressed, can be used as an iamb ; nor can it explain how beyond may show descending rhythm. For preponderating stress on -tom or be-, even though only slightly outweighing the other stress, would give to the word an improper and abnormal form. But this is just what hovering stress demands. Indeed Schipper agrees with Masson that / Burnt after them to the bottomless pit. " is too horrible ; and such barbai:ous readers are imaginary." And yet Schipper's scansion really results in the same thing, for in order to make an iamb out of bottom upon the principle of hovering stress, the last syllable must preponderate however little over the first. We shall see hereafter that the only proper treatment of such a word is to let the (expiratory) prose-accent remain on its proper syllable, and to indicate the other syllable as arsis by a specific verse-accent the characteristic of which is pitch. Let Mr. Mayor, the author of Chapter on English Metre, London, 1886, express himself under the three heads. He considers blank verse in general, not Milton's especially. Syllahificcction and Accent in the Paradise Lost. 19 (1) Supernumerary syllables he admits, but he limits them with these words : "As to trisyllabic substitution, it is plain that if we set no limits to this, the character of the metre is changed. . . . I think the limit of trisyllabic substitution is three out of five [feet]." " We must distinguish, however, between the different kinds of trisyllabic feet. The anapest, which may be considered an exten- sion of the iamb, is the most common ; the dactyl, which is similarly an extension of the trochee, is only allowable, I think, in the first and either the third or fourth foot ; e. g., we may say, Terrible \ their approach \ terrible | the clash of war, or Terrible \ the clash | of war \ terrible \ the din. Whether amphibrachis, i. e. iamb followed by an unaccented syllable, can be allowed in any place, except, of course, cases of feminine rhythm, is perhaps doubtful. I cannot remember any parallel to a heroic line such as the following, but I see no objec- tion to it : — / Rebounding \from the rock | the an\gry brea\kers roared.^^ in which the first foot is an amphibrach. "I have no doubt that we must recognize the substitution of tribrachs for iambs in English blank verse," as in / X X / Melody on branch and melocZ^/ *^ midair. Tennyson, Gareth. (2) " With regard to . . . accentual irregularities, excess of accent, ^. e. spondee, is allowed in any position, and I am inclined to think that . . . there might be four spondees in the line, supposing that the fourth or fifth foot remains iambic." " Defect of accent, the pyrrhic, may also be found in any posi- tion, but it is rare for two pyrrhics to come together, and perhaps impossible without a secondary accent falling on one of the syllables." He thus allows, erroneously, we believe, variation in the typical number of accents. (3) "I should be inclined to say that the limit of trochaic substitution was three out of five, provided that the final syllables remained iambic, otherwise two out of five." Thus we can get not only two but three successive trochees ; as well as final and non-csesural trochees. 20 SyUahification and Accent in the Paradise Lost. Mr. Mayor seems to favour anapsestic postponement of the arsis. For on p. 637 of The Transactions of the Philological Society (London), 1873-1874, in commenting on Abbott and Seeley'a English Lessons for English People, he finds fault with the scansion^ Oh, weep for Adona|is. The \ quick dreams. / Then tore with bloody ta|lon the \ rent plain. and says "It is plain that in these lines the is about the least important word, and is intentionally prefixed to the important words quick and rent to give them additional emphasis. In technical language, the is here a proclitic; so far from laying any stress upon it, a good reader would pass it over more lightly than any other word in the lines. I am unable, therefore^ to see the propriety of speaking of it as bearing the metrical accent." This scansional system Professor Mayor has applied to ShelleyV verse in a paper entitled Shelley's Metre, London, 1888, Printed for private subscription. (1) Professor Mayor would find a trisyllabic foot (anapaest) in this line from The Cenci: X X / To rend and ru-m. What say ye now, my Lords ? where ruin is metrically monosyllabic. (2) Another mode of varying the metre, thinks Professor Mayor^ is the adding to or the taking from the number of accents in the foot. This, he says, is " so common that the exception is to find a line which does not contain feet with either no accent or more than one accent. Thus / / X X / / . . The dry | fixed eye-lball, the | pale quivjering lip. has two spondees and one pyrrhic, altogether six instead of five accents." The ictus, we think, must really be placed as follows :; II I I I The dry | fixed eye|-ball, the | pale quiv|ering lip. (3) Great license is given to trochaic substitution. Thus Pro- fessor Mayor allows trochees to stand where not preceded by a caesura : / And wild I roses \ and ivy serpentine. in the last foot : SyUabification and Accent in the Paradise Lost. 21 / And heard the autumnal winds like light \ footfalls \ . and successively : / / If thou hast done | murders | , made thy \ life's path. The metrical accents on the words above italicized must, we hold, III I fall thus : roses, footfalls, murders, and perhaps made thy. The licenses of trochaic substitution and trisyllabic feet are frequently combined by Professor Mayor and he gives us such a scansion as : / X X That ever came | sorrowing \ upon the earth. This we must correct to : / That ever came | sorrowing \ upon the earth. In his article " The Translation of Beowulf and the Relations of Ancient and Modern English Yerse" in The American Journal of Philology, yol.YIl, 1886, pp. 46 if., Professor Gummere shows, naturally enough it seems when we note the general theory he is advancing, a tendency to underestimate the importance of certain of the arses in the heroic line. He believes that " we must regard Chaucer's great creation, heroic verse, as really founded on the national verse [of four stresses], not as an importation." That is, *'our heroic verse ... is simply the result of forcing the iambic movement (influence of foreign models played its part here) upon some late form of our four-stress verse. This process, in a word, reduces the pause [at the middle of the native line], and cuts down all triple measures (aside from cases of slurring); but it adds a new verse-stress, though this, in a majority of cases, has no syn- tactic force. Such a verse as Chaucer's ( C. T. 500) That if gold ruste, what shal yren do f is itself almost enough to support the above statement ; note the pause, the balance, the real movement : uu — — u|uu — u — as compared with the movement of the verse-scheme : u — u — u| — u — u— . Again, "The majority of heroic verses may be said to have really but four stresses. Further, any reader of English verse will remember a certain tendency, notably strong with Dryden, •22 Syllabification and Accent in the Paradise Lost. Pope, and Johnson, to balance lines in such a way that the verse falls in halves, with a slight pause after the second measure, or in the middle of the third (masculine or feminine pause), with a very weak third or fourth stress which in the case of a feminine caesura gives the effect of heaped-up light syllables.^ No secret island in the boundless main. Refund the plunder of the beggar'd land." In other words. Dr. Gummere, though not quite consistently with the earlier-quoted passage in which the expression " adds a new verse-stress " would make in and of arses, seems here to consider the proper reading (and if the reading of a line is not indicative of its scansion, what is ?) to be : / / . / /. No secret island ± in the boundless main. II II Refund the plunder 1 of the beggar'd land. where the dash represents a slight pause which compensates for the absence of the third arsis. The proper scansion, of course, raises in and of into the arsis position by pitch-accent. In his essay "An Answer to the Question What is Poetry?" in Imagination and Fancy, London, 1844, Leigh Hunt's remarks on metre are confined to purely aesthetic judgments, and upon this aesthetic basis he does not attempt to raise a system of the mechanics of verse. Symonds, too, in his Blank Verse (Scribner's Sons, New York, 1895), writes mainly from the aesthetic viewpoint. In his tech- nical exposition of verse he allows, we believe, too great freedom of scansion. " Successive trochees in the third and fourth places," he says, " of which there are several specimens in Milton, , . . are far from disagreeable in the English iambic." Again, while not so expressing himself explicitly, it is plain from his words that he scans two well-known lines from Samson Agonistes either with two successive trochees initially / / For his \ people \ of old : what hinders now f ^ Cf., says Gummere, Eieger oa the second auftact : Im Beowulf ist, wie iiberall, die anschwelluDg des auftactes mehr im zweiten als im ersten halbvers zu hause. AU- und Ags.- Verskunst, p. 59. Syllahificntion and Accent in the Paradise Lost. 23 and, TJni\versal\ly crowned with highest praises. or with only four feet, the first two of which would be trisyllabic : . . / / For his peo'j)le of old | lohat hinders now f . I I Univer\sally crowned \ with highest praises. In the proper scansions, / / 1^1 For his | people, etc., and Uni\versal\ly, etc. the propriety of / is witnessed by Ill 642, Of many a coloured plume sprinkfd with gold. Y 750, In their triple degrees, regions to which, while the propriety of -al is shown by III 33, Those other two equall'd with me in fate. IX 370, But if thou think tria^ unsought may find. Again, avoiding obvious synclisis and syncopation he seems to scan this line from P. L. with two trisyllabic feet : XX / X X / Submiss ; he rear'd me, and, Whom thou soughtesf I am. So the line from P. B. The one winding, the other strait, and left between. " affords," he says, " a good instance of what is meant by the massing of sounds together, so as to produce a whole harmonious to the ear, but beyond the reach of satisfactory analysis by feet. It is not an Alexandrine, though, if we read it syllabically, it may be made to seem to have six feet. Two groups of syllables — The one winding \ the other strait \ — ^take up the time of six syllables, and the verse falls at the end into the legitimate iambic cadence. At the same time it would no doubt be possible, by the application of a Procrustean method of elisions and forcible divisions, to reduce it to an inexact iambic, thus : X / X / TA' one win\ding th' ot\her strait and left betiveen." Thus he sees what we consider the true method of scansion only to condemn it. A good example of the defect of Symonds's treatment of metrics, a defect always present in a system which is founded upon a sub- jective judgment of the scansion required by the meaning rather 24 SyllabijiGation and Accent in the Paradise Lost. than upon regard to the mechanical structure, may be seen in the following sentence, which, though having nominally to do with dramatic blank verse, really expresses his teaching upon the scan- sion of blank verse in general : " The one sound rule to be given to the readers of dramatic blank verse, written by a master of the art, is this : Attend strictly to the sense and to the pauses ; the lines will then be perfectly melodious ; but if you attempt to scan the lines on any preconceived metrical system, you will violate the sense and vitiate the music." Recent essays on metre have been appearing in the Arden Shakespeare, an essay being appended to each separately edited play. I have examined those in the volumes containing Richard II, Richard III, Hamlet, and Cymbeline. Dramatic or free blank verse differs from Miltonic or strict blank verse chiefly in this, — while the latter allows an eleventh syllable only at the end of a line (making thus a feminine ending), the former may have a trisyllabic foot at any place in the line. Criticism of the one kind from the syllabic standpoint is not, therefore, wholly applicable to the other. (1) But though trisyllabic feet are allowable in free blank verse, nevertheless an hypermetric syllable is frequently elided or other- wise obviated even in the dramatic line. Thus Dr. Herford, the editor of Richard II in the Arden edition, scans Sent back like Hallowmas or shortest of day. Stoop with oppression of their prod(i)gal weight. Let it be tenable in your silence still. (Hamlet.) I I Reproach and dissolution hangHh over him. What says your majesty ? Sorrow and grief of heart. with an extrametric syllable in majesty before the csesura. I have been studying how I may compare. " Sometimes the number of syllables is less than the normal ten, the stresses remaining jive. This happens especially after a marked pause. . . . But it hardly became a regular type [i. e. when it occurs it is felt as unrhythmical). E. g. Your grace mistakes ; only to be brief. Ill 3, 9." Syllabification and Acoent in the Paradise Lost. 25 , (2) The necessity of five stresses to the line Dr. Herford puts well. He says "The essential structure of Shakespearian blank verse ... is a series of ten syllables bearing five stresses. The words stress and non-stress are here used for the metrical ictus^ or beat, and the pause between. It is essential to distinguish the series of stresses and non-stresses which form the rhythm, from the word- and sentence- accents which are accommodated to them. . . . The stresses may vary in degree ; syllables which bear a very slight natural accent being placed in a normally stressed place. Thus, To scarlet indignation, and bedew." (3) In regard to trochaic substitution he says, " Within limits, the alternate order of stress and non-stress may be inverted. As this causes two stresses to come together, and as two stresses can only be pronounced in succession when a slight pause intervenes, this inversion commonly coincides with a pause in the sense, and is thus found most often at the beginning of a line, [or] in the 3rd or 4th foot, sense-pauses commonly occurring in these places. . . . In the 2nd foot it is much less usual. Konig has reckoned that it occurs 34 times in Shakespeare in the 2nd place, against c. 500 in the 3rd, and c. 3000 in the 1st. ... In the fifth the inversion has hardly become typical (i. e. when it occurs it is felt as unrhyth- mical). . . . Two inversions may occur in the same line, . . . but we rarely find two inversions in succession and never three." Some of the other Arden essays are less satisfactory than Dr. Herford's. Thus (1) Dr. Chambers, editor of Hamlet, prefers tri- syllabic feet to the obviously proper scansions 1 am more an antique Roman than a Dane. Let it be tenable in your silence still. The light and careless Wwery that it wears. So Dr. Wyatt, editor of Cymbeline, prefers a trisyllabic foot to the scansion How bravely thou beGom\{e)st thy bed, \ fresh lily. Dr. Wyatt says, in regard to the (2) number of stresses in a line, that we may have feet of the form a a and xx, when a indi- ^The phonetic or acoustic nature of the metrical ictus, however, Dr. Herford does not explain. 26 Syllabification and Accent in the Paradise Lost. cates a stressed, x an unstressed, syllable. He gives us, however, no idea as to how one of these syllables is to be distinguished above the other as arsis. Dr. MacDonald, editor of Richard III, recognizes the necessity of distinguishing between two equally accented syllables and says in regard to feet of the type a a, though the remark holds equally good for the foot xx, "The same foot cannot . . . contain two stresses of precisely equal strength. The beat of the rhythm must be distinctly perceptible, if the line is to be metrical." (3) Position of stresses. Dr. Chambers allows too great license, we believe, to trochaic substitution, inverting without a sufficient preceding verse-pause the second foot in The wind \ sits in \ the shoulder of your sail. Hamlet, I, 3, 56. But, as Dr. Herford remarks, " it is probable that [in Elizabethan English] both prepositions and the definite article [and the indefi- nite, too?] often bore a stronger accent than now." Dr. Chambers makes two consecutive trochaic substitutions, and final at that, in Hamlet, I, 3, 101, / / Affection ! pooh ! you speak \ like a \ green girl \ . Dr. MacDonald gives a double trochaic substitution initially iu Richard III, III, 1, 188, / / Shall we \ hear from | you, Catesby, ere we sleep? And fearing to accent the preposition in IV, 1, 56, he scans A cockatrice hast thou \ hatched to \ the world. making an inversion without a sufficient pause before it. Dr. Wyatt makes an undesirable initial inversion and omits a desirable synizesis — an omission which necessitates a trisyllabic foot — in Oymbeline, II, 2, 20, / X X / X X / Bows to\wards her, | and would Mn|der-peep her eyes. Which should, we believe, be scanned / / Bows towards \ her, and \ would wn.|der-peep her eyes. He makes the forbidden final inversion in IV, 3, 9, / / The hope of comfort. But for thee \ fellow \ . Perhaps his most unsatisfactory scansion is the anapaestic post- ponement of the arsis which he finds in II, 2, 18, Syllabification and Accent in the Paradise Lost. 27 X X / / How dear|/2/ they \ doH ! ^Tis \ her breathing that. It will be observed that this scansion leaves the second foot without an ictus, while throwing two verse-accents into the third foot. Properly scanned, the ictus falls on they, of course, and not on doH: II I How dearjZ^/ they \ doH ! ^Tis \ her breathing that. Goswin Konig's treatment of Shakspere's verse, Der Vers in ShaJcsperes Dramen, Quellen und Forschungen, LXI, is seriously- vitiated, we believe, by such scansions as these : Thou hast done a deed, whereat | valour \ will weep. Coriolanus, V, 3, 134. / We shall be called | 'purgers \ not murderers. Julius Ccesar, II, 1, 180. / My heart \ prays for \ him, though my tongue do curse. Comedy of Errors, lY, 2, 28. In the above lines and many others, Konig finds himself obliged to make a trochaic substitution where there is no preceding verse- pause, due to the fact that he ignores the r6le played in verse by secondary stress which by a slight heightening of its pitch is fully capable of marking the rhythmic wave. The italicized words in // // // the above lines should bear the ictus thus valour, purgers, prays for. In a similar way the double trochees are to be altered in the faulty scansions : I. . I Villains, \ answer \ you so the lord protector ? Henry VI, Ft. I, I, 3, 8. TaJce the \ boy to \ you : he so troubles me. Winter's Tale, II, 1, 1. and others ; which we must read with the ictus so placed : // Villains, | answer | you so the lord protector ? Take the | boy to | you : he so troubles me. etc. Konig properly finds a final trochee very disturbing to the rhythm, and such a trochee he always attempts to obviate. This he does at times by considering that a thesis has fallen out earlier 28 Syllabification and Accent in the Paradise Lost. in the verse and that the trochee is really only a feminine ending, thus : / (^) /x Appear in person here | in court. | Silence \ . Winter's Tale, III, 2, 11. At other times he does away with the trochee by having recourse to hovering stress : The gods rebuke me, but it is tidings. Antony and Cleopatra, V, 1 , 27. But why not avail one's self of the secondary stresses inherent in the second syllables of these words and read J I Appear in person here in court. Silence. . . . 11 The gods rebuke me, but it is tidings. While Professor Gummere seems to take the method of scansion / / , / / , No secret island — in the boundless main. merely as a manner of reading, the strict verse-scheme really requiring the time of the dash or rest to be filled by the following unimport- ant word, Mr. John Mackiunon Robertson in an appendix entitled Accent, Quantity, and Feet in his Nev, Essays towards a Ot^itical Method, John Lane, London and New York, 1897, elevates this method of reading into a metrical type. A large part of this appendix — an appendix to the essay on Poe — is given up to a review of " the endless discussion " " as to the scansion of classical verse," — a discussion which justified Poe, "convinced of the hope- lessness of resort to academic tradition," in his endeavour " to construct a metrical system on purely rational grounds." In the not always very lucid criticism of Poe's system which follows, Mr. Robertson gives expression to some of bis own metrical ideas. " Poe," he says, " gives us the scansion : u— u— u — u— u — A breath \ can make | them as | a breath \ has made, making five iambs. This is rhythmically false ; no good reader would scan so. He would read, pedally speaking : u— u— uuu— u — A breath can make them | as a breath has made, and if we are to be logical in our pedal notation we should make the line one of two feet of five syllables each, as some of the ancient grammarians would actually have made it. What has Syllabification and Accent in the Paradise Lost. 29 really happened is a pause after them, making the real scansion run u— u— u Ou— u — A breath can make them as a breath has made, and so keep up an iambic movement, with the slight variation of one anapaest. And this is the cue of the rhythmical essence of the first line of Paradise Lost u — u— uo— u — o u — Of man's first disobedience and the fruit. That is to say, the movement is essentially iambic, with the easy variation of two anapaests. ... It may be argued that this method of analysis will serve to certificate the most unrhythmical verse at will as rhythmical ; but this is a needless fear. A bad verse is so because it puts in a continuous run diversities of step which disconcert us; and no aid from pauses can cure such jars. We may, by pausation, make a risky line of Milton nearly quite rhythmical, as in u — u— u u— uu — uu — Burnt after them to the bottomless pit ; ^ where the undue stress on them is now the only flaw. But no pausation can cure such lines as Mr. Lowell's Uu — uu — uu — uu — Forty fajthers of free|dom of whom ] twenty bred ... u '-' T "-'^ ~ V^ *-* u— Ouu— u Each has six | truest patjriots four discovjerers of e|ther, where the anapaestic intention stumbles as if in epilepsy."^ The objection which Mr. Robertson himself suggests may be made to his system forms its real condemnation : " This method of analysis will serve to certificate the most unrhythmical verse at will as rhythmical." Anything is allowed ; and we cannot but feel that the severe criticism of at least the first of the above lines from Lowell is due rather to Mr. Robertson's animosity towards that critic, plainly shown in his essay on Poe, than to the fineness of an ear which can tolerate the scansions / / / . /. Of man's first disobedience and the fruit ... , III I / Burnt after them to the bottomless 'pit. Dr. Charlton M. Lewis in his thesis entitled The Foreign Sources of Modern English Versification {Yale Studies in English, No. 1, Halle, 1898), covers extensive ground, tracing the rhythmic princi- ^I imagine (but who can tell ?) that ^ between after and than is a misprint. "In reality, the first of these lines is tolerable. 30 Syllabification and Accent in the Paradise Lost. pie from " the syllabic verse of some of the oldest parts of the Avesta/' through the quantitative classical metres, the Mediaeval Latin hymns, and early French verse, to English soil. Only in his last section, § 51, The syllabic principle in Modern English verse, does he express opinions which directly concern our modern pentametre. Our five foot verse, he says, "affords some recogni- tion — though feeble — to the syllabic principle." In a strict and thorough-going application of this principle the integrity of the number of syllables would be the one thing needful. It is this principle which, partially applied, justifies, thinks Dr. Lewis, the two consecutive trochees which he finds initially in / . / Palpi\tated, \ her hand shook, and we heard. The Princess, IV 370. and medially in Paradise Lost, X 178, / / And dust shall eat | all the \ days of \ thy life. and the three consecutive trochees in / / . / Harmon\izing \ silence \ without a sound. Epipsychidion (near the end). and, Belus \ or Se\rapis \ their gods, or seat. Par. Lost, I 720. The harshness of these inversions entirely destroy the iambic rhythm. These lines must of course be scanned / . // Palpi\tated, \ her hand shook, etc. II II And dust shall eat \ all the | days of | thy life. IIarmo\nizing \ silence \ without a sound. while in Par. Lost, I 720, Milton evidently said Serapis, cf Part II, The accent of various words. As Dr. Lewis considers that the presence of the syllabic princi- ple is shown clearly in the inversions he finds in the above lines, so he thinks it is conclusively proved in cases of anapceslic post- ponement of one of the five accents. This happens in "all those verses which are commonly described as containing pyrrhics followed by spondees." Thus he says " the line from The Cenci, The house-dog moans and the beams crack; nought else. Ill 2. may be divided into feet as follows : SyUahijioaiion and Accent in the Paradise Lost. 31 The hou8e\-dog moans \ and the | beams crack \ nought else, the third foot being a pyrrhic and the fourth a spondee. But if it be compared with the following, The house\-dog moans \ and the beams \ are cracked, it will be seen that the reading of the first four feet is exactly the same in the two cases, so far as time and stress are concerned.'^ The latter line is in the familiar Christabel metre, and the third foot is an anapaest : — so we see that the pyrrhic in the former line is not a foot at all, except to the eye. It seems doubtful wisdom, therefore, to try to reduce such verses to regularity by dividing them into dissyllabic feet at all. The better explanation is that while the postponement of the third accent is a variation from the normal, the strict observance of the syllabic rule keeps the verse within bounds." For a criticism of this improper scansion see Part II, 4. For the proper method of scansion : // / The house-dog moans \ and the \ beams crack \ ; ncmght else. see Part II, 3 a and 6. Mr. H. C. Beeching's criticism in The Athenceum, June 1st, 1901, of Professor Bright's method of scansion calls for a word in reply. The fundamental objection which Mr. Beeching finds to Dr. Bright's system is the acceptance of such accentuations as / / . / / . amorig, beyond, and many, parent. It is not now necessary to dwell upon the historical justification of such forms. In Part II this shall be done at some length. For the present, let us meet Mr. Beeching on his own ground, and, like him, appeal to the judg- ment of the unphilologieal reader. Would not this personage be inclined to accept the accentuation among, before in the following stanzas of Wordsworth and Tennyson ? I travell'd among unknown men In lands beyond the sea ; Nor, England ! did I know till then What love I bore to thee. — A life of nothing, nothing worth From that first nothing before birth To that last nothing under earth. ^ This statement is erroneous. 32 Syllabification and Accent in the Paradise Lost. Again, while the unphilological reader might be so bold as to accept a trochee for an iamb in any one of the first four feet of a blank verse, would even he allow such a rhythm-destroying inversion in the final foot? Would he not rather, of the two evils, if Mr. Beeching will, prefer as the lesser such an accentuation as surface, for example, in the line Which of us who beholds the bright surface f The form of expression we have just used indicates that we concede that to the unphilological reader there might conceivably be a possibility of choice between surface and surface. Wholly impossible, however, is it that anyone should refuse, to instance but one out of numberless similar examples, such an accentuation as lovest in the fifth stanza of Shelley's Invocation, where the rime fixes the accentual form beyond choice. II II I love all that thou lovest, Spirit of Delight ! The fresh Earth in new leaves drest And the starry night, etc. Thus even the unphilological reader will probably concede that Milton, Shelley, Tennyson, Wordsworth furnish us with surface^ III lovest, among, before. The insufficiency of the views of metricians has been perhaps fully enough illustrated in the above review, and justifies, we believe, the following pages. Syllabification and Aocent in the Paradise Lost. 33 PART I. SYLLABIFICATION. Divisions of Part I : — Syllabification. A, within a word. B, between words. {a and /3) Synizesis corresponds to Synclisis ^ Syneresis " " | ?'f f. H ^l^^^o"' ^ lEcthlipsisJ (7) Syncope - .f Apocope and L Apneresis (of vowels). Dieresis is treated along with Synizesis and Syneresis, Hiatus with Elision. Definitions : — Synizesis is the falling together of two contiguous vocalic elements within a word into a diphthong : chari-ot > chariot. This is a very common phenomenon. Syneresis is the falling together of two contiguous vowels within a word into a monophthong. This is not a common phenomenon. Even where it seems to occur it is difficult to distinguish it with certainty from Synizesis. Perhaps we have examples of it in cre-ature >- creature, Cana-an > Canaan. Opposed to Synizesis and to Syneresis is Dieresis, which is the standing apart of two contiguous vowels within a word so as to form two syllables : fe-alty, ide-a, Ba-alim, etc. This phenomenon is common. Syncope is the cutting out of a vowel within the body of a word. Except in certain verbal endings it is very rare. 3 34 Syllabification and Accent in the Paradise Lost. Elision is the blending of two contiguous vocalic elements in separate words into a monosyllable. It occurs in three forms : — (1) In Synclisis both the vocalic elements remain and form a I XX I I X I I X X I I XI diphthong : many a man >• many a man, able as he > able as he. Synclisis is very common. (2 and 3) In Crasis and Ecthlipsis one of the two vowels is absorbed in the other. In the written word these phenomena are hardly to be distinguished from Synclisis. (2) In Crasis the second vowel is absorbed in the first. An XX / X / example of this is perhaps No ingrateful food > No 'ngrateful food, V407. (3) In Ecthlipsis the first vowel is absorbed in the second. An X X / X / example of this is perhaps To whom thus the portress > T' whom thus the portress, II 746. Apocope is the cutting off of final, Apheresis the cutting off of initial, sounds, vocalic or consonantal. These phenomena are very rare, the only examples being i(n) th{e) (twice), and {o)f thy. Opposed to Elision is Hiatus, which is the standing apart of two contiguous vowels in separate words so as to form two syl- / X / lables : from all the \ ends. This phenomenon is very common. The correspondence between the phenomena of syllabification as occurring within a word and between words is to be noted. Thus Apocope and Apheresis, when vocalic, correspond to Syncope ; Crasis and Ecthlipsis to Syneresis; and Synclisis to Synizesis. So Hiatus corresponds to Dieresis. The above terms are employed in accordance with ten Brink's usage in his Chaucerian grammar. A. Syllabification within a word. a. Synizesis and dieresis of vowels. General rule : Two vowels in contact stand in dieresis if one of them is under the primary accent ; otherwise they suffer synizesis. 1. a before or after tonic vowel. Dieresis is regular^ : La-ertes, ^In the list which follows, the examples of pre-tonic a precede the semicolon, those of post-tonic a follow. So also for e, i, etc., under 3, 5, etc. Syllabification and Accent in the Paradise Lost. 35 H-ereal ; fe-alty, ide-a, re-al, the-atre, unagi^ee-ahle, di-adem, 3fo-ah, voy-age, Ba-alim (cf. Balaam under 2), etc. Important among this class are proper names in -ce-a, -e-a, -ei-a : Judce-a, Uzze-an, Tarpei-an. Exceptional synizesis, or perhaps preferably syneresis (monoph- thongization), is constant in creature. Varying usage is found in three words : diamond, Messiah, trial. In the last two, ia stands regularly in dieresis, but we have Messiah in XII 244, and trial in I 366. In diamond, on the contrary, synizesis outweighs dieresis, and we have di-amond in III 506, Y 634, but diamond in lY 554, Y 759, YI 364. 2. a + atonic vowel. Synizesis (or syneresis) is regular : Balaam, Canaan, Canaanite, Isaac, Israel^ etc. Exceptional dieresis in Senna-ar, Yarying usage. In Abraham the h is generally weakened and the two a's form one syllable. In XII 152, however, they stand in dieresis. Michael and Raphael vary : Of thunder and the sword of Micha-el. II 294. Thy Maker's image, answered Michael then. XI 515. / / . ^ Micha-el and his angels prevalent. YI 411. / / Michael, this my behest have thou in charge. XI 99. 3. e before or after tonic vowel. Dieresis is regular : be-atitude, Be-ehebub, Ele-ale, Cle-ombrotus, cre-ate (and its similarly accented derivatives), pre-amble, re-alities ; se-er, ve-hement (though perhaps the h is strongly enough sounded to keep the vowels apart), nigh-est, qui-et, embow-eVd, low-er, pro-em, cru-el, etc. Synizesis: satiety y 111 216, j^ursuers. Yarying usage : diet but di-eted ; variety YII 542, but vari-ety YI 640 ; soci-ety in / / Among unequals what soci-ety. YIII 383, but society in For solitude sometimes is best society. IX 249 ; * Sometimes Isra-el in P. B. 36 Syllabification and Accent in the Paradise Lost. piety and pi-ety ; generally high-er and high-est, but also higher and highest; generally hi-erarch, but Of hierarchies, of orders, and degrees. 4. e + atonic vowel. Synizesis is regular : atheist, Boreas, all-bounteous, Bethlehem (with weakened h), delineate, gorgeous, Leucothea, meteor (cf. mete-orous in 7), spontaneous, etc. Exceptional dieresis : recre-ant, Briare-os, Capre-ce, Pane-as, Pelle-as, Be-ersaba.. Proper names, due to their un familiarity, would naturally tend to keep the fuller form. In the last word this tendency is aided by the initial position of the atonic syl- lable, — a position which has a peculiar strength. Variation. Generally ethereal, but ethere-al V 499. 5. i before or after tonic vowel. Dieresis is regular : Bri-areos, Gui-ana, i-ambic, Sy-ene, Thy-estean, tri-umph, Ethi-opian (cf. Ethiop in 6), expi-ations, medi-atioii (cf. the varying forms of mediator below), Ophi-uchus (cf. Ophiusa below), Sogdi-ana, Susi-ana; Mahana-im, Mosa-ic, assay-ing, survey-ing, see-ing, enjoy-ing, vow- ing, bloiv-ing, fiu-id, vacu-ity, ly-ing, etc. ./ \ . .1 . \ . ./ . Synizesis : Serraliona, humiliation, propitiation (the recession of the secondary accent in the last two words is to be noted; cf. Part II, 3 c, and The accent of various words), Ophiusa, Damiata. oi is always treated as a diphthong : choice, foil, join, etc. Disobeying, obeying, displaying, viewing occur only final, and in that position seven times altogether. The great preponderance of the masculine ending may point to the probability of these words suffering Synizesis, but positive decision is impossible. Variation. Generally say-ing, but twice saying; almost uni- versally ru-in (and so its derivatives), but ruin I 91, and ruinous II 921 ; generally _^2/-m^, but flying twice; regularly dy-ing, but dying once; cZo-m^^ and do-ings generally, doing II 162; the Syllabification and Accent in the Paradise Lost. 37 noun being usually occurs in the final position where its syllabifi- cation is of course ambiguous ; when not final it behaves contrary to the rule, being only once be-ing II 865, to fiour times being ; deity and deities however illustrate the rule, being generally each three syllables; we have deities, however, in VI 157, and deity IX 885, XI 149; the form is ambiguous in The Deity and divine commands obeyed. V 806. Medi-ator X 60, mediator XII 240. 6. i + atonic vowel. Synizesis is regular. There is great wealth of illustration: Abdiel, audience, bestial, chariot, costliest, demoniac, JEthiop (cf. Ethi-opian under 5), Oeryon, idiots, inglorious, myriads, obsequious, Quintius, saviour, speedier, speediest (the con- stant way of treating comparatives and superlatives of adjectives in -y), superior, symphonious, tragedians, valiant, zodiac, etc. Syni- zesis affects some words the fuller forms of which would seem to be more euphonious : costliest, illustrious, stateliest, etc. The syllable suffering synizesis is often followed by another syllable : Adiabene, immediately^ vitiated, mysteriously, variously, gloriously, commodiously. Exceptional dieresis : expi-ate, soci-ably, the proper names Dari-en, Ithuri-el. In insati-able, vari-able, soci-able, dieresis is due to the fact that -a- is more suitable for an arsis than is -ble. Varying usage. Gladli-er VI 731 is contrary to every other case of -ier and -iest in P. i. ; thus VIII 47 gladlier. Gabriel and Uriel, though generally regular, show each one excep- tion,— Ga6ri-eZ initially IV 865, and Uri-el III 648. In I 112 ^As to the scansion of immediately in XI 477 and XII 87, we hare the choice between /. / / . / / . / . immediately a place, immediately inordinate, and / -/ / ■ /-/ / / immedi-ately a place, immedi-ately inordinate ; but the only possible scansion of VII 285, immediately the mountain, would show the former of the above pairs to be right. 38 Syllabification and Accent in the Paradise Lost. we find the adjective suppliant, in X 917 the noun suppli-ant. The difference is hardly due to the difference of grammatical category. It is more probable that the final position of the latter word tends to cause its fuller form. I have not collected tnany data to verify this hypothesis — i. e., that final position in the line tends to preserve the fuller form of the word, but examples like the following point to such a tendency, Final effemi-nate Non-final : effeminate gene-rate temperate exasper-ate regenerate dolo-rous dolorous. 7. before or after tonic vowel. Dieresis is universal : Cho- I / , / / aspes; cha-os, Nebai-oth, Galile-o, Le-o, mete-orous (cf. meteor in 4),, Chi-os, Hermi-one, li-on, vi-olated, vi-olet, etc. 8. + atonic vowel. Synizesis : arrowy, bellowing, echoing^ followers, Meroe, Samoed, Siloa. Exceptional dieresis in the the-ologians and pi-oneers is due to the strength of initial syllables. Aro~er preserves the fuller form perhaps because it is a proper name. / 9. u before or after tonic voweL Dieresis : fru-ition; Ima-us^ I I Asmode-us, Lycce-us, tri-umph, pi-ous. I I Exceptional synizesis : puissance, puissant. 10. M + atonic vowel. Synizesis is usual : assiduous, casual^ extenuate, fiuotuates, impetuous, Joshua, mellifluous, presumptuously, profluent, sensualist, transpicuous, visual, voluptuous, virtuousest, etc ^ Exceptional dieresis: vacu-ous Yll 169. Dieresis in situ-ate VI 641 may be due to the exceptional syncopation of the past- participial ending in a verb which ends in a dental (cf. Part I, A7). Any further shortening would be undesirable. Dieresis in sanctu- ary, insinu-ating, situ-ation, is due to the fact that of the two final Syllabification and Accent in the Paradise Lost. 39 syllables bearing secondary accent the next to last is the more suitable for use as arsis. Variations. In influence dieresis occurs seven times, synizesis twice. Possible the u is kept so often in a separate syllable by the double consonant preceding it, for when it suffers synizesis the second syllable starts off with the heavy consonant combina- tion /w. Superfluous, on the other hand, inclines strangely enough to the form superfluous which occurs three times, while we find superflu-ous but once, IV 832. The different behaviour of the two words is perhaps due to the fact that in the latter word the /can combine with the previous syllable much more easily than in the former word, thus in-flu-ence but superf-luous. Five times spiritual forms but two syllables, and is therefore scanned sph^itual. Four times it forms three syllables. It is impossible to say in the latter case whether it shall be scanned spiritu-al or spir-itual, spirit being as customary in Milton's usage as in -ual. The trisyllabic occurrences are (initial) IV 585, V 406, XII 521, (medial) V 573. ■ 11. In other tonic combinations synizesis is usual. This seems contrary to the general rule, -ower and -our are triphthongal in nature, due to a vocalic glide preceding the r ; but the triphthong never breaks up into two syllables save occasionally in the case o{ flower and power. Some of these words are bower, coivering, dower, hour, lour, powerful, shower, tower. Flower and power, too, are always monosyllabic except in one instance each : With flow-ers, garlands, and sweet-smelling herbs. IV 709. At which command the pow-ers militant. VI 61. Flower is very frequently final, in which position, due to the great preponderance of masculine endings, a word is best considered monosyllabic. Similarly we have diphthongs and triphthongs, not dieresis, in faey-y, fiery, prayers, sewers. Friars, which occurs but once and then finally, is from its position probably a monosyllable. loward and towards. This word in its two forms is almost always monosyllabic. In five cases, however, it must be scanned 40 SyllabiflGation and Accent in the Paradise Lost. as two syllables. Two of the lines containing its dissyllabic form, / / Straight to-ward heaven my wond'ring eyes I turned. Y III 257, Safe to-wards Canaan from the shore advance. XII 215, are, if scanned as suggested with inverted initial foot, ambiguous as to the accentuation of to-ward{s). But YI 648, When coming to-wards them so dread they saw. gives the word initial accent, as do also lines IX 495 and XII 296. yS. The development of syllabic nasals and liquids and their syni- zesis with contiguous syllables. An atonic syllable containing a nasal or a liquid tends to lose its vowel, letting the nasal or liquid become syllabic. The syllabic nasal or liquid may then suffer synizesis with the contiguous syllable. The cases are three : — I. The syllabic nasal or liquid undergoes synizesis with an atonic vowel : bordering > bord-r-ing >> bordering. II. The syllabic nasal or liquid undergoes synizesis with the tonic syllable. Of this there are two cases : — a. The syllabic nasal or liquid has developed from vowel + nasal or liquid. Then the synizesis is with, not the tonic vowel but, the tonic syllable, which must end in a very sonorous sound. Thus fallen y^falln '^falln. b. The syllabic nasal or liquid has developed from nasal or liquid -\- vowel. The synizesis is then with the tonic vowel. Thus flourishing ^ flour shing '^flourishing. Taking up these cases in turn we shall find that I, the synizesis of the syllabic nasal or liquid with the atonic vowel, is common ; while II a and 6, the (development and) synizesis of the syllabic nasal or liquid with the tonic syllable, is unusual. The similarity of these usages to the general rule for the synizesis and dieresis of vowels is evident (cf. Part I, Aa). I. I. Development and synizesis of I with an atonic syl- lable is fairly frequent : javelin, devilish, articulate, credulous (in IX 644 ; in P. R. II 166 the fuller form occurs), oraculous, Syllabification and Accent in the Paradise Lost. 41 pendulous, popular. In these latter words the u must have been yet a monophthong, — not a diphthong as at present. A set of words belonging here is formed by derivatives from words in -ble : fable-\-ing =:fabl -\-ing ";> fabl-ing (three syllables) y- fabling (two syllables). This is true synizesis of I, and when we add this category to the words first mentioned, the forms showing synizesis preponderate over the fuller forms now to be cited. The fuller forms are frequent : Sofa-la, vassa-lage, marvel-ling, fami-ly, deso-late, fraudu-lent, regu-lar, Baby-Ion, etc. Varying usage is found in perilous which is once dissyllabic (I 276) and once trisyllabic (II 420). We have orbicular in X 381, but orbicu-lar in III 718. Populous shows synizesis of I three times (I 351, 770, IX 445) against its fuller form twice (II 903, VII 146). With the last word we should compare the scansion popular, which is used in all four occurrences of this word. m. Development and consequent synizesis of syllabic m does not occur. Thus : tourna-ment, ele-ment, impedi-ment, astrono-mer, Solo-mon, etc. This constancy of usage enables us to decide between the two possible scansions of IX 494 and IX 905, which we must read / / . ^ So spake the ene-my of mankind, enclosed. Of ene-my hath beguiled thee, yet unknown. instead of scanning enemy, with hiatus between the y and the next word. n. Development and consequent synizesis of syllabic n is very common : covenant, bituminous, business, reasoning, reasonant, unseasonable. Further, the very large class of present participles of verbs in -e7i ; betokening, chastening, hardening, etc. But the fuller form of similar words is very common also : Gompa-ny, arse-nal, proge-ny, emi-nence, imagi-ning, coro-net, fortu-nate, etc. The words that vary are countenance, which is generally shortened, but occurs in its full form in III 730 ; original, which 42 Syllabification and Accent in the Paradise Lost. is shortened four times, but occurs thrice in its full form, II 375, IX 150, 1004, {originals in its single occurrence is un contracted) ; lumi-nous II 420, luminous VIII 140. We have luminary and luminaries each once jfinal, where it is impossible to decide certainly as to the scansion. The preponderance of the masculine ending, however, is in favor of the shortened forms. This is supported by YII 385, / . // With their bright luminaries that set and rose. Usage seems to be about equally divided between the full and the shortened forms in the words under this section. r. Development and consequent synizesis of syllabic r is so fre- quent as to be considered, perhaps, the rule : barbarous, nectarous (cf. necta-rine below), answering, artillery (cf. mise-ry below), collateral, considerate (cf. gene-rate, ite-rate below), delivey^ance (cf. igno-rance below), difference, feverous, liveries (cf. sente-ries below), mastery, utterance, arbo7-ous, corporal, labourers, savoury, adventurous, torturer. The fuller form, however, frequently occurs, as : adulte-ry, circumfe-rence, exaspe-rate, gene-rate, ite-rate, mise-ry, mode-rate, sente-ries, igno-rance, etc. The fuller form may perhaps be expected when the syllable following the r is not a common formative element : cata-ract, necta-rine, favou-rite, cormo-rant, Thamy-ris. Moreover the fuller form is to be expected when the shortened form would throw a less suitable syllable into the arsis : mise-rable, insepa-rably. Varying usage in the same word is frequent. In such words, however, the shortened form is the commoner one. Words of which the reverse is true are marked (*) in the following list of words showing varying syllabification : — Conqueror twice : conquer-ror I 323 : conquer-rors XI 695 dangerous generally : dange-rous II 107 desperate II 107 : despe-i^ate III 85 Syllahificaiion and Accent in the Paradise Lost. 43 different generally : diffe-rent I 636, IX 883 emperor twice : empe-ror II 510 / \ "^innumerable V 585 : innume-rable generally "^liberal IX 996 : libe-ral twice mineral I 235 : mine-ral IV 517 numerous generally : nume-rous XI 130 pondering generally \ ponde-ring VI 127 prosperous II 259 : prospe-rous XI 364^ reverence generally : reve-rence VIII 599 ^reverent in P. R. once : in P. L. 7'eve-rent three times temperance generally : tempe-rance XII 583 amorous generally : amo-rous IV 31 1, VIII 477, IX 1035 dolorous twice : dolo-rous II 619 odorous generally : odo-rous V 482 natural generally : natu-ral X 765 sulphurous twice : sulphu-rous VI 512. In the above list the preference for the fuller form of liberal and reverent is exceptional or accidental. In the case of innumerable, the fuller form is preferred, as the shorter form throws into arsis-position the ending -ble which is less suitable for this purpose than is the -a~. II a. The development of a syllabic liquid or nasal from vowel 4- liquid or nasal, and its consequent synizesis with the tonic syllable. This process, or at least the synizesis, is exceptional, not regular. ^ It is interesting to compare these two occurrences. The first is in the body of the line: prosperous of adverse. The second is initial: prospe-rous or adverse. Here we need a trochee plus an iamb, and therefore prosperous, though in just the same collocation as in the first passage, must be scanned with one more syllable. 44 Syllabification and Accent in the Paradise Lost. I. Generally dev-il and ev-il, but evil occasionally, and devil X 878. "" ^ m. There is no word in which the treatment of m according to II a. would be possible. n. -len. In this combination the fuller form occurs only in sullen. Otherwise : befallen, fallen, stolen, swoln. -ren, -ron. -ren occurs only once, i. e. in barren, which retains the full form, -ron occurs only in iron, which in like manner generally counts as two syllables. In II 878, III 594, XI 565, however, iron forms but one syllable. In the first and last of these lines. Of massy iron or solid roch with ease. Labouring, two massy clods of iron and brass. it would be possible to let the n of irn syllabify with the initial vowel of the following word. But this kind of slurring we shall later find to be quite exceptional, save in the case of I in the ending -ble. Moreover in III 594, With radiant light, as glowing iron with fire. synclisis of the n with the next word is impossible, and here at least we must have irn > irn. o ^ ^ -ven. This combination in certain words retains its full form : clov-en, rav-en, riv-en, generally the adjective ev-en, and various compounds of wov-en. It is interesting to note concerning this last word that when Milton uses the past participle where a mono- syllable is metrically required, he does not change wov-n > wovn, but employs the n-less form wove ; as in III 352, Their crowns in-wove with amarant and gold, -ven inclines, however, to the form -vn: graven, given, seven (and its compounds), the noun even (wherever we can be sure of its scansion, i. e. medially ; it is frequently final), the adverb even ; the adjective even is, however, generally dissyllabic, though it is Syllabification and Accent in the Paradise Lost. 45 shorteDed in VI 544 ; generally driven, though driv-en certainly at least twice, VI 738 and VII 185. When final its scansion is of course ambiguous ; as it is also in III 677, which naay be read in two ways, / / Who justly hath driv-en out his rebel foes. ; / — / . / Who justly hath driven out his rebel foes. Heaven, simple and compounded, is regularly monosyllabic. Not infrequently, however, it is treated as two syllables. There is always a question how heavenly -\- vowel shall be scanned. Thus we have the two possibilities for I 138, As far as gods and heavenly f essences. As far as gods and hea-venly essences. The preponderance of the form heaven and the not great infre- quence of such an hiatus as -ly t essences, would point to the first scansion as preferable. The other lines in which the same question arises are VIII 453 and X 641. In XI 871 we have both synizesis and synclisis, / . / . At present, heavenly instruction, I revive. -zen, -zon. Here the full form is customary : braz-en, froz-en, imblaz-onry, seas-on, treas-on. Among the four words that show varying syllabification, prison and reason are regularly dissyllabic, though we find prison in I 71, VI 660; and reason in I 248, VIII 591, IX 559. Chosen is twice dissyllabic, I 8, III 183, and twice monosyllabic, I 318, IV 691. Risen, which occurs rather frequently, goes contrary to the regular usage, being always monosyllabic in P. L., though once in P. R. it appears as a dissyllable, II 127. r. The full form — with a single exception — is constant : cov-er, hith-er, pover-ty, riv-ers, and generally ev-er. The one exception is ever III 244. At first sight we might think that we have further examples of forms similarly shortened in the monosyllabic use of pillar 46 SyUabification and Accent in the Paradise Lost. II 302, XII 202, 203 ; whether X 57, XI 296 ; savour X 269 ; river III 358. But we find that in each of these cases the follow- ing word begins with a vowel. We have accordingly synclisis rather than synizesis. Cf. Part I, B. II b. The development of a syllabic r + vowel, and its con- sequent synizesis with the tonic vowel, is most exceptional. It takes place in flourishing in its only occurrence, P. R. Ill 80. Authority occurs a number of times, but the shortened form authority we find only once, P. R. II 5. Both of these reductions, it is to be remarked, are found in P. R. 8pirit(s), however, it was Milton's custom from the beginning of P. L. to use as a monosyllable. The fuller form occurs about 30 times out of 100, or 1 in 3. / .\ / . \ / \ In spir-ited, spir-itless, spir-itous, the secondary accent on the derivative syllable is used as arsis and shortening does not take place. In other words of the form o^ flourishing the shortened form is never used : car'-avan, cher-ishing, mer-its, glor-ify, etc. ; idol-atries, civil-ity, pol-icy, vol-uble, etc. ; rem-idy, calam-ity, Trem-isen, etc. ; min-ister, un-ison, in-nocence, etc. 7. Syncopation. Syncopation (except in certain verbal endings) is most excep- tional. All the cases are as follows : capHal II 924, XI 343, XII 383 {eap-i-tal occurs I 756) ; CapHoline IX 508. No other instances occur, and words which would seem to lend themselves excellently to syncopation are not so treated : cap-i-tol (P. R. IV 4:1)Jelic-i-ty, dif-fi-cult, etc. The second person singular ending -est is regularly syncopated both in the present and in the preterite. Regularly syncopated is the preterite and past-participial ending -ec?, save in verbs whose roots end in d or t. There are, however, a goodly number of exceptions to the syncopation of -ed, and we have not infrequently wing-ed, blessed, curs-ed, arm-ed, Syllabijicfition and Accent in the Paradise Lost. 47 forJc-ed, etc. When the adverbial ending -ly is added to a past participle in -ed, the fuller form seems to have been preferred by Milton. Is this preference due to a possible acoustic uncertainty as to form that might have resulted from syncopation ? Be that as it may, even to the present day, when -ed is elsewhere uniformly syncopated (except after t and d), the fuller form is preserved before -ly. Thus Milton writes : Deserv-edly thou griev'st, composed of lies. P. R. I 407, and similarly in P. R. lY 133. This confirms the scansion of II 914, / / Confus-edly, and which thou must ever fight. which otherwise might have been read Confusedly, and, etc. We have on the other hand a few cases in which syncopation of the past-participial ending in -ed takes place after a stem ending in t, with consequent simplification of the double dental : uplift I 193 (cf. uplifted I 347, II 7), unsusped IX 771, convict X 83; and these words in -ate: elevate II 558, alienate V 877, infuriate Yl 486, situate VI 641, frustrate XI 16. B. Syllabification between Words. Following ten Brink in his Chaucerian grammar we use the term Elision as comprising Synclisis {many a man, both vowels retained as a diphthong), Crasis {No ingrateful food Y 407, loss of the second vowel), and Ecthlipsis {To 'whom thus the portress II 746, loss of the first vowel). It is generally unnecessary to distinguish these subdivisions of Elision. It will be seen that Elision, whether Synclisis, Crasis, or Ecthlipsis, always means the monosyllabification of two con- tiguous vocalic elements. -e of the, final ^ of -hie, and -y are very frequently elided before (or suffer synclisis with) a following initial vowel. If, however. 48 Syllabification and Accent in the Paradise Lost. Miltou's preference seems here to elide, the exigencies of the metre often decide between hiatus and elision. Accordingly we may get such a line as XI 345, Fi^om all the f ends of the earth to celebrate. Other less usual elisions are : 1. {th)e, -y, -{b)l, elided before h : the: I 520, 577, H 759, III 560, IV 610, VI 79, VII 322, 371, VIII 45, IX 52, X 204, 684, XI 372, 583. -y: III 677,V 366, IX 905, X %Q, 722, 874, 906, XI 333, XII 340. -ble: IV 250. 2. the elided before w: VIII 613 [the way). 3. Other vocalic endings elided before vowels : to: I 323, IV 67, V 271, 360, 383, 576, 614 {into), Qll, 725, VI 316, 636, VII 350, IX 648, 992, 1140, 1185, X 110, 334, 467, 486, XI 170, 236, 596, XII 437, 499, 515. Other o's ; I 470, 584 {Morocco), II 450, III 108, V 407 (no), 628, IX 296 {though), 1082, X 203, XII 611. -ow: I 558,11 518,111 120, V 575, X 717, 1092, 1104, XI 757, XII 613. be, thee, me, he, we: I 245, II 703, III 3, IV 758, 839, V 107, 553, 563, VIIE 316, IX 121, 152, 546, 570, 746, X 75, 149, 762, 766, 769, 795, XI 689. -tue (generally in virtue; twice in continue): II 314 {continue), III 586, IV 371 [continue), 848, VI 703, VII 236, IX 110, X 884. thou: X 121, 758. journey: V 559. , they: X 567. I: VIII 611. my: X 468. 4. Other vocalic endings than {th)e, 4, -y, elided before h : to: I 524, 525, 749, II 98, 746, 968, V 447, 467, VI 814, 909, IX 1153, X 12, 594, XI 453. thou: X 198, XI 347. 6e; VIII 649. virtue : X 372. Syllabification and Accent in the Paradise Lost. 49 (5) n and r elided before vowels, n; IX 764 {eaten), IX 1025 {forbidden), X 116 {garden), X 983 {begotten), r: II 302 {pillar). III 358 (nVer), YIII 498 {mother), X 269 (savowr), X 57 and XI 296 {whether), XII 202 and 203 (pi7/ar). There follow a few lines in which care must be used in making the elisions : / / Created hugest that swim the ocean stream. I 202. / / GamboWd before them ; the unwieldy elephant. IV 345. Firifwe in her shape how lovely ; saw and pin' d. TV 848. No in grateful food ; and food alike those pure. V 407. Of rainbows and starry eyes. The waters thus. VII 446. / / Leads up to heaven, is both the way and the guide. VIII 613. Because thou hast hearkened to the voice f thy wife. X 198, (showing also apheresis in Y.) / / Against God only, I against God and thee. X 931. A pillar of state ; deep on his front engraven. II 302. The phenomena so far treated under B correspond to a and /5 under A. Corresponding to A7, Syncopation, we have double apocope (the cutting off of a final sound) in the change of in the to i' th', I 224, XI 432 ; and apheresis (the cutting off of an initial sound) of of thy to 'f thy, X 198. These are the only cases of apocope and apheresis in the poem. The following erroneous scansions containing the italicized tri- syllabic feet cannot be corrected by elision, apocope, or apheresis. , , / \ With impetu-ous recoil and jarring sound. II 880. , / And Ti7'esi-as and Phineus, prophets old. Ill 36. / \ . And Cbrpore-al to incorporeal turn. V 413. \ . / . And propiti-atwn ; all his works on me. XI 34. \ / By humili-ation and strong sufferance. P.B. I 160. / / On Lemnos, the Age-SLn ile ; thus they relate. I 746. ' 50 Syllabification and Accent in the Paradise Lost. Before thy feKoi^s, am6iYi-ous to win. VI 160. \ / Therefore thy Awmifl-atioa shall exalt. Til 313. Of sorrow unfeign'd, and AwmtYi-ation meek. X 1092, 1104. Proper aecentuatioa (and a consequent simple synizesis) will in each case correct the scansion. These lines will be therefore treated in Part II. Concluding remark to Part I. From the above data I think we may be sure that Milton's epic line is written upon a strictly decasyllabic basis. If he meant the vowels, liquids, and nasals, which we have treated above, to form independent and separate syllables, why did he not at times intro- duce hypermetric syllables so hedged about by consonants as to prevent that amalgamation of neighboring vocalic sounds into one syllable which we have advocated ? But this he did in all in only the seven cases pointed out above, — three times in capital, once in Capitoline, twice in in the, and once in of thy. With these excep- tions, the so-called extra-metric syllables are always in contact (generally vocalic) with a syllable with which they can combine by the simplest phonetic processes — either elision (generally in the form of synclisis) or synizesis — to form a (diphthongal) monosyllable. Syllabification and Accent in the Paradise Lost. 51 PART II. NUMBER AND POSITION OF ACCENTS. The ideas which underlie my treatment of the number and position of accents are not my own but are due to Prof. Jas. W. Bright's insight and originality. A suggestive exposition of some of the more important points of his metrical theory is made by Dr. Bright in two articles, the first being " Proper Names in Old English Verse " in the Publications of the Modern Language Association of AmeriGa,Yol. XIV, pp. 347 fF., the second "Con- cerning Grammatical Ictus in English Verse " in An English Miscellany Presented to Dr. Furnivall in Honour of his Seventy-fifth Birthday, Clarendon Press, Oxford, 1901, pp. 23 ff. The various theories as to the proper manner of reading poetry may be called the sense-doctrine, the ictus-doctrine, and the rhythm-doctrine. The sense-doctrine teaches that the only way to preserve the sense of a line of poetry is to read it exactly as if it were prose. At times this method of scansion gives good rhythm as in the line / / / II Of that forbidden tree whose mortal taste. but frequently the rhythm is ruthlessly violated : / / . / , /. Of man^s first disobedience and the fruit. This theory so ignores the music of the line that it may be dis- missed from further consideration. The ictus-doctrine demands that in reading verse the words standing under the ictus shall always be strongly stressed, and that no other words may be strongly stressed. This method of scansion has the virtue of always giving a rhythmic reading, which may at times be consonant with sense, as in / / / / / Of that forbidden tree whose mortal taste. 52 Syllabification and Accent in the Paradise Lost but is quite as often at odds with it. Thus in the line Nor served it to relax their sensed files, to, though an arsis, deserves beyond question only a weak stress ; while in Of Thammuz yearly wounded: the love-tale, love, though only a thesis, must yet for the sense be stressed. The ictus-doctrine can suggest no reading for these lines. Thus as the sense-doctrine disregards rhythm for sense, so the ictus-doctrine disregards sense for rhythm, and like the first must be dismissed. It is plain that we must have regard to both sense and ictus. Let this more comprehensive method of scansion be known as the rhythm-doctrine, or (to anticipate our argument by the name), as it also might be called, the pitch-doctrine. In exposition of this rhythm- or pitch-doctrine, we must first call attention to one of the most essential characteristics of poetic enunciation, — monotony or uniformity in the utterance of all the syllables. Professor Lloyd ^ recognizes four several types of oral English : the vulgar, the careless, the careful, and the formal types. These types are described as differing chiefly in the matter of syllabic stress, the more elevated kinds of utterance containing few or no syllables which are stressless. ^It is obvious enough,^ says Professor Bright, ' that in formal utterance the language has qualities {which may be described as musical) which are available for artistic use, and that these qualities are bound up with the careful observance of not only the principal but also the subordinate stresses of the It is possible to state more exactly what accents, obscured or neglected in ignobler utterance, are conserved in more exalted diction, and especially in that of poetry? Such a statement is possible. From the present day back to the earlier Anglo-Saxon poetry, English poets have always made use of the right to place the ictus not only upon the second member of substantive com- pounds, but also upon such derivative syllables as -lie {-ly), -nesSy -ig (-2/), -er, -en, -el, -or, -est, -ing, etc. Cf. the marked ictus in '^Northern English: Phonetics, Orammar, Texts. London: D. Nutt, 1899, p. 30. Syllabification and Accent in the Paradise Lost. 53 / But coloured leaves of latter rose-blossom, Stems of soft grass, some withered red and some / . , / Fair and flesh-blooded ; and spoil splendider . / / ■ Of marigold and great spent sunflower. Swinburne, The Two Dreams. Here the ictus is in conflict with the natural primary stress of the word, and the language in responding to the exigencies of verse, yields a new class of stresses, — new, that is, from the point of view from which prose-stresses are usually observed. In prose there is a kindred accentual phenomenon which will throw light upon this phenomenon of verse. When stressed syl- lables are to be contrasted or emphasized in prose, their normal accent (^stress + pitch) is simply increased : not good, but bad. What is done when unstressed syllables are to be contrasted or emphasized ? A reading of these sentences will suggest the answer : / . / , / It has not been repaired but impaired. I'm going towards town / not from town. The same phenomenon is observable in words which have not normally initial accent, when used as vocatives / / / / and exclamations : Elizabeth, Jerusalem, Precisely, i^emarkable. We see how easily syllables may be made prominent without disturbing the fixed word-accent. Further, the accent of these emphatic prose-syllables bears a strong resemblance to the marked verse-stresses in the above-quoted lines from Swinburne. Now it can be easily seen that the new prose-stress is not a word-stress equal to the regular word-stress in expiatory force ; nor a reduced form of the expiatory word-stress, which would be nothing more than a secondary accent in prose. Rather, this new accent is a stress with a rising inflexion, a pitch-accent. The same is presum- ably true of the nearly related new verse-stress under consideration. We may say, therefore : When the verse-accent or ictus stands in conflict, it is attended by the pitch-accent. This conclusion is strengthened by an historical argument hinted at above. We saw that the poets have always felt at liberty to use for arses syllables having a secondary stress. But a marked characteristic of secondary stress is the important part played in 54 Syllabifieation and Accent in the Paradise Lost. it by the element of pitch. We are thus led in another way to the same result, i. e. that the ictus in conflict is marked by pitch. To bring out somewhat more clearly the characteristics and advantages of the rhythm- or pitch-doctrine of scansion, let us contrast it with the doctrines of sense and ictus above described. The advocates of the sense-doctrine claim that it alone enables the meaning of the verse to be properly preserved. In reality, however, the rhythm-doctrine is quite as efficient in the preserva- tion of the meaning of the line, for the characterization of the ictus by pitch enables us to retain the stress upon important words standing in the thesis. On the other hand, unimportant syllables standing in the arsis can be marked by the pitch-ictus without giving them an undue (stress-) prominence which would obscure the sense. The ictus-doctrine teaches that the rhythm can be properly preserved only when there is coincidence of stress and ictus. Accordingly when a thesis is stressed or an arsis unstressed, this doctrine offers no possible scansion of the line, but condemns it as unrhythmical. The rhythm- or pitch-doctrine, finding no invariable connection between stress and ictus, is delivered from this difficulty. For remembering that the ictus (and thus the rhythm) may depend much upon pitch, we see that a light syllable if distinguished by pitch is sufficient to mark the rhythmic wave, while a stressed syllable if relatively reduced in pitch can easily stand in the thesis. It is evident therefore that the rhythm- or pitch-doctrine com- bines the advantages of the two theories and preserves at once the sense and the rhythm of the words. ^ Symbols and Terms. (a). Ordinarily a syllable having primary stress, ^, stands in the arsis with the stress used as ictus. (yS). So a syllable having secondary stress, '**^ , ordinarily stands in the thesis, the secondary stress being disregarded metrically. Thus as examples of (a) and (/3) : ^ For another statement of Professor Bright's theory cf. Dr. Julian Huguenin,^ Secondary Stress in Anglo-Saxon (a Johns Hopkins University diss.), Baltimore : John Murphy Co., 1901, p. 6, note 1. SyUabification and Accent in the Paradise Lost. 55 / / \ / / \ / Of that forbidden tree whose mortal taste. (7). A syllable under secondary stress may, however, by an increase of the element of pitch, become available for use as an arsis. The accent it then bears is distinguished by heightened pitch without increase of stress, thus differing from both ordinary primary and ordinary secondary stress. This accent shall be represented by '', and may be conveniently termed the pitch-ictus. (S). Further, the pitch-ictus, //, may also rest upon an unstressed syllable. (e). A stressed syllable may stand in the thesis, if its pitch, relatively to the contiguous arses, is low. Its accent shall then be represented by ^, and may be conveniently termed the reduced primary stress. Thus as examples of (7) (S) (e) : // . A // Which of us who beholds the bright surface. (f). Unstressed syllables are represented by '^. [t]). a general symbol for the arsis is -, and for the thesis is <-'. (6). When two stressed syllables are used as an iamb by the reduction of the pitch on the first syllable (cf. e), the resultant foot may be called, with regard to the heavy stress still remaining on the thesis, a heavy iamb, ^ ^ . Cf. A / I mean of taste, sight, smell, etc. (c). When two lightly stressed syllables are used as an iamb by the placing of the pitch-ictus on the second, the resultant foot may be called, with regard to the light stress still remaining on the arsis, a light iamb, x //. Cf. X // Nor served it to relax, etc. (k). When the combination of a heavily stressed + a lightly stressed syllable is used as an iamb by the reduction of the pitch of the first syllable and the addition of the pitch-ictus to the second syllable, the resultant foot may be called, with regard to the combination of heavy stress still remaining on the thesis and light stress still remaining on the arsis — a stress-combination which constantly tends to turn the foot into a trochee — an unslnble iamb, A //. Cf. A // Which of us who beholds the bright surface. 66 SyllabiJiGation and Accent in the Paradise Lost. {\). This conflict between the stress-relation which favors one form of the foot (in k a trochee) and the pitch-relation which demands the other form (in k an iamb) is termed, simply, conflict. Conflict has two manifestations. One is as in k. Cf. further 6-13 below. The other manifestation of conflict is jitst the reverse of the above, the stress-relation favoring an ascending rhythm, while the pitch-relation demands a descending rhythm. Cf. // A Yet en|vied ; a\gainst me | is all their rage. Cf. further 3-4 below. It is to be noted that in each case of conflict, whether ^ 1 1 or 1 1 ^, it is the j9i7cA.-relation which determines the form of the rhythm. 1-4. Light and Heavy Iambs (x //, a /). 1. The light iamb is common in Miltonic and all blank verse : — X // Nor served it to relax their serried files. VI 599. Though the type-line may be said to consist of five normal iambs (x /, or '^ /), the commonest form of the line has one light iamb. Less commonly two such feet occur : X // X // ^ Transfix us to the bottom of this gulf. I 329. I cannot cite an example of two successive light iambs in Milton. If they occur they are extremely rare. 2. The heavy iamb is common in Miltonic and all blank verse : _ A / By conquering this new world, compel me now. IV 391. The heavy iamb is never used throughout the whole line ; it may, however, occur a number of times, and consecutively : A / A /_ I mean of taste, sight, smell, herbs, fruits, and flowers. VII 527. A / A / A / His full wrath whose thoufeeVst as yet least part. X 951. 3. Light iamb -\- heavy iamb. — Stress Conflict. In 1 and 2 we have had examples of light and heavy iambs in what might be called a normal environment. In such conditions the pitch-accent has only to indicate which of two syllables having equal stress shall stand at the crest of the rhythmic wave. Suppose, Syllabificaiion and Accent in the Paradise Lost. 57 however, that the light iamb is followed by a heavy iamb. Here a lightly stressed syllable shall be used as arsis though directly followed by a more heavily stressed syllable. This is of course easily accomplished by the use of pitch-accent on the first arsis and a reduced primary stress on the following thesis, but, as the stress-relation remains unchanged, there results one of the two manifestations of conflict (cf. Symbols and Terms \). Let us first consider this phenomenon when X // A / a. The conflict occurs in two words. Examples : hut to bring forth X // A / . _ ,X // A /. I 217, when with fierce winds I 305, inclination or sad choice II X // A / X // A / 524, wounded the love-tale 1 452, me the dear pledge II 818, etc., etc. I find over 175 instances of this kind of combination of light + heavy iamb (with consequent conflict) in the poem ; but in some cases the method of scansion might vary. It is to be remarked that the combination of light + heavy iamb (when the conflict is between different words) is particularly apt to stand at the end of a line, as in I 30, 40, 82, 97, 122, 217, 238, 268, 256, 379, 452, 453, etc. It may happen that \ b. The conflict occurs in the syllables of one word, as : a goddess // A / ' \ // A / \ // A among gods IX 547, sentence beyond dust X 805, reluctance against God X 1045, etc. This phenomenon is of infrequent occurrence. There follow all the instances. Only three times (marked /.) does the word form the arsis of the fourth and the thesis of the fifth foot. Against IV 71, YI 813, 906, X 931, 1045; arr!^ong III 283, VII 623, IX 547 ; becomes XII 409 ; before YIII 464 ; between VII 473; beyond II 7, V 159, X 463, 805; confused II 615 ; //A //A // A // A except II 1032 (/.); obscene I 406 ; obscure II 132 (/.) ; supreme I 735 (/.), II 210; unless VIII 186.^ ^ An analogous accentual development is to be observed in the following words : //,../ ,^,,,„ // / // / // / // / ambitious V 1 160, corporeal V 413, impetuous II 880, quintessence III 716, propitiation XI 34, humiliation III 313, X 1092, X 1104, P. B. 1 160. In ttese words we 58 Syllabification and Accent in the Paradise Lost. It is to be observed that in each case the syllable receiving the pitch-accent is a prefix of prepositional or adverbial nature, and may therefore properly bear, due to its independent idea-content, a marked stress. To this statement supreme forms an exception. Here the prepositional adverb is itself the root of the word and is hence peculiarly prone to receive the accent. c. The combination of light + heavy iamb when non-final may be easily changed into triple rhythm. Thus in part of II 18 X // A / X / and the \ fix^d laws j of heav'n the conflict of the stress on jix'd with the pitch on the will find satisfaction in the form and the jix'd \ laws of heaven. This is more stable than the first form, in that it does not show conflict ; but in the second form there is still a reduced A , A primary stress (on laws), as there was in the first form (on jix'd). The second form is therefore not so perfectly regular as to induce the reduction of the first form thereto. When the combination of light -j- heavy arsis is final (as it generally is) ^ there is even less inducement to change it to triple X // A / rhythm. Thus the final feet of I 217 hut to \ bring forth, if reduced to triple rhythm, would satisfy the conflict and form a X X / A foot with feminine ending, thus, but to bring forth. But the have, not the choice of a secondary accent as arsis accompanied by the suppres- sion of the primary stress to thesis (as in the words above cited), but simply the choice of an unusual one of two possible secondary stresses as arsis, leaving the primary accent of the word still free to form the next arsis. That is, we should /// // / expect propitiation rather than propitiation. This recession of secondary accent is doubtless nearly related to the Germanic principle of recession of primary stress. A phenomenon just the reverse of the above, — i. e. the passing of the secondary stress towards the end of the word, is to be seen in Tennyson's use of palpitated for the more usual valpitated in ' / // Palpitated, her hand shook, and we heard. The Princess, IV 370. // A // Rhythmically this may be likened to Milton's use (later to be treated) of universal II I for universal, which is, however, grammatically an example of the use of a secondary accent to the neglect of the primary stress, not of the choice between two secondary accents. ^ Cf. 3 a above. Syllabification and Accent in the Paradise Lost. 59 impropriety that the strongly stressed word forth should sink to the position of second syllable in a feminine combination — a position that is properly very light — makes the first scansion of two light + two heavy syllables when final the preferable one. For examples of this combination used non-finally in triple and double rhythm, cf. from Browning's Glove, X X / A X / From the poor \ slave whose club \or bare hands'] X //A / X / and, And the\fix'd laws \ of heav'n. II 18. For examples of this combination used finally in triple and double time, compare another line of The Glove, X X / A And bade him make sport \ and at once stir (riming with : Up and out of his den the old monster), X // A I and I 217 : bid to bring forth. These comparisons seem to show that when non-final the com- bination is suited both to triple and double rhythm ; while, as final, its preferable and more dignified use is in double rhythm. 4. Anapsestic postponement of Arsis. There is a scansion of two light + two heavy syllables which may suitably be here pointed out as erroneous. It may be called the anapaestic postponement of the arsis and is just the inverse of the phenomenon called trochaic substitution to be considered under (5). Like most errors in scansion it arises from paying undue attention to the logical or prose-emphasis. It may be represented thus : X X / / I 217, Hoio all his malice serv'd but to bring forth. The arguments against such a scansion are several, (a) It is condemned by an appeal to the rhythmic sense. (6) It leaves one X X _ foot {but to) without an arsis or verse-accent, and it gives to another foot {bring forth) two ictus, (c) A comparison with Anglo-Saxon usage. Trochaic substitution, or something very much like it, is freely allowed in Anglo-Saxon verse. Thus of type B, XJ.>^2, types E and C may be considered as variations due to trochaic , substitution in the first and second feet respectively ; thus E = ^ X X ^, and C = X ^ ^ X. go in type A,^^^^, inversion of the 60 Syllabification and Accent in the Paradise Lost. first thesis and second arsis (or trochaic substitution) produces D, IJ-XX. There is one, and only one, more imaginable combination of two theses and two arses, namely xx II. This would be a variant formed from B, x - x -, by an accent-shift the exact inverse of trochaic substitution affecting the second and third verse-ele- ments, or in other words by the anapaestic postponement of the accent. Bat this form x x ^ -^ is forbidden in Anglo-Saxon verse. It is this very same faulty rhythm which renders invalid, I believe, the reading X X / _ / Pursues as inclination or sad 5. Trochaic Substitution. The normal combination used for making an iamb is x /. 'We have seen in (1) and (2) that verse-ictus or pitch can turn x x and ^ ^ into iambs, thus x // and ^ ^. We may go further. Even a descending prose-combination, -^ x^ or ^ ^ , can be turned into an iamb by pitch, thus ^ //. For example A // VII 311, . . yielding fruit \ after \ her kind, or A // II 544, Through pain \up by\ the roots. . . . Here, as in (3), there is conflict of pitch and stress. But here, unlike the phenomenon treated there, reconciliation of the opposing accents may under conditions take place. For initially or after a strong caesura the pitch-accent passes forward and coincides with the stress and we get trochaic substitution. Thus initially : / X X / / X X / Wasting the earth II 502, Truce to his rest II 526 ; and after the caesura, where the substitution may be called ccesural : / X X / of truce ; \ I ris\ had dipt. XI 244, / X X / by thee, \ vile as \ I am. X 971. The restrictions to the passage of -^/Z to ^x (or ^^), are: (a) it shall not occur in the final foot (cf. 6) ; (6) it shall not occur twice in succession (cf 9) ; (c) it shall occur only after a distinctly marked pause, initial or csesural (cf. 7 and 8). As we have seen in the Introduction, these rules, deduced from genera] usage and an appeal to the rhythmic sense, are often violated by metricians. Thus Mr. Bridges inverts the final foot of X 840, Syllabijlcaiion and Accent in the Paradise Lost. 61 / . Beyond all past example and future. It were to be wished that the sense of the line described his method of scanning it. Unfortunately he has not been the only- one so to read it, nor yet will he be. Further, he inverts two successive feet in VI 906, / / As a despite \ done a,\gainst the \ Most High. and makes a non-csesural inversion in I 253, / A mind \ not to \ he changed by time or place. Do levelly stressed syllables ever take on the trochaic form ? Mr. Bridges says {Milton's Prosody, p. 17) "Initial weak feet are almost always made up of two monosyllables, and a slight accent will be given in reading to the first of them so that the foot is really inverted," an inverted foot being of course another expres- sion for trochaic substitution. He cites I 498, And in luxurious cities, etc. Again (p. 19) he says, "As a general rule, when the first foot is weak it will strengthen itself by a slight conventional inversion in spite of the sense, e. g. I 259, We shall be free . . . ." This, however, is a matter of individual choice, and I should prefer to read each of the above-cited initial feet as an iamb. According to the above-stated restrictions (a) and (b), trochaic substitution may occur in the verse as a whole as follows : in foot 1, (2), 3, or 4 ; in feet 1 and 3, 1 and 4, (or 2 and 4). The most frequent place for it is of course in the initial foot. Since a caesura after the first foot is rare, inversion in the second foot is very infrequent. It probably occurs in / X X / Hell! What do mine eyes with grief behold. TV 358. Inversion at once in the second and fourth feet would be even rarer. Trochee + Heavy Iamb = E^ A marked and beautiful effect, which I have not seen recognized by the metricians, is gained when the trochee precedes a heavy iamb, as in X 96, / X A / Came the mild judge .... 62 Syllabification and Accent in the Paradise Lost. The similarity frequently pointed out between type E of Anglo- Saxon verse and the common combination of a trochee + an iamb, becomes well-nigh an identity, if, more narrowly, we specify E^ and, for the modern combination, a trochee + a heavy iamb. / X \ / / X A ./ Compare 'mo7'])orbed stred and Came the mild judge. I find about 175 examples of this combination, counting both its initial and its csesural occurrences. Some of the cases could of course be differently scanned. Tlie combination is much less frequent csesurallv than initially. Initially : / X A ' / / X A / . Nor the deep troM. I 28, Both the \ lost hap\piness. I 55, / X A / On the firm brimstone. I 350, etc. Csesurally : / X A / , _ Shall long usurp ; ere the third dawning light. XII 421, / X A / _ _ Stood whispering soft, by a fresh fountain side. IV 326. 6-8. The Unstable Iamb in Normal Environment. 6. The final unstable iamb. Whatever licenses are allowed to the earlier part of the verse, the final foot must always leave on the ear the true iambic cadence. If therefore a word of the form ^ X ends a line it must always be used as an unstable iamb A //. Thus, A // Which of us who beholds the bright surface. VI 472, A// Beyond all past example and future. X 840. For a list of the unstable iambs, see 7. 7. The non-ccesural unstable iamb formed by one word. Since trochaic substitution may occur in the body of the line only after a csesura (cf. 5), we shall have to read the words the other tvay I I f\ II I II Satan, as the other way Satan not the other way Satan, although Satan is our natural pronunciation of the word. That is, if a naturally trochaic word appears in the body of the verse non- csesurally, we must have recourse to the pitch-accent to give to the word an iambic form. As it is often a matter of opinion whether or not a strong enough csesura for a following inversion (= an inversion-cassura) Syllabification and Accent in the Paradise Lost 63 exists at any given point, there is a corresponding difference of opinion whether the combination ^x shall go into verse as a trochaic substitution ^ x or as an unstable iamb ^ H . For instance in Gods, yet confest \ later \ than heaven or earth. I 509, A // / X shall we consider later to be an iamb later or a trochee later 'i That is, does an inversion-csesura exist before it? In extreme cases there can be no doubt ; thus in V 750, In their triple degrees ; regions in which. there is as unquestionably no inversion-csesura after their, as their is one after degrees. Accordingly we read A // / .X . In their triple degrees ; regions in which. with triple as an unstable iamb, but regions as a trochee. In order that we may have a criterion which is not largely subjective, let us say that an inversion-csesura shall be considered to exist only at a punctuation point. According to this, I find in the poem some 69 cases of words the ordinary forms of which are trochaic, but which must be used as an unstable iamb. An approximately full list of these words, and of other similar unstable iambs which resist the trochaic form either on account of their final position (cf. 6) or on account of an immediately preceding trochee (cf. 9), is as follows. Only one passage is cited for each word. The order is alphabetical according to the ending which bears the verse-stress. A // A // A // -ace: surface YI 472; -age: image VII 527; -al: equalVd II A // A// A // III 33, (uni)versal YI 34, trial IX 370 ; -edge : knowledge YII A // 543; -en: loaden TV 147, garden Ylll 299; -ent: orient YI 524 ; -er: never I 159, under 1 345, later I 509, whether lY 907, ever Y 810, other Y 884, mightier YI 32, after YII 311, worthier IX 100, over X 253, waters X 79; -es: races IX 33; -est: busiest XI 490 ; -y : rallied (= rally'd) YI 786, glo7^y (the second) X 722, only X 936 ; -ing : rolling II 873, shading III 357, listening YII 106, borrowing YII 177, gardening IX 203, roving IX 575, dropping IX 582, blowing X 289, arming XI 374 ; -I : 64 Syllabification and Accent in the Paradise Lost. sprinkl'd III 642, triple Y 750 ; -on : visions XI 3.77 ; -u : argues II 234; -ure: nature's IV 207, azure^ VII 479 Juture X 840. A // A // Trisyllabic words ^c audibly Y 11 518; bottomless Y S6Q ; A // A // A // A// infinite V 874 ; odorous V 483 ; popular VII 488 ; {in)visible III 586, VII 122 ; i;^MZe IX 436 ; carbuncle III 596. The following notes will justify, with one or two exceptions, the employment of the secondary stresses in the above words : (Old) / . / / . / / Fr. surface, image, equal, universal. Trial, composed of the M. E. vb. trie ' to pick out ' + the suffix -al, would be accented finally / . / in analogy to such words as equal, universal. Knowledge, coming from the M. E. vb. knowleche which accord- ing to Kluge and Lutz presupposes an A.-S. *gecndwlcecan in which the secondary stress on the second member of the com- pound would frequently fall in the arsis, is thus justified in occurring as knowledge. From A.-S. times the past-participle ending -en was capable of bearing the ictus. / ./ (Old) Fr. gardin, orient. Words in -er. Nouns of agency in -er might utilize as ictus the secondary stress on the derivative element in A.-S. Thus / \ se fiscere. In the above list there is no word of this nature ; but other words in -er, whether the -er be original as in waters, whether, under, etc., or derived from -r(e), -r{a) as in never, later, etc., have conformed to the accentuation of nouns of agency in -er. 1 For the trochaic form of the word cf. I 297, I I II On heaven^ azure ; and the torrid climb. The unstable iambic form occurs in VII 479, / / / . . . and purple, azure and green. and in IX 429, Carnation, purple, azure, or specMt .... In the last two cases Mr. Bridges scans quite erroneously, / / / . . . and purple, azure and green. Carnation, purple, azure or specMt .... This elision of -ure is not to be accepted. ^ For further remarks on bottomless and popular cf. 12 ; on infinite and odorous cf. 9. Syllabification and Accent in the Paradise Lost. 65 races : the nom. plu. ending -as does not have a secondary stress in A.-S. busiest : the secondary stress on the superlative ending -est in A.-S. may ajEford an arsis. The forms rally and argue are to be held together. In the French forms there is a following syllable which receives the / / accent : rallier, arguer. When in passing into English the Romance inflexional ending is laid aside, the accent continues final. Glory, coming from the Fr. form glorie which ends in a mute e that never receives the accent, should retain the root-accent of / . its source. We may say either that glory is accented finally in analogy to other Romance words, or under the influence of native words in -He, an ending always bearing secondary stress and capable of standing in the arsis. Such a word is only < dnllc. The ending -i7ig inherits its accent from A.-S. -ing (-ung), which, however, was a nominal not a participial derivative. It is interest- ing to observe that the ending -end of the present participle in A.-S. bore a secondary stress which could always be used as arsis. The accent on the syllabic I in the frequentative ending of the native word sprinkle and in the ending of the French word triple is not etymologically justifiable. It must therefore be due to the influence of some analogy. (Old) Fr. vision, azur, nature, etc. Of the trisyllabic words, the accents of a number are explained by the Lat. (-Fr.) accents : audibilis, odorus, visihilis, voluhilis, carbunculus. II In bottomless) we see the utilization of the secondary accent on what is felt to be a derivative syllable in analogy to the accent resting on many derivative syllables from A.-S. times. infinite is probably under the influence of the simple -^ovA. finite. popular would seem to be anomalous. Cf. Lat. populdris, and Fr. populaire. 8. The unstable iamb formed by two words. The syllables form- ing the unstable iamb may belong (not as in 7 to the same word, but) to separate words ; as in VI 796, 5 66 Syllabification and Accent in the Paradise Lost A // Against \ God and \ Messiah, or to fall. A // A // Other examples : is low raise and I 23, a veil down to TV 304, A II uncropt falls to IV 731 . The two-word composition of the ma- stable iamb is much more common than its one- word composition. It seems also to be far less unstable than the unstable iamb com- posed of one word, often not inverting into a trochee even after an inversion-caesura. Examples with a marked caesura and yet perhaps not showing inversion are A // Deep malice to conceal, couch' t with revenge. IV 123. A // Squat like a toad, close at the ear of Eve. IV 800. One may, however, invert these feet after the strong caesura. Of course a most important element to be considered in each case, where inversion is a matter of choice, is the relative strength and weakness of the first and second words. For all kinds of relation are possible, from almost level stress (when inversion would not be expected) to the case where the first syllable so out-weighs the second that inversion becomes necessary. 9-11. The unstable iamb in combination with other feet than stable iambs. 9. Trochee + unstable iamb. Inasmuch as two successive trochaic substitutions would destroy the iambic rhythm of the line, four syllables which would naturally form two trochees are put into the heroic line as a trochee + an / , A // unstable iamb. Thus we scan : Save what sin hath VI 691, II A II ■ / All I A II Universal VI 34, By the waters XI 79, To the garden VIII 299, / A // . IX A II Thro' the infinite host V 874, Spirits odorous breathes V 483, I A Jl In the visions XI 377, etc. The instances of such unstable iambs when composed of the syllables of one word are included in the list given in 7. The relative weight of the syllables composing the unstable iamb may vary. Thus more nearly levelly accented than the above examples are the iambs in : Where and what art II 681, On a sunbeam IV 556, Soon as midnight V 667, etc. These may perhaps be considered heavy, rather than unstable, iambs ; Syllahificdlion and Accent in the Paradise Lost. 67 A / as we must certaiuly consider the iamb in : Nor the deep tract 128. Remark. The unstable iamb is more unstable after a trochee, as in this section, than after a stable iamb, as in 6, 7, and 8. The reason is easy to see. In the first case, as for exam- / X A // pie in To the garden YIII 299, the suppressed primary accent on gar- tends to the arsis-position, being more prominent not only than -den but also than the. It is doubly hard, therefore, for gar- to sink to the thesis. X / A// In the second case, however, as for example in confest later I 509, the inducement to put ta- in the arsis arises solely from its prominence over -ter. The preceding syllable -fest does not (as in the first case) tend to throw la- into the arsis, the two being equally heavy. 10. Unstable iamb + heavy iamb. When the unstable iamb is in the body of the line, it will usually be both preceded and followed by the normal iamb, as in X / A // X / What I is low | raise and \ support. Here there is conflict between i-aise and and, but and and sup-^ with their level light stress, and low and raise, with their level heavy stress, show no conflict. They are differentiated into arses and theses (without conflict) by the pitch-accent. But if, in the place of the light syllable sup-, we have a heavy syllable, then evidently there will be conflict between and and the following as well as the preceding syllable. The effect is peculiar and marked : /\ II ^ I And out of good still to find means of evil. I 165. Behold me then, me for him, life for life. Ill 236. A II A I Ceased warbling, but all night tun'd her soft lay. VII 436. T- A // A / For gods ! Yet him God the Most High vouchsafes. XII 120. Initially and after a strong caesura the first foot of this com- bination may be inverted, giving a combination already treated in 5. Thus, To judge them with his saints; him the Most High. XI 705 ; and initially, 68 Syllabification and Accent in the Paradise Lost. I X A / Came the mild judge. X 96. 11. Unstable iamb + unstable iamb. The combination treated in 10 is rather jolting, and, moreover, is somewhat unstable, tending to invert the first foot to a trochee. Both the jolting effect and the unstability are increased if the arsis of the second foot is lightened and we get the combination unstable iamb + unstable iamb ; as in X 178, A // A // And dust shall eat all the days of thy life. This rhythm can hardly be considered satisfactory. Moreover it is very unstable, inverting the first foot after a verse-pause and giving the combination of trochee + unstable / X A // //A // iamb, already treated in 9 ; as. By the waters XI 79, Universal VI 34, etc.; where conflict remains only in the second foot. Perfect stability would be gained only if the second foot, too, became a trochee ; but this would make the forbidden combination of two successive trochees, and is therefore avoided. 12-13. Light iamb + unstable iamb. In the Remark at the end of 9 we saw that the unstable iamb shows special instability when it follows a trochee, as in / X A // To the garden, due to the fact of its being preceded by the trochee'fci (light) thesis. It is, in like manner, particularly un- stable when preceded by the (light) arsis of a light iamb, as in X // A // remem\ber and \fear to VI 912. This combination may be followed by the normal foot or stable iamb and then we get — 12. Light iamb -f unstable iamb + stable iamb, as in X // A // X / . . . remem\ber and \ fear to \ transgress. A // Here the instability of the unstable iamb /ear to is not increased by the following foot, as to and tra,ns- are in expiration about equal (in weakness). The instability of the foot (caused by the conflict between its two elements) is, however, heightened by the weakness of the preceding arsis and, which accordingly stands in conflict with /ear. Such a combination must stand in its unstable Syllabification and Accent in the Paradise Lost 69 form, for let reconciliation of the two conflicts take place and we get XX / X X / -ber and fear to transgress. But by so doing we lose an arsis, the resulting series of theses and arses forms two ascending trisyllabic feet, and the whole line becomes / / / / Yet fell; remember and fear to transgress. This stable form of the line cannot, of course, stand in pentametre- verse, as it has two trisyllabic feet and only four arses. The unstable form, therefore, with its conflicts must be preserved. Other examples of this phenomenon are : X // A // X / X // A //X / to the bottomless pit VI 866, in her popular tribes VII 488. These words show the use of a secondary stress as arsis immedi- ately following the primary word-stress. Cf. the unstable iambs in 7. Further examples are : X // A // X / ethe\real quin\tessence \ of heaven. Ill 716. X^// A J/ X / fel\lows am\bitious \ to win. VI 160. X II A II X I And cor\poreal \ to in-. V 413. X II A II X /. With im\petuous \ recoil. II 880. These words show the use of secondary stresses as arses both preceding and following the primary stress. Cf. 3 b and footnote. Caution. The last three words must not be scanned ambiti-ous, I II , I II corpore-al, impetu-ous. For thus there is made with (the last syllable of) the preceding word a faulty trisyllabic foot, X X / // With impet\u-ous | , etc. II , ,1 , II . I . Cf. propitiation for propiti-ation in footnote to 3 b. 13. Light iamb + unstable iamb + heavy iamb. When an unstable iamb, preceded by a light iamb, is followed, not by a normal iamb as in 12, but by a heavy iamb, we get the combination light iamb -\- unstable iamb -\- heavy iamb = 70 Syllabification and Accent in the Paradise Lost. X // 1 A // 1 A / = done a\gainst the \ Most High. YI 906. This form is, on account of the conflicts, unstable, and tends to the more stable form of two trisyllabic feet, thus, X X / X A / done against \ the Most High. Here we evidently have a reconciliation of the conflict between a- and -gainst. But, further^ there is a reconciliation of the conflict between the and Most; for in a trisyllabic foot, whether of the / \ \ / \ / form X X x^ X X Xj or x x x (as here), a distinct secondary stress, here indicated by (""), is always possible on the syllable immedi- ately preceding the arsis. Thus in the trisyllabic foot the Most High, Most, suitably to its importance in meaning, stands in the position of what may be called the strong thesis. In VI 906 and in every other parallel occurrence in heroic verse the unstable form is of course required. The Accent of Various Words. Levelly accented compounds. In 3 and 7 we find a number of words which by utilizing a secondary stress could pass into verse as a trochee or an iamb, whereas their more natural form is an iamb or a trochee. This use of the secondary accent is to be distinguished from the phenomenon seen in the use of such a word as man-kind as either a trochee or an iamb. Such a word is so nearly levelly stressed in natural pronunciation that a slight suppression of the pitch of either element will leave the other an arsis. Thus Milton writes in I 36, A / , The mother qf\ man-kind \ , but in III 275, / X / A / Found out \for man-\kind un\der wrath .... Similarly he uses as arsis either the first or second syllable of adverse, also, archfiend, consort, elsewhere, falsehood, forthwith, headlong, henceforth, into, meanwhile, mid-night, off-spring, side- long, sometimes, tenfold, thereby, therefore, therein, unblest, unwieldy, upright, warlike, whereat, wherefore, wherein, whereof, whereon, without, etc. Good-will, sea-beast, star-bright, and other like words, have final accent. But as they are of only a single occurrence, they Syllabification and Accent in the Paradise Lost. 71 -doubtless might as well have initial accent, and have, therefore, like the above words, really level accent. Certain nouns and adjectives. Among manv nouns and adjectives of foreign origin not having / / . . / recession of accent, important are : blasphemous, crystalline, instinct (noun as well as adjective), insult, maritime (cf. Lat. maritlmus), I I . I . I obdurate, remediless, retinue, sinister, and numerous others, having initial accent at present. Miscellaneous words, chiefly substantival. II I . II I II I II ^ I II ^ I Accessories, acceptable, receptacle, contribute, attribute. II I II I In propitiation XI 34, humiliation III 313, X 1092, X 1104, (P. R. I 160), we have examples of the employment not of an unusual primary but of an unusual secondary stress. Cf. foot- note to 3 b. Certain proper names. I I Aegean. In I 746 we must read Aegean not Aege-an, the accent we should expect, according to the Latin-English method of accenting classical words, from both Latin Aegceum and Gk. / / ^Airyaiov. Thus, On Lem\nos th^ Ae\gean He. The form in P. B. IV 238 is ambiguous. Thus, / / ' / Where on \ the Ae\g€an shore ... / / / or. Where on \tN A€ge\-an shore. ... Pearce's note to I 746, quoted in Todd's Milton, 1809, II 364, runs : " So he pronounces Aegean in P. R. IV 238. Fairfax led the way to this manner of pronouncing the word, or rather to this poetical liberty ; for in his translation of Tasso, C. I. 60, he says, / Cer Aegean seas, throvgh many a Greekish hold. Again, C. XII. 63, / As Aegean seas, etc." 72 Syllabification and Accent in the Paradise Lost. Serapis I 720. We must read / / Belus or Serapis, etc. Pearce's note on this word, quoted in Todd's Milton, II 361,^ runs: "There are authorities which may serve to justify in Milton this departure from the classical accent upon the second syllable of Serapis ; for we read in Martianus Capella ' Te SerSpin Nilus, etc,' and, in Prudentius, ' Isis enim et SerSpis.' " But may not Milton's accent and the shortening of the penult in these late Latin writers both go back independently to one and the same thing, namely the Greek accent of the word which was SapaTTi? ? Tiresias III 36. The ordinary form of this word in English is Tiresias, the penult being short (Gk. "Yetprjala'i) ; but we must scan III 36 thus // / And Tiresias and Phineus, prophets old. Has not Milton here taken the accent of the original Gk. form Teip7]ai,a