∞∞∞∞∞∞∞
INTRODUCTION
SYSTEMS OF PROSODY
The texts hereunder all adhere to the quantitative-foot system(1) of prosody. Although Bryant(2) was composing in Christabel metre before 1837, and Whittier(3) in strictly syllabic verse in 1829, the accentual and syllabic systems of prosody either failed to cross the ocean, or were repelled(4) as heretical. Before proceeding to the differences of detail between the texts below, or between any of them and Saintsbury, it is appropriate to mention the rival systems that have influenced the theory (and rarely the practice) of versification.
Any system of prosody can be applied a priori or a posteriori.(5) Poe, despite his express distrust of merely mathematical reasoning, is the Aristotelean prosodist. He applies his active faculty of thought to deduce, ex tohubohu, a system of feet, metres, verses, and stanzas, enlivened here and there by judiciously, i.e. unexpectedly, disposed rhymes. Saintsbury, despite his many points of agreement with Poe, is firmly Baconian: while we may occasionally doubt whether his system can have been [page xvi:] induced from the documents, it is unquestionable that he has studied the documents.(6)
The syllabic system(7) is, of course, the basis of modern French versification.(8) It works, after a fashion, on the decasyllables of Pope, and even of Milton. It is the correct system for scanning verses — it is too much, pace Whittier, to say poems — written to hymn-tune metres.(9) James VI Stuart(10) and Edward Bysshe(11) are the classic writers on non-hymnodic scansion by syllables.
The accentual system(12) may be paraphrased thus: place four distressed syllables in a line; fill with unstressed syllables ad libitum; the result will be an English verse. As an account of the construction(13) of Christabel it is inadequate: first, monosyllabic rhyme requires that the fourth distress end the verse; sec ond, except perhaps in ‘Tu — whit — tu — whoo!’, distressed syllables do not regularly abut. The system has lately attracted the notice of Harvey Gross,(14) who argues that Anglo-Saxon poetry rests on a basis of four heavy distresses to the line, and pretends that Pound and Eliot, inter al., have revived the Anglo-Saxon line.
The accentual-foot system(15) retains from the quantitative-foot system the Greek names of feet, and scansion by vertical bar lines between the feet: discarding the terms short and long, it marks the beat within the foot by accents of varying shape. This compromise is not new: Felton(16) wrote in 1842 of taccented trochees and dactyles’ — accentual is perhaps a better word. Against the accentual-foot system we object that the application of the names of feet is made to fight the etymology: the tribrach,(17) the amphibrach,(18) the amphimacer(19) directly imply the proscribed short and long; the dactyl(20) implies them by a well-known metaphor.
The musical system,(21) holding that the scansion of verses should be expressed by musical time-signatures, bars, notes, and rests, has been elaborately exposed by Sidney Lanier.(22) Allen(23) objects that Lanier's scansion is not self-evident. We consider that Tovey's remarks s.v. rhythm in the 1929 Encyclopaedia Britannica(24) are a cogent refutation of Lanier. The specific use of [page xviii:] rests makes the system more convenient than the foot system for recording feet of one syllable and no syllables.
THE QUANTITATIVE-FOOT SYSTEM
The texts below may be divided between the go-as-you-please foot system (Poe 1837, 1843, 1846; Bryant; Brown) which denies, or rather never asserts, that all homologous feet shall occupy exactly equal times;(25) and the metronomic(26) foot system (Poe 1847, 1848, 1850; Whelpley) which asserts that they shall. To deduce a complete comparative system from these texts would be further comi1 plicated by the presence of irrelevant matter: Poe attempts to bring Latin hexameters, asclepiads, and sapphics under what should be a system of English scansion; Whelpley does the same for Greek iambics. We attempt an outline of English foot prosody based on these texts; the order of points is modelled after Saintsbury, Historical manual, pp. 30-35.
1. In English versification, syllables of varying length are grouped into feet; and feet into verses, i.e. lines.
2. Two lengths of syllables, short and long,(27) are recognized by all. Poe 1848 adds three species of shorter-than-short (one of which recurs in 1850) and two of longer-than-long; and we have ventured(28) to add one longer yet.
3. The long syllable is reckoned equal to two short: in go-as-you-please scansion, a conventional or approximate equality; in metronomic scansion, an absolute equality.
4. Feet of two and of three syllables are recognized by all; Poe adds feet of one and of four.
5. The monosyllabic foot, peculiar to Poe, is always long, or rather longer-than-long.(29) It is regularly used at the end of a verse, but it can be used in the interior. We consider it the most natural way to bring 7-syllabled verses(30) under the foot system; Brown is content to scan these indifferently as trochaic of four feet catalectic and iambic of four feet acephalous.(31)
6. Of disyllabic feet, all admit the iambus and trochee. Poe admits the spondee, apparently because of its occurrence in Latin hexameters, and calls it the rudiment of verse; Whelpley admits it, and, apparently with the view of showing that English blank decasyllables are isochronous with Greek iambics, makes it the backbone of the former.(32)
7. Of trisyllabic feet, all admit the anapaest. Bryant admits the tribrachys as a foot of substitution for the iambus; it is equal in metronomic time to the iambus, and would have saved Poe some Procrustean adjustments in constructing the bastard iambus. Poe rejected it, we presume, by an extension of the argument that rejected the pyrrhic.(33) Brown admits the dactyl, with the important qualification that dactylic verse is uncommon;(34) Whelpley, as a foot of substitution for the iambus; Poe, as a stick to beat Longfellow withal. We argue for the amphibrachys below.
8. Trisyllabic feet peculiar to the metronomic Poe are the bastard iambus(35) and the bastard trochee.(36)
9. The metronomic Poe introduces three tetrasyllable feet: the bastard anapaest and bastard dactyl,(37) and the quick trochee.(38) He has misgivings about the existence and desirability of the quick iambus.(39)
10. The metrical quantity of syllables depends not on the vowel quantity,(40) but on distress or on consonants that impede the enunciation — preferably both.
11. All monosyllables, according to Poe, are common(41) — but he [page xx:] condemns Bryant and Byron for making a, an, the long.(42)
12. For an iambus, all recognize the substitution of a trochee; all except the metronomic Poe, an anapaest (he must use a bastard iambus); Bryant, a tribrachys; Whelpley, a dactyl.
13. For a trochee, Brown allows the substitution of a dactyl; Poe, a bastard trochee or (rarely.) a quick trochee.
14. For an anapaest, Brown allows the substitution of an iambus; Poe, a spondee.
15. For a dactyl, Poe and Whelpley allow the substitution of a spondee. Felton(43) concedes — as a matter of fact, not of law — that trochees are inserted in what pass for English hexameters.
16. In alexandrines, Poe demands a pause after the third foot.
17. Poe admits stanzas of unequal lines — even Pindarics — but impatiently rejects lines of one foot.44
18. Blank decasyllables are admitted by all; English hexameters are condemned by Poe and treated in passing by Whelpley. All other English verses in this book are rhymed.
19. Double rhymes are quoted by Poe from Byron and by Whelpley from Milton. Triple rhymes are exhibited by Brown and Poe, but only exempli-gratia.
20. The orthodox place for rhyme is at the ends of verses. Brown exhibits triple rhymes in hemistichs; Poe argues that rhymes should be put in the best place by ear, regardless of convention.
21. Poe considers rhyme an important binding force in a stanza, and stigmatizes as ‘loose and ineffective’ stanzas in which any rhyme occurs more than twice.(45)
22. About the connexion of rhyme with the foot system, the writers below are silent. We have essayed to forge a connexion: see pp. 67 ff. below, note 59.
DIFFICULT SCANSIONS
Four of Poe's lyrics are scarcely tractable by Poe's quantitative- foot system — perhaps by any system, viz:
To these it is convenient to add
[5]. Longfellow's verses beginning ‘The day is done, and the darkness’,(49) which Poe confessed(50) that he could not scan.
1. The accentual system forces this lyric into the four-stress mould by counting each printed verse as a hemistich, and allotting it two stresses: a correct but jejune and inadequate scansion.
The syllabic system, applied a posteriori, begins by making a syllable count of the verses, as printed; and finds 24-syllabled verses, 186-syllabled, 117-syllabled, 3 8-syllabled, and 2 [page xxii:] 9-syllabled. After eliding carriers in v. 94, many a in v. 108, and (with a touch of lockjaw) environs in v. 117, it revises the count to 24-syllabled verses, 206-syllabled, 97-syllabled, 48-syllabled, and 19-syllabled (v. 102). A Procrustean prosodist will now state, nonobstante caleulo, that there are nine couplets of alexandrines, printed in hemistichs and marred by damnable licences; a liberal prosodist, wishing to save the phaenomena, will find a decasyllable couplet (vv. 82-85); two couplets of fourteeners (vv. 90-93, and, with elision of the, vv. 102-105); three S.M. stanzas (vv. 106-117); and twelve miscellaneous numeri lege soluti.
The go-as-you-please foot system will note the general intern* tion to be iambic of three feet,(51) and mark the variants as tro** chaic and anapaestic substitution and lines of two and four feet; the accentual-foot system will reach much the same foot-division with different names. The metronomic foot system, as we apply it, differs from the above in two points: first, we find tribrachyses in v. 87 and bastard iambi in vv. 89, 90, 91, 92, 93, 94, 97, 99, 100, 102, 108, where liberal scansion finds anapaests; second, we scan thy Nesace in v. 102 and Infinity in v. 104 as iambi with two added syllables each, for the reasons sketched at p. 70 below.(52) Poe's scansion would be likely to contain two infelicities: first, he could scan vv. 82-85 in dactyls; second, vv. 90-93, 94, 97, 99-100, 102, in trochees.
2. The accentual system deals with this as with the previous lyric: indeed, a capital shortcoming of four-stress-ism is its impotence to discriminate the metre of
from that of
The syllabic system notes with satisfaction that the lyric, in general, will go to the hymn tune ‘The Lord is my shepherd’.(55) Verses 92-95 do not exactly fit this mould? — then the organist must vamp a little.
The scansion given this lyric by the foot system depends on whether the amphibrachys(56) or the anapaest be taken as the principal foot. We offer both scansions: we favour the amphibrachic. Could Nesace have ejaculated ‘Ligeia! Ligeia!’ if it were given to her to know that Saintsbury would hack the symmetry of that verse into an iambus, an anapaest, and a supernumerary syllable?
Taking the quatrain opposite note 54 as a model of this lyric, [page xxiii:] we scan vv. 112 § 114 with two amphibrachyses, 113 § 115 with an amphibrachys and an iambus. On metronomic principles this will not do: the amphibrachys is a tetrachronic foot; the iambus, trichronic. The metronomist can scan may be and from thee as spondees without great strain. In v. 71, away will not pass as a spondee; to be both amphibrachic and metronomic, we must give -way time and a half (at the end of a verse, this is easy) and scan
Th2e moonbe2am | a2w3/2ay | .(57)
Anapaestic scansion cuts each verse into an iambus and an anapaest, and adds a hypermetrical syllable to accommodate the double rhymes in vv. 112 $ 114. This scansion is not metronomizable: while we may make a spondee of Ligei, Thy im, or Thy mu, we cannot torture The in The moon (in v. 71) into the needed length. [page xxiv:]
Poe — we judge by his mishandling of ‘The bride of Abydos’ — would have ignored the rhymes, and scanned in continuous dactyls, varied by spondees and caesuras: vv. 68-99 would emerge from this process as
Nea2th | blue-be2ll o2r | streamer Or | tufte2d wi2ld | spray That | keeps
fr2om th2e | dreame2r Th2e | moonbe2am a2 | way Bright | bein2gs th2at |
ponde2r Wi2th | half cl2osi2ng | eyes O2n th2e | stars whi2ch yo2ur | wonde2r
Ha2th | drawn fr2om th2e | skies Ti2ll th2ey | glance thr2o' th2e | shade
an2d Co2me | down t2o yo2ur | brow Like | eyes o2f th2e | maide2n wh2o |
calls o2n yo2u | n3/2ow A2 | rise fr2om yo2ur | dreamin2g I2n | vio2le2t | bowers
To | duty2 b2e | seemin2g The2se | star-li2tte2n | hours And | shake fr2om
your | tresse2s E2n | cumber2'd wi2th | d3/2ew The | breath o2f the2se | kisse2s
Th2at | cumbe2r th2em | too O | how wi2thout | you Lo2ve Cou2ld | ange2ls
b2e | blest Those | kisse2s o2f | true lo2ve Th2at | lull'd y2e t2o | rest
Up | shake f2rom yo2ur | wing Each | hinde2rin2g | thi3/2ng th2e | dew o2f
th2e | night i2t wou2ld | weigh do2wn yo2ur | flight And | true lo2ve ca2 |
resse2s O2 | leave the2m a2 | part Th2ey ar2e | light o2n th2e | tresse2s Bu2t |
lead o2n th2e | hea2rt.
This lame scansion, which contravenes what we apprehend to be the reading-flow — amphibrachic where Poe can find double rhymes, ana’ paestic where he makes do with single rhymes and enjambment — can be protracted to the end of the lyric; the only instructive variation is the bastard dactyl witchery may in v. 153.
3. The metre of ‘The haunted palace’ offers difficulties to any system of scansion. Because our attempts to scan it require continual reference to the text, we reprint it in full; premising our conclusion, that no composer imbued with the foot system would have offered ‘The haunted palace’ as a strophic song.(58)
In the greenest of our valleys
By good angels tenanted,
Once a fair and stately palace —
Radiant palace — reared its head.
5
In the monarch
Thought's dominion —
It stood there!
Never seraph spread a pinion
Over fabric half so fair!
Banners yellow, glorious, golden,
10
On its roof did float and flow, [page xxv:]
(This — all this — was in the olden
Time long ago,)
And every gentle air that dallied,
In that sweet day,
15
Along the ramparts plumed and pallid,
A wingedx odour went away.
Wanderers in that happy valley,
Through two luminous windows, saw
Spirits moving musically,
20
To a lute's well-tunedx law,
Round about a throne where, sitting
(Porphyrogene!)
In state his glory well befitting,
The ruler of the realm was seen.
25
And all with pearl and ruby glowing
Was the fair palace door,
Through which came flowing, flowing, flowing,
And sparkling evermore,
A troop of Echoes, whose sweet duty
30
Was but to sing,
In voices of surpassing beauty,
The wit and wisdom of their king.
But evil things, in robes of sorrow,
Assailed the monarch's high estate.
35
(Ah, let us mourn! — for never morrowx
Shall dawn upon him desolate!)
And round about his home the glory
That blushed and bloomed,
Is but a dim-remembered story
40
Of the old time entombed.
And travellers, now, within that valley,
Through the red-litten windows see
Vast forms, that move fantastically
To a discordant melody,
45
While, like a ghastly rapid river,
Through the pale door
A hideous throng rush out forever
And laugh — but smile no more.(59)
Mere inspection, not committed to any system of scansion, discloses 48 verses, printed in 6 stanzas of & verses each, with di syllabic rhymes(60) in the odd verses and monosyllabic rhymes in the reader were tempted to voice the b in entombed. [page xxvi:] even verses; the sixth verse of each stanza being short. The accentual prosodist will get off fairly well by assigning one distress to each sixth verse and two to each of the others; but this paradigm breaks down at v. 27, which requires three distresses, and can receive five.
The syllabist, counting syllables as printed, finds that the short lines number 1 3-syllabled and 5 4-syllabled; the ordinary lines, 1 4-syllabled, 4 each 6- § 7-syllabled, 16 each 8- § 9-syllabled, and 110-sy1labled. Elisions in vv. 3, 9, 17, 18 alter this count to 14-syllabled, 46-syllabled, 67-syllabled, 168-syllabled, and 159-syllabled. The best that can be made of the metre by counting syllables is 9.8.9.8.9.4.9.8(61) with many inexplicable shortages.
Go-as-you-please foot scansion, either accentual or quantitative, has three options: to scan in trochees, with Time in v. 12 a monosyllabic foot, and anacrusis in vv. 13-16, 23-48; to scan in iambi, with v. 6 a trisyllabic foot, In that in v. 14 a substituted trochee, and vv. 1-5, 7-11, 17-21 acephalous; to scan every line for itself, in the manner of Poe's scansion of ‘The last leaf’ (pp. 73-76 below.)(62) This third scansion, retaining Poe's enjambment of lines of less than two feet and Poe's name caesura for the monosyllabic foot, would run roughly thus: vv. 1, 3, 7, 11, 19, 21, four trochees; vv. 2, 8, 10, 20, three trochees and caesura; v. 4, dactyl, two trochees, and caesura; vv. 5/6, five trochees and caesura; v. 9, two trochees, dactyl, and trochee; vv. 12, 14, 22, 30, 46, trochee and iambus; vv. 13, 15, 23, 25, 29, 31, 33, 35, 37, 39, 43, 45, four iambi and hypermetrical syllable;(63) vv. 16, 24, 32, 34, 46, four iambi; vv. 17, 27, dactyl and three trochees; v. 18, trochee, dactyl, trochee, and caesura; vv. 26, 40, trochee and two iambi; vv. 28, 48, three iambi; v. 38, two iambi; vv. 41, 47, iambus, anapaest, two iambi, and hypermetrical syllable; vv. 42, 44, trochee and three iambi. Poe's metronomic system accommodates ‘The haunted palace’ by ignoring the rhymes and scanning in continuous trochees: a scansion of vv. 9-16 should be a sufficient illustration.
Banner2s | yello2w | glor4iou4s | golde2n | On it2s | roof di2d | float an2d |
fl3/2ow | This al2l | this wa2s | in th2e | olde2n | Time lo4ng a4 | go An2d |
ev4er4y | gentl2e | air th2at | dalli4ed I4n | th3/2at | swe3/2et | day A2 | long
th2e | rampar2ts | plumed an2d | palli4d A4 | winge2d | odo2ur | went a2 | wa3/2y | .
The scansion we prefer for these verses is paeonic.(64) After making the elisions in vv. 3, 9, 17, 18, we scan v. 6 as an anapaest; the 4-syllabled verses as fourth paeons; the 6-syllabled, identification. [page xxvii:] as a trochee (or iambus) and a fourth paeon; the 7-syllabled, as a paeon (first or third) and an anapaest; the 8-syllabled with monosyllabic rhyme, as a second paeon and a fourth paeon; the 8- syllabled with disyllabic rhyme, as a paeon (first or third) and a third paeon; the 9-syllabled, as a tribrachys, a trochee, and a third paeon (except v. 27, a dactyl and three trochees).
The shortest and most convincing account is given by the prosodist that relies on musical analogies: faced with a monody in lines of unequal length, with rhymes and off-rhymes intermingled, and accompanying (or accompanied by) a fantasia on the guitar,(65) he pronounces it a talking blues.(66)
4. The general metrical structure of ‘Ulalume’ is clear: ten Praed stanzas(67) (the eighth imperfect) farced out from the normal 8 lines to 9, 10, 11, 12, or 13. Except in stanzas & § 9, we can [page xxviii:] reverse the farcing by deleting vv. 3, 12, 14, 22, 25, 32, 41, 46-48, 53, 60, 63, 66, 70, 97, 100; the residuum is eight metrically and syntactically complete octastichs. Poe seems to have imitated the Praed stanza from Byron's verses beginning,Though the day of my destiny's over’.(68)
To Saintsbury's go-as-you-please quantitative - foot scansion of the Praed stanza(69) we have nothing to add. Accentual-foot scansion of ‘Ulalume’ is sufficiently illustrated by Allen's scansion of the first stanza.(70) The strict accentual prosodists will be loth to give more than two distresses to the normal verse.(71) The syllabist will have tolerable success, recognizing a mixture of 8- syllabled and 9-syllabled verses, which disyllabic rhyme lengthens to 9 § 10 syllables; but lockjaw is to be feared from the necessary elision of flickers in vv. 66 § 70.
Metronomic scansion in anapaests, continuous through each stanza, requires a foot not in Poe's apparatus, the acephalous anapaest (p. 129, note 23, below): The skies, at the very outset, will not pass as a spondee. We scan vv. 20-29:
Our talk | ha2d be2en se | r3iou3s an3d so | be3r Bu3t ou3r thoughts | th2ey
we2re pal | si2ed an2d sere | Our mem | o3ri3es we3re trea | ch3erou3s an3d
sere | Fo2r w2e knew | no2t th2e month | wa2s O2cto | be3r An3d w3e marked |
no2t th2e night | o2f th2e year | Ah night | o2f al2l nights | i2n th2e
year | We no | te3d no3t th3e dim(72) | la2ke o2f Au | be2r Tho2ugh once | w2e ha2d
jour | ney2ed do2wn here | We2 r2emem | ber3ed no3t th3e dank | ta2rn o2f Au |
be3r No3r th3e ghoul- | ha2unte2d wood | la2nd o2f Weir | .
The underscored sere illustrates the licence of rhyme between an anapaest and a bastard anapaest;(73) the italicized Auber illustrates that continuous scansion in isochronous feet is antagom* istic to the parallelism of the verses.(74) Surely appositional repetition(75) enforces that the reading flow of dim lake of Auber and dank tarn of Auber shall be identical; but the first Auber is scanned as equal to l | long syllables, the second to l | . We have, then, what Poe's delicate ear must condemn as unequivalent feet — the unutterable flesh. To save the phaenomena we specifically adn mit that disyllabic rhyme in anapaestic verse requires a fraction of a foot;(73 [[76]]) vv. 26-29 can then be scanned without enjambment:
W2e note2d | no2t the dim | la2ke o2f Au | be2r
Though once | w2e h2ad jour | ney2ed do2wn here |
W2e re2mem | ber3ed no3t th3e dank(76) | ta2rn o2f Au | be2r
No2r th2e ghoul- | ha2unte2d wood | la2nd o2f Weir | .
Huxley's allegations that ‘Ulalume’ is composed in a walloping [page xxix:] dactylic metre(77) and the stanzas to Augusta in a dactylic permanent wave(78) are surprising — especially since we state below(79) that Poe never composed in dactyls. We recall that to wallop (sc. gallop) is the recognized affect of anapaests,(80) not of dactyls. [page xxx:] Huxley had enough Latin to tell an anapaest from a dactyl if he had bothered to. From the facts that Huxley quotes Milton from memory,(82 [[81]]) that his gallery of horrors adds Moore(82) and Hood(83) to Poe and Byron, and that he speaks of dactyls, we conjecture that he dashed off his essay at a country-house weekend, with no books of reference except a volume of Poe's poems § essays (probably the Everyman, but possibly Brimley Johnson's); noted that in Rationale Poe spoke of dactylic lines in ‘The bride of Abydos’; and transcribed the word dactylic without bothering to verify what Byron's, or Poe's, metre was. The fact that Poe himself could easily have scanned, or attempted to scan, ‘Ulalume’ in dactyls (rejecting the initial The, or even scanning it as a caesura) will not save Huxley from the imputation that, in Vulgarity ‘ he scanned Poe like a mouton de Panurge.
We are unwilling to quit the subject of ‘Ulalume’ without mentioning that Campbell's reading of immemorial as memorable (84) is untenable; Poe clearly took the word in its radical sense without memory: but choice between the glosses whereof memory surviveth not and wherein [my] faculty of memory was absent we leave to the metaphysical psychologists, with what assistance they can derive from v. 22.
5. To the twelve lines cited by Poe it is expedient to add vv. 25-32:
25
Read from some humbler poet
Whose songs gushed from his heart,
As showers from the clouds of summer,
Or tears from the eyelids start;
Who through long days of labor,
30
And nights devoid of ease,
Still heard in his soul the music
Of wonderful melodies.
The syllabic system can make nothing of these verses. It counts, in the blank lines, 2 7-syllabled, 5 8-syllabled, 2 9-syllabled, and 1 10-syllabled; in the rhymed lines, 2 6-syllabled, 5 7- syllabled, and 3 8-syllabled; and pronounces the normal stanza to be 8.7.8.7, which strikes an average between all by fitting none.
To scan Longfellow in feet, it is convenient to begin with Bryant's ‘delightfully delicate and beautiful’ translation of Iglesias, pp. 34-36 below. Bryant's metre is paradigmatically regular, and is iambic of three feet with hypermetrical syllable in the blank lines;(85) and the paradigm will be recognized in Longfellow's vv. 29-30. Now substitute anapaests (or bastard iambi) ad libitum, and trochees at the beginning of vv. 6 § 25 — substitutions well within Poe's comprehension — and Longfellow's metre emerges.
Why, then, did this scansion defeat Poe? Psychologically, it is arguable that, his feud with Longfellow & Outis still warm in [page xxxi:] his heart, he would or could not praise Longfellow without damning him as well.(86) Metrically, the difficulty is with the hypermetrical syllable. Although Poe's practice embraces double rhyme in anapaestic verse (regularly in ‘Ulalume’) and an eleventh syllable in blank verse (quite frequent in ‘The Coliseum’(87) , Poe's theory has not digested this syllable. In that theory any syllable remaining after dividing a verse into feet must either be made a foot (which works in trochaic verse, and even in elegiacs) or be carried forward in that chimera, a continuous scansion. Continuous scansion in iambi would make an iambic trimeter (six feet) of each of Longfellow's distichs, and entail such quick iambi(88) as me that my soul, ing that is not, ly as the mists.
A word should be said about Poe's suggestion that Longfellow's anapaests are confounded with dactyls.(89) The single verse
Gleam through the rain and the mist
is obnoxious to scansion as two dactyls and a monosyllabic foot: but embedding a verse so scanned in a quatrain yields 9.7.9.7, a metre very different from Longfellow's:(90)
Watch while the | lights of the | villages
Gleam through the | rain and the | mist,
Flooding my | spirit with | intricate
Woes that for | ever per | sist. [page xxxii:]
CONCLUSIONS
1. The quantitative-foot system is alien to the hearts of most, and the thoughts of many, English composers: but, discreetly used, it can give a more detailed account of the phenomena of English verse than its competitors.
2. As elaborated and metronomized by Poe in Rationale, the system is incomplete. We offer the following additions, congruous with Poe's fabric: the hemiola,(91) the acephalous anapaest, the tribrachys, the amphibrachys, and the fractional foot induced by disyllabic rhyme or feminine endings in blank decasyllables.
3. Anacrusis — a fractional foot at the beginning of a verse — is consistent with metronomic scansion, but seldom wanted. A distressed syllable in anacrusis(92) is an unnecessary complication, because such syllable can be scanned as a foot; an additional unstressed syllable before an iambic or anapaestic line is scanned as a bastard iambus or bastard anapaest; anacrusis in dactylic metre is rare because dactylic metre is rare; and so anacrusis is practically confined to trochaic measures. If it occurs throughout, we replace the trochaic scansion by iambic; if it occurs sporadically, it upsets the equality demanded by metronomic scansion.(93) There remains the possibility of anacrusis as a regularly recurring accident of trochaic stanzas — say in the fourth line of every hexastich: such metre is theoretically feasible, but we have not seen the experiment tried, and we suspect that the bump would grow tedious on repetition.
[The following footnotes appear at the bottom of page xv:]
1 The same as Saintsbury's foot system, Historical manual, pp. 19-36 et passim. We introduce the compound epithet quantitative-foot for distinction from the accentual-foot system, below.
2 In ‘Rizpah’. See p. 26, note 23, below.
3 Gay Wilson Allen, American prosody, New York: Octagon Books, 1966, p. 131, attempts to scan ‘The spirit of the North’, which is in perfectly regular 7.7.7.7.D, by the accentual-foot system.
4 So Poe on Coleridge, p. 121 below, text opposite note 3.
5 Poe, ‘Mellonta tauta’, Harrison 6:205, offers, as an alternative to the a priori and a posteriori roads to Truth, the great highway of Consistency. We have not seen a consistent system of English scansion, and ask — for information — whether there can be induced, from the corpus of Milton's verse, laws that will apply, with a minimum of pulling and hauling, to all writers from (say) Gower to Tennyson.
[The following footnotes appear at the bottom of page xvi:]
6 Though Harvey Gross (in his introduction to Historical manual, p. xxi) appears to be damning Saintsbury with faint praise, his imputation of ‘thoroughgoing empiricism’ is probably intended to make the same point.
7 Saintsbury, Historical manual, pp. 14-18.
8 Félix Gaiffe et al., Grammaire Larousse du XXe siècle, Paris, 1936, pp. 406-408, 412-419, wisely expose French versification in terms of syllabes, eschewing the term pied altogether; and cf. the quotation from Fauchet, p. 109 below, note 48 s.f.
[The following footnotes on page xvii:]
9 These may be roughly defined as 8.8.8.8, 8.7.8.7, 8.6.8.6, 7.7.7.7, 7.6.7.6, 6.6.8.6, 6.6.6.6, and their extensions to 6, 8, 5 12 lines. Even some apparent decasyllables may be best referred to the syllabic system: take the hymn beginning ‘Turn back, O man, forswear thy wicked ways’.
We forbear to speculate how, or why, the scholars erected a bulkhead between poetry and piety: but it is easy to verify that neither Sternhold § Hopkins nor Tate § Brady figures in Allen's or Saintsbury's index.
10 Ane schort treatise, conteining some rewlis and cautelis to be observit and eschewit in Scottis poesie ‘ Edinburgh, 1585. What we have seen is the reprint in Robert S. Rait, ed., A royal rhetor* ician, &c., New York: Brentanos, 1900, pp. 1-27. It will be seen that, although James does not define fete, they comprise all the syllables of a line except the hypermetrical syllables occasioned by double and triple rhyme, or following the tenth syllable in blank decasyllables. James's Treatise is, in general, more admired than read; and Saintsbury must have translated fete, at aventure, as feet rather than pieds.
11 See p. 54, note 28, below.
12 Saintsbury, Historical manual, pp. 6-13. See p. 121, note 3, below.
13 Save that Coleridge, being a genius, may be assumed not to have constructed Christabel: he wrote it currente calamo, and analysed it (badly) afterward.
14 Sound and form in modern poetry, Ann Arbor: Univ, of Michigan Press, 1964.
15 Allen, American prosody, throughout. The syllable-stress system of Gross, Sound & form, is much the same thing.
16 See p. 79, note 80, below.
17 Allen, American prosody, pp. xxvi, 328.
18 Allen, pp. xxvi-xxvii, 323.
19 Allen, pp. xxvii, 323. 20Allen, pp. xxvi, 324. 21Allen, pp. 281-283.
22 The science of English verse, New York: Scribners, 1880.
23 American prosody, p. 283.
[The following footnotes appear at the bottom of pages xvii and xix:]
24 What we have seen is the reprint in Donald Francis Tovey, The forms of music, Cleveland: Meridian Books, 1956 ‘ pp. 174-191. See, [page xix:] for the point at issue here, the passage on pp. 181-184, beginning ‘Let us now try’; and imagine, if possible, how Tovey would have barred a twelve-bar blues.
[The following footnotes appear at the bottom of page xix:]
25 Saintsbury's prosody is emphatically of this school: Historical manual, p. 32, &c; and see p. 131, note 25, below.
26 We owe this epithet to Allen, American prosody, p. 58.
27 There is not and need not be any consensus what makes a syllable long. Effectively, any syllable is long which scansion declares to be long and the reader can dwell on without shockingly disturbing the normal enunciation. Length can arise from quantity of the vowel or diphthong; from concurrence of two or more consonants; from distress (not necessarily heavy) ; from heavy punctuation enforcing a pause (Jose Garcia Villa's commas are inadequate). The root of Whelpley's spondaic scansion seems to be that only when a syllable has failed all these tests will he pronounce it short.
28 See p. 123 below, note 7 s.f.
29 Saintsbury, no metronomist, concedes that it requires time and a half at least: Historical manual, p. 23.
30 In particular, Namby Pamby.
31 Brown does not use the term acephalous. Allen [American prosody, pp. xxx, 325) speaks of a headless line, which is quite correct, but sorts ill with Greek terms.
32 On p. 126 below, note 15, we posit a further disyllabic foot as requisite in metronomic scansion.
33 See p. 126, text opposite note 17. Quaere: what would Poe, or Whelpley, have made of the molossus?
34 See p. 185.
35 See p. 115; and p. 156, text opposite note
36 See p. 116.
37 See p. 130.
38 See p. 124.
39 See p. 134; and cf. p. 27, note 23.
40 This did not prevent nineteenth-century printers from indicating syllabic quantity by marks attached to the vowel.
41 See p. 58 below, text opposite note 43.
[The following footnotes appear at the bottom of page xx:]
[[42 There is no text for note 42 in the original printing]]
43 See p. 79, note 80; and cf. Saintsbury, Historical manual, p. 32, note 1: ‘The combination of dactyl and trochee in English, however, will not produce the same effect as the combination of dactyl and spondee in Latin or Greek.’
[The following footnotes appear at the bottom of page xxi:]
44. We consider these lines legitimate: see p. 77, note 76, below.
45 This is our gloss on the first paragraph on p. 110; Allen (American prosody, p. 60, note 10) finds the concept of the loose stanza ambiguous. The only common taut stanzas are the quatrain and the sonnet (either Petrarcan or Shakespearian).
46 Campbell, pp. 37-38, vv. 82-117; pp. 42-45, vv. 68-155.
49 Henry Wadsworth Longfellow, ‘Proëm’, in The waif: a collection of poems, 5. ed. ‘ Boston: Wm. D. Ticknor & Co., 1846, pp. [ix]-xi. Quoted in full by Poe, ‘The poetic principle’, Harrison 14:276-277.
50 Poe, ‘Marginalia’, Harrison 16:79-81:
The conclusion of the Proëm in Mr. Longfellow's late “Waif” is exceedingly beautiful. The whole poem is remarkable in this, that one of its principal excellences arises from what is, generically, a demerit. No error, for example, is more certainly fatal in poetry than defective rhythm; but here the slipshodness is so thoroughly in unison with the nonchalant air of the thoughts — which, again, are so capitally applicable to the thing done (a mere introduction of other people's fancies) — that the effect of the looseness of rhythm becomes palpable, and we see at once that here is a case in which to be correct would be inartistic. Here are three of the quatrains —
[5]
I see the lights of the village
Gleam through the rain and the mist,
And a feeling of sadness comes o'erx me
That my soul cannot resist —
[9]
A feeling of sadness and longing
That is not akin to pain,
And resembles sorrow only
As the mists resemble the rain.
[41]
And the night shall be filled with music,
And the cares that infest the day
Shall fold their tents like the Arabs,
And as silently steal away.
Now these lines are not to be scanned. They are referable to no true principles of rhythm. The general idea is that of a succession of anapaests; yet not only is this idea confounded with that of dactyls, but this succession is improperly interrupted at all points — improperly, because by unequivalent feet. The partial prosaicism thus brought about, however, (without any interference with the mere melody,) becomes a beauty solely through the nicety of its adaptation to the tone of the poem, and of this tone, again, to the matter in hand. In his keen sense of this adaptation, (which conveys the notion of what is vaguely termed “ease,”) the reader so far loses sight of the rhythmical imperfection that he can be convinced of its existence only by treating in the same rhythm (or, rather, lack of rhythm) a subject of different tone — a subject in which decision shall take the place of nonchalance.
Now, undoubtedly, I intend all this as complimentary to Mr. Longfellow; but it was for the utterance of these very opinions in the “New York Mirror’‘ that I was accused, by some of the poet's friends, of inditing what they think proper to call “strictures” on the author of “Outre-mer.”
[The following footnotes appear at the bottom of page xxiii:]
51 Saintsbury would call it iambic dimeter brachycatalectic.
52 Although Nesace/Infinity will not pass muster as a trisyllabic rhyme, it is an exceedingly feeble monosyllabic rhyme; cf. p. 32 below, text opposite note 65.
53 Campbell, p. 37, vv. 82-85.
54 Campbell, p. 44, vv. 112-115.
55 This hymn, whose metre the hymnals catalogue as 11.11.11.11, is a metrical paraphrase (by James Montgomery) of the psalm, set to the air of Thomas Koschat's popular song ‘Verlassen, verlassen, verlassen bin i’.
56 Saintsbury, Historical manual, p. 268:
AMPHIBRACH. — A foot of three syllables — short, long, short (^ — ^) — literally “short on each side.” According to some, this foot is not uncommon in English poetry, as, for instance, in Byron's
The black bands | came over
The Alps and | their snow,
as well as individually as a foot of substitution. Others, including the present writer, think that these cases can always, or almost always, be better arranged as anapaests —
The black | bands came o | ver
The Alps | ‘and tlieir snow,
and that the amphibrach is unnecessary, or, at any rate, very rare in English.
Byron's verses (from ‘The deformed transformed’) have been suggested as a source for Poe's metre: see Campbell, p. 185, note 68.
57 It is a shortcoming of our proposed extension of the metronomic- foot system to feet not contemplated by Poe, that it must assign the same numerals to feet (or fragments of feet) so different to the ear as the first two syllables of an amphibrachys (above) and the acephalous anapaest (p. 129, note 23, below).
[The following footnotes appear at the bottom of page xxv:]
58 It is arguable that whenever Poe's lyric muse took up the pen she ignored his apparatus of feet and wrote by ear: this extreme position is not necessary to the present scansion and we do not pursue it.
59 Poe, The raven and other poems, New York: Wiley & Putnam, 1845, pp. 29-30. What we have seen is the reprint on pages 277 £ 281 of Harry Clark Snider, An edition of the poems in Poe's last collection based largely on his own critical principles, diss. Ph.D. Univ. Michigan, 1962. (Univ. Microfilms, no. 63-6969) The indentation of vv. 6, 14, 22, 30, 3$, 46 is ours, not Wiley § Putnam's.
60 Or attempts at rhyme: the rhymes at vv. 1/3, 2/4, 13/15, 17/19, 34/36, 41/43, 45/47 are inexact; as would be vv. 38/40 if the reader were tempted to voice the b in entombed.
[The following footnotes appear at the bottom of page xxvii:]
61 9.8.9.8 is an uncommon metre, although isolated quatrains can be found in ballads mostly in 8.8.8.8; apparently the more natural way to introduce disyllabic rhymes into 8.8.8.8 is as 8.9.8.9. For an example of 9.8.9.8 executed consistently (for 104 verses) see Rego Wraddey, The tangle of zeal (New York 1960). And see Saintsbury, Historical manual, p. 92, quatrain ascribed to D. Lewis.
62 But such scansion cannot be reconciled with Poe's dictum (p. 132 below, text opposite note 31) that no one verse furnishes a satisfactory indication of metre.
63 Poe never expressly recognizes in his prosody that disyllabic rhymes occur in iambic verse: but the more Poe-like scansion of four iambi and caesura outrages the reading flow.
64 To paeonic scansion in the quantitative-foot system corresponds syncopation in the accentual-foot system: see Alien's profitable analysis of ‘The Raven’ in American prosody, pp. xxx-xxxi, 74-76.
65 Poe, (The fall of the house of Usher’, Harrison 3:284:
It was, perhaps, the narrow limits to which he [Usher] thus confined himself upon the guitar, which gave birth, in great measure, to the fantastic character of his performances. But the fervid facility of his impromptus could not be so accounted for. They must have been, and were, in the notes, as well as in the words of his wild fantasias (for he not unfrequently accompanied himself with rhymed verbal improvisations), the result of that intense mental collectedness and concentration to which I have previously alluded as observable only in particular moments of the highest artificial excitement.
66 It is a pleasure to credit Mr David E. Humez with this identification.
[The following footnotes appear at the bottom of page xxvii and xxix:]
67 Saintsbury, Historical manual, p. 114. By abruption of the last verse the Praed stanza becomes the Dolores stanza: [page xxix:]
When the Ducklet's dark doom is decided,
We will trundle him home in a trice:
And the banquet, so plainly provided,
Shall round into rose-buds and rice:
In a blaze of pragmatic invention
He shall wrestle with fate, and shall reign:
But he has not a friend fit to mention,
So hit him again!
[The following footnotes appear at the bottom of page xxix:]
68 Byron, ‘Stanzas to Augusta’, Coleridge 4:54-56. Quoted in full by Poe, ‘The poetic principle’, Harrison 14:287-289.
This attribution of Poe's source is implicit in the essay ‘Vulgarity in literature’ by Aldous Huxley (1930; what we have seen is the reprint in Robert Regan, Poe: a collection of critical essays, Englewood, N. J.: Prentice-Hall, 1967, pp. 31-35.) Huxley juxtaposes (pp. 32-33) Poe's verses 1-19 and Byron's verses 9-12.
69 Saintsbury, Historical manual, pp. 112-115.
70 Allen, American prosody, p. 77.
71 But these verses seem to require three distresses:
22 Our mèmories were treàcherous and sère;
37 Astàrte's bediamonded crèscent
103 This sìnfully scìntillant plànet.
72 Five dentals in a bastard anapaest is a strain on lingual exert1 tion: it were better (although less Poe-like) to admit an amphit1 brachus and scan
W2e note2d | no2t th2e dim | .
73 See p. 70, note 59, below.
74 See p. 133, note 31.
75 Allen, American prosody, pp. 76-77.
76 We offer this foot as a crux for Saintsbury's assertion (Historical manual, p. 289) that the paeon is ‘unnecessary in English verse’. The scansion
We remem | bered not the | dark tarn | of Au | ber
eliminates the paeon at the expence of making four feet where the ear — our ear at least — apprehends only three.
77 Regan, Poe, p. 32.
78 Regan, p. 33.
79 p. 111, note 52; p. 145, note 59.
80 Allen, American prosody, p. 16, text opposite note 9; and p. 170.
[The following footnotes appear at the bottom of page xxxi:]
81 Regan, p. 33.
82 Regan, p. 33; cf. Poe, ‘The poetic principle’, Harrison 14:282
83 Regan, p. 33; cf. Harrison 14:283-287.
85 In Saintsbury's terminology (Historical manual, p. 216 s.f.) iambic dimeter, the blank lines catalectic, the rhymed lines brachyi-catalectic. Note that a mere syllabic formula, 7.6.7.6, fails to distinguish this metre from that of ‘Good king Wenceslaus’, which is trochaic dimeter.
86 Sidney Phil Moss, ‘The Poe-Longfellow imbroglio’, Poe's literary battles, thesis Ph.D. Univ. Illinois, 1954 (Univ. Microfilms, no. 10524), pp. 96-151, devotes sixteen thousand words to this feud without citing the marginale of note 50 above — presumably because it is printed in vol. 16, rather than vol. 12, of Harrison.
87 Campbell, pp. 75-76.
88 See p. 27, note 23, and p. 134, below.
89 See note 50 above.
90 The effect of 11.10.11.10 (one foot longer than the above) is even more ludicrous : see p. 68 of Robert Graves, ‘Harp, anvil, oar’, in Gross, Structure of verse, pp. 52-71.
[The following footnotes appear at the bottom of page xxxii:]
91 The foot of note 15 on p. 126 below.
92 Pace Saintsbury, who wrenches the hexameters of Kingsley's Andromeda into five-footed anapaestic measure with anacrusis and redundancy. The simultaneous occurrence of these accidents impresses us as evidence that the prosodist barred the verse wrongly.
93 Cf. the scansion of ‘The haunted palace’, above.
∞∞∞∞∞∞∞
Notes:
In the original, the scansion examples give the numeric marks above and below the relevant characters, which is not really practical to accomplish in HTML/CSS. In an attempt to successfully represent this same idea, the above presentations go above or below the line, following the relevant characters, as much as possible. The compromise does fail to convey the distinction of when these numbers are given, in the original, above or below individual letters or between letters.
∞∞∞∞∞∞∞
[S:0 - JAG68, 1968] - Edgar Allan Poe Society of Baltimore - Bookshelf - EAP: The Rationale of Verse — a preliminary edition (Greenwood)