TRANSLATION AND TRANSLATIONS

PRINCIPLE OF COMPENSATION 75

Compensation. In such cases a translator is not

bound to take the nearest idiomatic correspondent to

the original. Suppose this presents some peculiarity

of form, say an antithesis, which cannot be reproduced

because its members are not such as would occur in

contrast in the translating language, then, if antithesis

is an indispensable part of the impression to be conveyed

by the original, the translation may be made

antithetical in some other way.

When reading Charles Lamb's ' Complaint of the

Decay of Beggars in the Metropolis' I came upon a

translation of the Epitaphium Canis byVincent Bourne,

' most classical and at the same time most English of

the Latinists,' which seems to be an example of unconscious

compensation. Bourne had written

Exiguum hunc Irus tumulum de cespite finxit,

etsi inopis, non ingratae munuscula dextrae,

carmine signavitque brevi dominumque canemque

quod memoret fidumque canem dominumque benignum,

in which the fourfold que of the last line and a half at

once arrests attention. Lamb renders

This slender tomb of turf hath Irus rear"d,

Cheap monument of no ungrudging hand,

And with short verse inscribed it, to attest,

In long and lasting union to attest,

The virtues of the Beggar and his Dog.

Here in the first part the rendering is not unreasonably

free; but at the end, where the Latin is more

obstinate, it lapses into paraphrase, in which however

the effect of the quadruple que seems to be traceable

in the otherwise motiveless doubling of the phrase 'to

attest.' In rendering Lucan's 'Dream of Pompey'

VII 9 (no. 7) I had to sacrifice the impressive poly76

TRANSLATION

syllable Pompeiani\ but I hope the reader will feel

I have given him some compensation in the following

line 1

.

Great caution undoubtedly should be exercised in

Compensation. To adopt part of a phrase of Tytler

op. cit. p. 22 'the superadded idea shall have the most

necessary connection with the original thought,' and

nothing must be introduced which the author, were

he his own translator, might not be expected to

approve. But I think Sir George Young is too absolute

whenhe says (Preface to his Sophocles, 1 888),

'

I heartily

repudiate the doctrine of compensation whereby, when

beauty has been missed, other ornament is imported

to make up the general effect.' For the principle

underlying compensation is that the translator should

deliver full weight. The metaphor is well illustrated

by a striking passage of Cicero de optimo genere dicendi

14 'non uerbum pro uerbo necesse habui reddere sed

genus omne uimque uerborum seruaui. non enim ea

adnumerare lectori putaui oportere sed tamquam

appendere!

1 The shift here is stylistic. It must be distinguished from shifts

required by some difference of idiom. As where tu, cruye, tyw do not

convey a contrast of persons but mark emphasis such as is expressed by

Eng. do, did, or by stress on a particular word as in Ter. Hec. 153

1 reddi patri autem cui tu nil dicas uiti

| superbumst,' not ' with whom

you can find no fault' but ' with whom nofault can be found',' Hor. Od.

I q. 16 'neque tu choreas,' ' nor dances.' So <T6ye Plato, Gorg. 527 D,

tyu t(j.ol, Demosth. Phil. Ill 17, where English would stress the

verbs. Mr Tolman (p. 56), forgetting that printed English now refuses

to indicate the emphasis of speech even where it can, gives for Tetat

c'est moi ' the cumbrous rendering

' The state it is I

' instead of ' / am

the state,' not the same as '

I am the state' with which he confuses it.

CHAPTER III

TRANSLATION OF VERSE

Up till now we have been considering Translation

in its general aspects. We now consider it in relation

to special forms.

Notwithstanding some uncertainty as to the exact

lines of demarcation, the world of literature is still

parted into two great continents, Prose and Verse, and

our cardinal principle would seem to require that

prose should be translated by prose and verse,

if possible, by verse. On the first half of this proposition

there is no controversy. About the second,

though at first sight equally self-evident, there has

been no little disagreement.

Verse in itself is a more powerful engine than prose ;

it has a further range and its impact is heavier. Hence

the sacrifice entailed by rendering verse into prose is

a very real one, and one which we are not surprised to

hear from Mr Archer the author of 'Peer Gynt' refused

to allow. His decision, which is that also of most

translators of modern poems and of many translators

of ancient ones, accords with the considered judgment

of the accomplished scholar who has translated the

Aeneid into both, that the metrical form of the

original is a feature which a translator is

bound to preserve. Long before Abraham Cowley,

in the Preface to his Pindarique Odes, pertinently

asked '

I would gladly know what applause our best

pieces of English Poesie could expect from a Frenchman

or Italian, if converted faithfully, and word for word

78 TRANSLATION

into French or Italian prose.' How far a verse, that is

a good verse translation of verse surpasses a

rendering in prose a few examples will show.

How poor appears the Loeb translation of Juvenal,

haud facile emergunt quorum uirtutibus obstat

res angusta domi,

It is no easy matter anywhere for a man to rise when poverty

stands in the way of his merits,

when set by Johnson's

Slow rises worth by poverty depress'd.

Wickham's of Horace, Satires II ii.

Cur eget indignus quisquam te diuite ?

Why is any in want who does not deserve it, while you have

wealth ?

by Pope's

How dar'st thou let one worthy man be poor ?

and the Loeb translation of Catullus,

odi et amo. quare id faciam, fortasse requiris.

nescio sed fieri sentio et excrucior.

I hate and love. Why I do so, perhaps you ask. I know not,

but I feel it, and I am in torment.

when compared with Mrs Krause's

I hate yet love. You ask how this can be.

I only know its truth and agony.

Why we say a good verse translation, the version of

Thomas Moore will show :

I love thee and hate thee, but if I can tell

The cause of my love and my hate, may I die !

I feel it alas ! I can feel it too well,

That I love thee and hate thee, but cannot tell why.

Interesting proof of this superiority of verse to prose

is supplied by the results of Metaphrase1

. In his

1 Above, p. i.

VERSE SUPERIOR TO PROSE 79

little book of 'Echoes from the Greek Anthology

(1919),' Mr J. G. Legge has played the part of metaphrast

to a number of the prose renderings of Dr

Mackail.

I cite the following:

NcumXe, p.T) irtvdov T'IVOS ev6d8( rvftjSos 08' ei/it

dXX' avros ITOVTOV rvy^ave xpqororepov.

A. P. VII 350.

Mariner, ask not whose tomb I am here, but be thine own

fortune a kinder sea. MACKAIL.

Seafarer, ask me not whose tomb I be,

But mayst thou chance upon a kinder sea !

LEGGE.

IloXXa XaXelf, avdpaire, XaMa ' &* T>i^H M*1"" ^<pov.

crt'ya >cai /xeXe'ra 5>v en rbv ddvarov. XI 300.

Thou talkest much, O man, and thou art laid in earth after

a little; keep silence, and while thou yet livest, meditate on

death. MACKAIL.

The grave is near, waste not in talk thy breath ;

Keep silence, man, and living think on death.

LEGGE.

Hov croi TQOV f<elvo iraXivrovov ol r OTTO creio

irTjyvvpfvoi p.e<raTr)v (s Kpadirjv Savants;

irov nrtpd; trov Xa/tTras 7ro\va>8vvos ; ts ri 8t rpiercra

erre/x/xara \tpv\v fxfis KPaT^ &' e>7r> aXXo 0ep{? ;

OVK airb nav8rifj.ovt (i>f, Kvirpidos 0118' dnb yair^s

c(/i( not v^airfs tuyovos ev(ppoo~uv7)s.

dXX' f'-yw fg nadaprjv Hfpoirutv (pptva Trvpvbv dvairrat

fiifjiadir)! ijrvxrjv 8' ovpavov elcravdyo)

tK 8' dpfT&v crrpdvovs TTHrvpcw ir\fKw &>v dtp? ticdcmjs

rov(r8e <pfp(>)i> irpa>T<a rw <ro<pir]s <rrf(po[juii.

App. Plan. 20 1.

Where is that backward-bent bow of thine and the reeds that

leap from thy hand and stick fast in mid heart ? Where are thy

wings? Where thy grievous torch? And why earnest thou

three crowns in thy hands and wearest another on thy head ?

I spring not from the common Cyprian, O stranger, I am not

born from earth, the offspring of sensual joy ; but I light the

8o TRANSLATION

torch of learning in pure human minds and lead the soul upwards

into heaven. And I twine crowns of the four virtues ;

whereof carrying these, one for each, I crown myself with the

first, the crown of Wisdom. MACKAIL.

Where is thy back-bent bow, those shafts of thine

That quiver in mid heart, so surely sped ?

Thy wings ? Pain-dealing torch ? And wherefore twine

Thy hands three wreaths, a fourth upon thy head ?

Stranger, no common Cyprian's child you see,

One born of earth, offspring of sensual joy :

In human minds I light, if pure they be,

Learning's bright torch and souls to heaven convoy.

Of the four virtues wreathed, in hand I bear

These three, the first, Wisdom, as crown I wear.

LEGGE.

Not only is verse as a representative of verse in

itself superior to prose, but in certain circumstances

its only tolerable representative. If the original is such

as is described by Horace, Satires I 4. 56 sqq. :

his ego quae nunc

olim quae scripsit Lucilius eripias si

tempora certa modosque et quod prius ordine uerbum est

posterius facias praeponens ultima primis,

non, ut si soluas ' Postquam Discordia taetra

belli ferratos postis portasque refregit'

inuenias etiam disiecti membra poetae,

then, as Horace says, remove the form of verse and you

strip offthe only thing which distinguishes it from prose.

It follows that acceptable translation of Comedies,

Satires and compositions where the true poetic quality

is either absent or present in slight degree, as in the

Civil War of Lucan, can only be achieved through

verse.

If now verse is the appropriate form to render verse,

why is prose ever used instead ? Why should the prose

VERSE IN HOMERIC TRANSLATION 81

translations in the Loeb Classical Library so much outnumber

the verse ? For more than one reason. Verse

to be acceptable, even to its composer, calls for more

skill than prose. Again, as a rule it takes more time to

produce. But given the skill and the leisure for translating

in verse, is there still any reason for preferring

prose? It is undoubtedly true that in prose

we may come nearer to the constructions and phrasing

of the original. And so, regarded as an aid to the

understanding of a text or as an abbreviated commentary

thereon, a prose translation has an advantage,

as W. G. Headlam has said, in the Preface to his translation

of Aeschylus (1904): 'Prose has a proper function

of its own, a separate and different one to show how

the Greek is to be construed. It is superior for that

purpose, and should be content, I think, if it can

achieve it without more offence than necessary.' He

adds ' The true spirit and effect are only in the power

of verse to give.' And our present concern is with

a substitute for an original, not with helps to its study.

In the case of Homer, where translation into

verse has peculiar difficulties to encounter, a

somewhat different argument is advanced for prose

in the Preface to Butcher and Lang's

'

Odyssey,'

p. vii: 'The epics are, in a way, and as far as manners

and institutions are concerned, historical documents.

Whoever regards them in this way, must wish to read

them exactly as they have reached us, without modern

ornament, with nothing added or omitted. He must

recognise, with Mr Matthew Arnold, that what he now

wants, namely the simple truth about the matter of

the poem, can only be given in prose,

' for in a verse

translation no original work is any longer recognisp.

6

82 TRANSLATION

able.'...Without this music of verse only a half truth

about Homer can be told, but then it is that half of

the truth which, at this moment, it seems most necessary

to tell. This is the half of the truth that the

translators who use verse cannot easily tell. They

must be adding to Homer....A prose translation cannot

give the movement and the fire of a successful

translation in verse ; it only gathers, as it were, the

crumbs which fall from the richer table, only tells the

story without the song. Yet to a prose translation

is permitted, perhaps, that close adherence to the

archaisms of the epic, which in verse become mere

oddities.'

There is nothing here to traverse but the admission

that translators in verse must be adding to Homer,

thus making the poet 'unrecognisable.' The necessity

for such additions as are instanced is, as we have seen,

for the most part self-imposed.

The contention we are now considering seems partly

true and partly false. It is true that verse has restrictions

of its own. This the Romans saw when they

distinguished verse and prose as metrically bound

(uinctus) and free (solutus). But there is an offset.

In vocabulary verse is freer and fuller than prose ;

and poetic liberty or, to use the time-honoured name,

'poetic licence' is not confined to words but extends

to order and constructions. This is specially important

in English where an inflexible syntax and

stereotyped order are serious impediments to the

facile writing of prose. It need not then surprise us

much that Messrs Archer say (op. cit. p. xix) :

' We

have found by experiment that the fact of writing in

measure has frequently enabled us to keep closer to

QUASI-METRICAL PROSE 83

the original than would have been possible in prose

'

;

or that Mr Leonard found in translating Lucretius

(above, p. 8 n.) that ' for accuracy of meaning verse was

preferable to prose.' Granted however that on the

whole the advantage is with prose, is the gain of

fidelity to the expression sufficient to outweigh the

infidelity to the form ?

There is however another course. Does a man feel

that he must translate, that he cannot versify, and

that to prose he is ashamed ? Then let him translate

in Prose so specially modified as to evoke the

Idea and Associations of Verse. This is not the

'

poetical prose

' of writers like Ruskin, whose diction

has a large infusion from poetry but whose rhythms

are the rhythms of prose, but prose so constructed as

to convey to the reader distinct suggestions of verse.

Such Quasi-metrical Prose will not, it is true,

communicate the impression of a poem ; but it will

lift the translation off the level of ordinary prose and

will secure for the translator the freedom of employing,

without offence or sense of disharmony, the vocabulary,

the constructions and the order of verse. This, I take

it, is what Professor Tucker means when he says,

Preface to his Choephori, p. v :

' For the purposes

of a work like this it appears imperative to render to

the best of one's ability in language which, though

is not appvdpos and though <ra^>7??, not ra-

It is the prose I have endeavoured to write

for the Loeb translation of Tibullus.

Since Verse, where possible, is to be rendered

by verse, what duty does our translator owe to

the Metre of his original? That he render it

by the metre with which it best corresponds.

62

84 TRANSLATION

Take the particular case of Epic verse. No one

doubts now that the hexameters of Greek Epic are

most fitly rendered by hexameters in Latin. But it

was not always so. When literature began at Rome,

the hexameter had not been evolved, and hence the

poet Liuius Andronicus translated the Odyssey into

Saturnians, the native measure of the Romans.

Saturnians are as unlike Homeric hexameters as

anything can be. But Liuius had no choice. The

Saturnian was then the only Latin metre with associations

suitable for his purpose, that is, the only

corresponding one 1

. This old Italian measure, for

Naeuius the vehicle of original as for Liuius of translated

Epic, was forced after a struggle short but sharp

to yield to the hexameter, which Ennius the father of

Roman poetry brought in from Greece and harnessed

to the service of the Latin Muse. The hexameter

displaced its rival, not only because it was intrinsically

superior but because it could be transplanted

without harm. What Ennius did for epic, others

did for elegiac and lyric metres, and with little less

success.

The reason of this success, the secret of the correspondence

of Greek and Latin metres, concerns our

present enquiry. The common basis of both

Greek and Latin Metre is quantity, that is, the

normal quantity of syllables. Whatever the basis

of English metre may be I will not here launch out

on so stormy a sea the normal quantity of syllables

1 On the literary standing of the Saturnian at this period see more in

Miss E. M. Steuart's paper on 'The Earliest Narrative Poetry of Rome'

in the Classical Quarterly for January 1921, pp. 31 sqq., which has

appeared since the above was written.

CLASSICAL QUANTITATIVE METRES 85

it assuredly is not. Hence, to speak particularly of

the Greek hexameter, it is difficult to find any feature

of it which cannot be reproduced in Latin, while in

English the trouble is to find anything that can.

Observe the correspondence between Latin and

Greek evinced by lines which can be transferred

almost bodily from the one speech to the other:

Et cycnea mele Phoebeaque daedala chordis.

LUCRETIUS 2. 505.

icai KVKVda pf^-ij $ot/3fta re Sai'SaXa

Amphion Dircaeus in Actaeo Aracyntho.

VERG. Buc. 2. 24.

i' AipKaTor ev 'Aicrat'ep

'

Orphei Calliopea Lino formosus Apollo, ib. 4. 57.

'Op(pet KaXXiOTreia Ai'i/cp KaXXioro? 'A.ir6X\o)v.

Glauco et Panopeae et Inoo Melicertae.

Georg. i. 437.

rXauKO) Km HavoTrcir) 18' 'li/awo Mf\tKfprr).

And then consider what S. T. Coleridge, no mean

craftsman in verse, has given (after Schiller)

1 as specimen

representatives of ancient classical metres :

The Homeric Hexameter Described and Exemplified* :

Strongly it bears us along on swelling and limitless billows,

Nothing before and nothing behind but the sky and the ocean.

The Ovidian Elegiac Metre Described and Exemplified3

:

In the Hexameter rises the fountain's silvery column,

In the Pentameter aye falling in melody back.

1 The '

reproductions' of Classical metres in Modern German have of

late been trenchantly handled by Professor Wilamowitz in his ' Griechische

Verskunst,' pp. 5 sqq.

2 I

3

!! But German can do worse, as witness this quoted by Wilamowitz,

op. cit.,

' Ware die Welt nicht die Welt, ware denn Rom auch nicht

Rom.'

86 TRANSLATION

Single English verses, conforming to Greek and

Latin laws of metre and prosody, have indeed been

written, though far fewer than many persons think.

But that classical quantitative measures can

be domesticated in our speech there is nothing

to show and a great deal to disprove. The

tyranny of stress in English, its very different syllabation

1

, its glut of consonants and the preponderance

of monosyllables in its poetic diction present a combination

of obstacles which the skill even of a Tennyson

finds it hard to surmount 2

. Anyone may convince

himself of this if he will take two or three lines of

English

' Classical '

Verse, replace the English words

by Latin or Greek ones of the same metrical value,

and consider the Latin or Greek lines that result.

We may reasonably ask from those who recommend

their 'classical metres' for English use that they

should be in some agreement about them. Every

genuine hexameter in Latin literature, from Ennius

to Ausonius, can be recognised for what it is by anyone

who knows the quantity of the Latin words that

compose it. But who could tell that Dr Bridges'

Like a lion nor alone, for with him two followers went,

Iliad 24. 573

and Dr Rouse's

Utter tumult stirred in the gulf and the rock rebellowed

(op. cit. p. 27)

1

Disputation on the merits and demerits of the English hexameter

appears seldom if ever to take this into account. Hence much futile

discussion as in the Classical Review of 1917. To a Roman the syllabation

of dis-in-her-it (for di-si-nhe-rit) would indeed have seemed strange

and perverse, but the metrical effect of the word with its four closed

syllables would have been . Compare the Times Literary Supplement

of June 16, 1921, p. 388.

2 His imitations are not free from metrical faults.

'CLASSICAL' METRES IN ENGLISH 87

were in the same metre unless both be labelled hexameter

?

When, if ever, agreement is forthcoming, when, if

ever, experiment and criticism have evolved a form

which experts can accept as something better than a

travesty of the classical measure, the translator of

classical verse will have a new instrument at his disposal.

Till then, to use the phrase of Matthew Arnold

himself (op. cit. p. 169),

' we must work with the tools

we have.' And these tools are the native metres of

English. For the present it suffices to say with Mr

Omond (op. cit. on p. 91) that in English 'the hexameter

still awaits development1

'; and what is true of the

hexameter is, a fortiori, true of the other measures,

such as the so-called English 'Sapphic,' in which

performance and promise are both insignificant

2

. And

I fear the same must be said of Robinson Ellis's

courageous attempts to translate Catullus in the metres

of the original (1871).

In iambic and trochaic verse the chasm between

ancient and modern does not gape so widely,

and in rhythm the English measures correspond fairly

to the classical. But pure iambics cannot be repro-

1 The purely

' accentual hexameter ' with its frank indifference to

syllabic quantity has the sole merit of ensuring that the first syllable of

each foot shall be long :

This is the forest primeval the murmuring pfnes and the hemlocks.

Its supporters now are few, and to those who have any feeling for the

ancient verse it is simply a monster.

2 The movement of the '

English Sapphic,' that of Canning's

' Needy

Knife-grinder' travesty, is that into which English readers of the

Horatian Sapphic who have not been better instructed inevitably fall,

although its falsity to its model has many times been exposed. The

Sapphic metre has been perverted in Germany also where it is given an

iambic movement : Wilamowitz, op. cit. p. 6.

88 TRANSLATION

duced in our tongue. Ellis's translation of Catullus

XXIX 5, 'cinaede Romule, haec uidebis et feres,' Can

you look on, look idly, filthy Romulus, would to a Roman

have given the effect of

| | |

^ -

|

^ -

\

^ -,

nor would he ever have conceded that The puny

pinnace yonder you, my friends, discern, was ' a pretty

exact representative of a pure iambic line.'

RHYME

Modern verse has, or at least has had, two

distinguishing marks, Metre and Rhyme. Of

these, Rhyme may be dispensed with, but, speaking

broadly, Metre cannot.

If the original is in rhyme, Correspondence

suggests that the translation should be in rhyme too.

A good reason, but not necessarily a decisive one.

Messrs Archer found that they could not render 'Peer

Gynt' adequately if rhyme were used, and so with the

assent of the author their translation is rhymeless, and

what appears to be the best rendering of the 'Divina

Commedia' into English verse abandoned both the

rhymes and the stanza of the original.

If the original is Unrhymed, the argument

from Correspondence fails; and rhyme, if used,

must be used for other reasons. They must be weighty

if they are to outweigh the disadvantages which its use

involves. Rhyme is always a handicap, and in our

language a very heavy one. Mr A. Loring, in his

Rhymer's Dictionary, says (p. xxx): ' Rhymes are so

limited in number in the English language that great

licence is accorded to the poet.' The rhymer's troubles

were known to Shakespeare, whose humorous Benedick,

89

after speaking of the 'even road of blank verse/

proceeds :

Marry, I cannot show it in rhyme. I have tried. I can find

no rhyme to '

lady

' but '

baby,' an innocent rhyme ; for ' scorn '

'horn' a hard rhyme, for 'school' 'fool' a babbling rhyme,

very villainous endings. No, I was not born under a rhyming

planet.

If rhyme hampers the composer, how much more the

translator. John Ruskin, in a letter to D. G. Rossetti

upon Cayley's translation of the ' Divina Commedia,'

says 1

:

'

I think Mr Cayley has failed simply by attempting

the impossible. No poem can be translated into

rhyme for the simple reason that in composition a poet

arranges his thoughts somewhat with respect to his

rhyme. The translator cannot do this, and therefore

must sacrifice all grace and flow to his rhyme and

sometimes truth also.' Ruskin's observation is acute,

though its expression is too absolute2

. But others

have been no less outspoken, as Cowper,

' a just

translation of any ancient poet in rhyme is impossible,'

and F. VV. Newman, 'the exigencies of rhyme

positively forbid faithfulness.' Do we really need proof

of these statements ? If so, might it not be enough to

quote two stanzas of what, under the title of 'The Two

Grenadiers,' a rhymester makes the English vocalist

associate with the verse of Heine and the music of

Schumann ?

How bitterly wept then the Grenadiers

At hearing the terrible story,

And one then said '

Alas, once more,

My wounds are bleeding andgory !

'

1 Ruskin's Works, Library Edition, vol. 36, p. 189.

2 A translator is not bound to render turns in the original which are

merely concessions to rhyme. See p. 34.

90 TRANSLATION

The other said ' My sun is set ;

With thee I would die gladly.

But I've a wife and child at home ;

Without me they fare badly}

Here is a pauper indeed. But none are rich enough

to pay the toll that rhyme exacts from its votaries. By

turning the limelight, so to say, on what is a mere

detail, it spoils for us the effect of Lord Curzon's

otherwise fine translation of a stanza of Francois

Coppe*e, Ruines du Cceur, op. cit, p. 92 :

Mon coeur e*tait jadis comme palais romain,

Tout construit de granits choisis, de marbres rares.

Bientot les passions comme un flot des barbares

L'envahirent, la hache ou la torche d la main.

Long ago my heart was like a Roman palace,

Made of choice granites, decked with marbles rare ;

Soon came the passions, like a horde of vandals,

Came and invaded it, with axe and torch aglare.

The question of the Suitability of Rhyme

must be considered apart from its difficulty. It has

been held by some writers that in rendering ancient

verse it should not be used at all. Mr T. S. Omond,

' Arnold and Homer,' p. 83, says :

' Rhyme adds to the

translator's task ; it is as characteristically un-Greek

as it is characteristically English and modern. Therefore,

if it can be discarded without loss, it obviously

should be discarded.' So Plumptre, Preface to his

translation of Sophocles (first edition i865),explaining

why he ' thought it right to exclude it altogether,'

says:

' It is not merely that it enhances the labour of

translating ; that might be overcome by greater diligence

or greater skill. A much more serious objection

is that it is hard to escape a sense of incongruity in

EFFECT OF RHYME 91

this union of what is essentially modern with what is

essentially ancient 1

.'

On the other hand, Miss Swanwick (op. cit. supra,

p. 3) strongly advocated the use of rhyme to represent

the lyrical element in ancient poetry, thinking

the objection that 'the exigencies of rhyme forbid

faithfulness' to be overstated, and Sir T. H. Warren,

op. cit. p. 127, says:

' Rhyme...belongs to what is

called the genius of the English language. It comes

under the head of idiom and equivalent and, subject

to the consideration stated above, should be used as

such.'

Practice also diverges as much as theory. Dr Way

in his Euripides rhymes the choruses but not the

dialogue ; Professor Murray both, and Professor

Wilamowitz, in his Hippolytos, neither. Further

inquiry then seems desirable.

The effect of Rhyme is to bind and unify. It

makes one whole of the verses covered by the rhyming

words, giving us in place of single lines groups of

two, three or more. It follows then that in continuous

measures or successions of similar lines (in Greek

Kara o-rt^oi/) rhyme is an unsuitable adjunct. So

Matthew Arnold (whom Mr Omond follows, op. cit.

p. 83) :

' There is a deeper, a substantial objection to

rhyme in a translation of Homer. It is that rhyme

inevitably tends to pair lines which in the original are

independent, and thus the movement of the poem is

changed.' The rejection of rhyme for the rendering

of continuous verse disposes of the heroic couplet of

Dryden and Pope as well as of the ballad measure

1 In a subsequent edition (1880) he deferred to opinion so far as to

add a rhymed version of the choral odes in an appendix.

92 TRANSLATION

of Conington's Aeneid and the Spenserian stanza of

Worsley's Odyssey.

It has been vainly urged in arrest of this judgment

that the constriction and monotony of the couplet

groups may be mitigated by various devices, for example

by interspersing occasional triple rhymes as in Dryden

and Pope or carrying on the sentences from one

couplet to the next as is done by William Morris.

For why should we introduce at all a feature whose

proper function we at once set about to obscure ?

The only form remaining available is the Blank

Verse of Milton's 'Paradise Lost' and of Gary's

translation of Dante's ' Divina Commedia ' ' the recognised

English vehicle of the epic,' Warren p. 123.

The same measure is obviously proper for whatever

is written in Continuous Hexameters. But for special

reasons rhyme is sometimes preferred. Epigram and

Antithesis run naturally in pairs ; and it has been

thought that, for example, in Lucan the rhymed couplet

of Nicholas Rowe is a more suitable measure than

the blank verse of Sir E. Ridley. It should however

be observed on the one hand that it is no part of a

translator's duty to cast about for a more appropriate

form than the original, and, as we may gather from

Martial (VI 65, i), the Hexameter was not generally

regarded as a suitable vehicle for epigram, and on the

other hand that the ' Night Thoughts

' of Dr Edward

Young are sufficient proof that blank verse and epigram

are perfectly compatible.

More however may be said for rhyming in Amoebean

verse, as in Lord Bowen's translation of Vergil's

Bucolics and in Mr J. L. Scott's Eclogues of Calpurnius,

in whose rendering I, 28 32

RHYMED OR BLANK VERSE? 93

Ornytus. Non pastor, non haec triuiali more uiator,

sed deus ipse canit ; nihil armentale resultat

nee montana sacros distinguunt iubila uersus.

Corydon. Mira refers ; sed rumpe moras oculoque sequaci

quam primum nobis diuinum perlege carmen,

appear as follows :

Ornytus. No traveller, no shepherd here

His wayside leisure seeks to cheer.

The style a very God betrays ;

No ring here of bucolic lays,

Nor alpine jodels intersperse

Their pauses through the sacred verse.

Corydon. Your words are strange ; yet prithee waste

No time, but read me o'er in haste

This song divine, and let your eye

Your rapid tongue accompany.

English poverty in metres does not permit

differentiation between the Hexameter and the Iambic

of drama. And accordingly the ten-syllabled line of

Engolish blank verse has been used for both. With

some translators however Dryden's example has proved

more potent than Shakespeare's. Of Dryden's tragedies

all but one have the dialogues rhymed ; and this choice

is defended in his 'Essay of Dramatic Poesy' pp. 90 sq.

(ed. Ker). His last one ' All for Love '

is in blank

verse, 'not,' its author says, 'that I condemn my former

way but that this is more proper for my present purpose.'

Professor Gilbert Murray's choice of rhyme for

his translation of Euripides is defended in the third

volume of the ' Essays and Studies of the English

Association' by arguments more subtle than convincing.

' Rhyme ' he says

'

gives to the verses the formal

and ringing quality, remote from prose, which seems

to my ear to be needed ; it enables one to move

swiftly like the Greek and to write often in couplets

94 TRANSLATION

and antitheses like the Greek '

(p. 22). Again :

' Blank

verse, having very little material ornament, has to rely

for its effect on rich and elaborate language. Rhyme

often enables you to write lines as plain and direct as

prose without violating the poetical atmosphere.' The

first reason here is just a personal preference. The

second was Matthew Arnold's reason for saying that

Pope's rhymed couplets were more Homeric than

Cowper's blank verse, because they were more '

rapid.'

Yet it is hard to see why rhymed verses must be

swifter than rhymeless. All depends upon the handling.

Sir George Young commends the blank verse of 'Julius

Caesar' for its rapidity and J.W.Donaldson's excellent,

if little known, translation of the Antigone is not

lacking in speed. It is true that rhyme enables one

to write in couplets and antitheses. And, where this

is appropriate, it is an advantage. Thus Mr B. Drake

in his edition of Aeschylus Eumenides (1853) used

rhyme for the couplets 68 1 sqq. in which Apollo and

the Eumenides spar during the counting of the votes 1

.

But it is no advantage to be perpetually forced into

couplets or tempted into antitheses.

In translating ancient drama the songs

must be di fferentiated from the dialogue. But

if the dialogue is rhymed, how can this be secured 2

?

Outside Tragedy there is still less reason for

rhyming dialogue. Even Dryden says that blank

1 The rhyming should have stopped at 700. It is out of place in the

speech of the Eumenides 701 3.

1 Mr Platt, not using verse, in his free translation of the Agamemnon

of Aeschylus into Biblical prose, discriminates by treating the dialogue

like the narrative and the choruses like the lyrics of the Old Testament,

'a distinction to the eye and ear... with which we are acquainted in every

rationally printed Bible.'

DIFFERENTIATING USE OF RHYME 95

verse is

' most fit for comedies where I acknowledge

rhyme would be improper.' Nor can I guess why

Dr Way, disregarding Shelley's example and his own

practice elsewhere, imported rhyme into the Cyclops.

For satyric drama is tragedy burlesqued ; and as such

it must follow the tragic form.

We pass now to consider for what ends and in what

circumstances Rhyme may or should be employed.

First Rhyme may be used as a second means

of differentiating Poetry from Prose, or in other

words for distinguishing grades of poetical quality.

Hence many translators of ancient drama who use

unrhymed verse in the dialogue change to rhymed

in the choruses. So Miss Swanwick, and Dr Way in

tragedy and Dr Rogers in comedy, and numerous

translators whose work is included in Mr A. W.

Pollard's ' Odes from the Greek Dramatists.' Fitz-

Gerald and Browning agree in this, though they agree

in nothing else. But rhyme does not have it all its

own way. It is discarded for example in the translations

of Sophocles by Donaldson, Plumptre and

Whitelaw.

A decision is not easy. But Sir T. H. Warren, op. cit.

p. 128, pertinently observes: 'In the lyric, rhyme seems

almost necessary to counterbalance the loss involved

in forsaking the form of the original Latin and Greek.'

And since the sharp distinction between the metres

of dialogue and chorus cannot be reproduced in

English, rhyme appears to be admissible as a substitute,

notwithstanding the sacrifices of fidelity in

details that its admission will entail.

The differences noted in translators' attitudes and

96 TRANSLATION

behaviours towards rhyme are somewhat surprising

till one reflects that with many of them the determining

factor is not so much its appropriateness as

its attractiveness. Hence Sir T. H. Warren goes so

far as to say that it is part of the genius of the English

language. If its spell falls on a translator who, like

Professor Gilbert Murray, was born under a rhyming

planet, all can see what will happen. The translator

likes the rhymings, the reader likes them, and who

can prove that the author would not have liked them

too? But those who are less gifted by nature will

shrink from taking upon themselves the chains of

rhyme for all their glitter and clang.

Another use of rhyme is to reproduce the

Grouping of verses as in Couplets or Stanzas,

or again as in Strophes and Antistrophes. The

echo of the rhymed syllable in the rhyming is a satisfaction

to the hearer who has been waiting for the

correspondence. It conveys to him the impression

that something has been finished ; and he is ready

for a fresh start. This function of rhyme may be

seen in Shakespeare's practice of rounding off a scene

in blank verse with a final rhyming couplet Rhyme

then is usually and properly employed in translating

the Elegiac Distich.

Nor is it less appropriate in rendering the quatrains

of Horace and others. But here it is necessary to

qualify.

In the first place since a single rhyme, to the word

that ends the last line of the stanza, is sufficient to

unify and give the reader the feeling of a whole,

a second rhyme is a needless luxury, a burden

on the translator, certain to injure the translation. I

RHYME IN TRANSLATING HORACE 97

have given examples, in the article already cited, from

Mr Lathom's translation of the Odes of Horace, which,

like Conington's and Gladstone's, uses the double

rhyme. I will add one from Conington himself. For

Horace II xvii. i 4:

Cur me querellis exanimas tuis ?

nee dis amicumst nee mihi te prius

obire, Maecenas, mearum

grande decus columenque rerum,

Conington gives

Why rend my heart with that sad sigh

1

?

It cannot please the gods or me

That you, Maecenas, first should die,

My pillar of prosperity.

Sigh, in the first line, is naught as a translation of

querellis but is demanded by the rhyme as a fellow to

die. Throw the double rhyme away, and you may write:

Why with thy plaints thy friend unman ?

Nor heaven would will it nor would I

That thou, my fortune's stay and pride,

Maecenas, first should die.

This function of rhyme is distinct from two

others with which indeed it is often conjoined, the

sharpening of a contrast and the heightening

of an emphasis.

Upon this Sir A. T. Quiller-Couch has some

suggestive observations in his paper on 'The Horatian

Model in English Verse' op. cit. pp. 66 sqq. Commenting

on the construction of Marvell's Horatian stanza:

Nor called the gods in vulgar spite

To vindicate his helpless right ;

But bowed his comely head

Down as upon a bed,

1 A verse which Thomas Moore might have written.

P. 7

98 TRANSLATION

he says: 'In Marvell's stanza we do in sense and

sound get the Horatian falling close almost perfectly

suggested. Yes: but not quite perfectly, I think. For

why? Because the ear is all the while attending for

the rhyme 'head,' 'bed.' That is the nuisance with

rhyme : it can hardly help suggesting the epigram,

the clinch, the verse 'brought off' with a little note of

triumph....Your ear expects the correspondent, and

'you are not quite happy till you get it,'

' and he goes

on to explain clearly and convincingly that Milton's

deviations from the forms of the sonnet in Shakespeare

and Petrarch were designed for the express object of

muffling this insistent echo of rhyme. Further on p. 74

he tells us that 'the Horatian secret can only be

captured' in 'delicate metres divorced from rhyme.'

He names one, Collins' 'Ode to Evening': 'There,

if anywhere in English poetry, shall we find the secret

of Horace's 'falling close.'

'

There is something more to be said. Agreeing that

the tone of Marvell's poem is truly Horatian, I do not

feel the same about its structure. It is not the rhyming

that is here at fault but the rhymer. The impulsive

force of rhyme lessens as the distance between the

rhyming words increases. Marvell's close is a rhyme

after an interval of only five syllables while the words

that provide it have the least euphonious ending, a

short vowel followed immediately by an explosive

consonant 1

. In it I can hear nothing that suggests

a falling close, rather, if one must turn to simile, the

nailing of pairs of boards together, first a longer and

then a shorter pair.

1 This is what makes the ending of Coleridge's 'Ovidian' pentameter

so peculiarly hideous.

CONSISTENCY IN USE OF METRES 99

But compare the stanza given above on p. 97. There

the interval is thirteen syllables and the rhymed words

end in pure vowels. Will not this be an approach to

a 'falling close'? Or take the stanza of 'In Memoriam'

where the interval between the first and last rhymes

is twenty-three syllables. When it ends, has the rhyme

any force left in it to make a 'clinch' or indeed anything

at all except its unifying property? That from this

point of view this stanza is not unsuited to render the

Alcaics of Horace can be seen from an elegant translation

of Odes in xxix bythe late H. C. F. Mason,one of

the most brilliant composers among my former pupils

1

.

Our discussion then points to this conclusion:

Rhyme may, we can hardly say must, be used in

translating stanzas of classical poetry; but it should

be used as sparingly as possible. In a quatrain

there should be two rhyming lines only, and

these not the last and the one before it, but the last

and the second, or perhaps better still the last

and the first.

When a translator has selected a measure he

is boun d to adhere to it. He must take his choice

' for better or worse." This was the view and practice

of Coning ton in his translation of Horace, and it

has been adopted amongst others by Mr Cudworth,

who says p. xi :

' Each translated ode must conform in

general appearance, division into strophes and length

and number of verses to its prototype, and each

instance of any given Horatian meter must invariably

be rendered into its English analogue as selected by

the translator.'

1 Compositions and Translations by the late II. C. F. Mason, Cambridge

Press, 1903, edited by H. H. West, with Memoir by R. C. Gilson,

p. 107.

72

ioo TRANSLATION

Conington's rule was challenged by Gladstone

whose arguments I have examined in the

Classical Quarterly in the article already cited, pp.

288 sq. Gladstone in his contention that 'the translator

from Horace should both claim and exercise the

largest possible freedom in varying his metres, so as

to adapt them in each case to the original with which

he has to deal' urged 'two fundamental objections'

against the rule. The first, which is based on the

principle of Commensurateness, was 'that the quantity

of matter which the poet has given in the same forms

of stanza is by no means uniform; and if uniformity

is to govern the translation, the space available for

conveying what has to be conveyed will sometimes be

too great, and sometimes too small.' This objection

cuts too deep. For the inequalities are not penned up

in different odes. They are found in different parts

of the same ode, and of course no one dreams that the

stanzas in the same ode should be of different lengths.

Gladstone again will not have it that 'any one English

metre which the translator may have chosen for one

Horatian ode will be equally supple, and equally

effective, for conveying the spirit and effect of every

other ode which Horace may have found it practicable

to construct under the same metrical conditions.' But

I have pointed out (1. c.) that it is hardly true to

suggest that, apart from rhyme, the English metres

are less supple than the Latin metres which they are

employed to render, that in his emancipation from the

restrictions of quantity and disregard of the concurrence

of vowels (hiatus) the verse-writer in modern

English has a freedom unknown to the Roman, and

that it is not the case that our English metres have so

marked an individuality as to unfit them for the various

THE IDEAL TRANSLATOR 101

uses to which the Latin metres are turned. I did not

question that a translator might obtain for himself

some occasional ease and relief by shifting from one

metre to another, and that, if his design were only to

render single odes, such variation might perhaps be

deemed excusable. But the translator of the whole

collection can claim no such liberty. It is not the least

characteristic feature of Horace's lyrical compositions

that the same metre is employed for odes of a

very different spirit. This sameness in diversity

is of the essence of his art, and to obliterate it in translation

is an infidelity of the highest order, to be condemned

the more unflinchingly because it is likely to

escape the reader for whom the translation is intended.

Non cuiuis homini contingit adire Corinthum ; and

this is true above all of those whose traffic is between

foreign and native speech. The ideal translator must

be a master of both the languages with which he has

to deal. His mastery of the foreign tongue must be

critical ; of his own practical. The non-coincidences

of language will tax his skill and care to the utmost.

If he would avoid ambiguity and misrepresentation,

he must be continually on his guard against the

erroneous or extraneous suggestions in '

equivalents.'

These he must so hedge in and circumscribe that

in all partially corresponding expressions the correspondences

alone shall be brought before the reader.

In a word he must be an Expert Qualifier.

If he elects to render Verse by Verse, he must have a

sufficiency of versifying skill and a special command of

the vocabulary of poetry. Poetic feeling he must possess.

But need he be a poet ? Dryden indeed demanded

as much. 'To be a thorough translator of poetry a man

102 TRANSLATION

must be a thorough poet.' But there is force in what

Lewis Campbell says, Preface to his translation of

Sophocles:

' Mr Arnold in speaking of the drawbacks

alludes to the danger of too much originality. In that

respect... it may be said... that this coy Muse yields

more readily to one who is not a poet ; for a poet cannot

step off his own shadow.' No writer can write away

from himself for long ; poets least of all. What havoc

one poet's individuality may make of another's, Moore's

version of Catullus p. 78 above is enough to show.

Moore failed through lack of perception ; but other

poet translators through want of self-control.

Linguistic knowledge and literary capacity are

essential for every kind of translation ; but Insight is

just as indispensable. Some would construe '

insight

'

as 'sympathy' with the author, or they would add

'sympathy' thereto. If this sympathy means the

appreciation which insight and study have produced,

we may agree. But many who use the word intend

something more emotional, such as admiration, love

or esteem. Aversion, to be sure, is likely to mar a

translator's work. But so, though more subtly, will

partiality, as may be seen from a study of 'sympathetic

' translations. Translating, we must add, is an

exercise of the intellect ; and sentiment has no place

in its performance.

Lastly our translator should have diligence and

conscientiousness in the highest degree. He should

shrink from no labour that may improve his work.

An infinite capacity for taking pains must be his

substitute for genius.

TRANSLATIONS

RETROSPECTIVE TRANSLATIONS

* HORACE, Odes I xxiii

VlTAS inuleo me similis, Chloe,

quaerenti pauidam montibus auiis

matrem non sine uano

aurarum et siluae metu.

nam, seu mobilibus uepris inhorruit

ad uentos foliis seu uirides rubum

dimouere lacertae,

et corde et genibus tremit.

atqui non ego te, tigris ut aspera

Gaetulusue leo, frangere persequor.

tandem desine matrem

tempestiua sequi uiro.

HORACE, Odes II xiv

EHEV, fugaces, Postume, Postume,

labuntur anni ; nee pietas moram

rugis et instanti senectae

adferet indomitaeque morti.

non, si trecenis quotquot eunt dies,

amice, places inlacrimabilem

Plutona tauris qui ter amplum

Geryonen Tityonque tristi

compescit unda, scilicet omnibus

quicumque terrae munere uescimur

enauiganda, siue reges

siue inopes erimus coloni.

frustra cruento marte carebimus

fractisque rauci fluctibus Hadriae;

frustra per autumnos nocentem

corporibus metuemus Austrum.

RETROSPECTIVE TRANSLATIONS

To Chloe

THOU shun'st me, Chloe, like a fawn

That seeks on desolate glen and hill

Her fearful dam; whom breeze and woods

With idle terrors thrill.

Its leaves aquiver to the wind

Has bramble rustled ? Thro' the brake

Has the green lizard push'd a way ?

Heart, knees are all aquake.

Yet no Gaetulian lion I

Or furious tiger seek thy gore.

Then, maiden for a husband ripe,

Follow thy dam no more.

To Postumus

FRIEND, friend, the years are fleeting fast,

Ah me ! nor will devotion stay

The nearing march of wrinkled age

And death that who can slay ?

Nay not, if every passing morn

Smoke of three hundred bulls arise

To tearless Dis, who Tityos pens

And Geryon's triple size

'Twixt dismal waters all must cross,

Who from earth's bounty draw our food,

Whether poor husbandmen we be

Or princes of the blood.

In vain we 'scape th' ensanguined field

And Hadria's hoarsely breaking surge,

Shrink from the South wind's sickly breath,

Of autumn hours the scourge.

ic6 RETROSPECTIVE TRANSLATIONS

uisendus ater flumine languido

Cocytos errans et Danai genus

infame damnatusque longi

Sisyphus Aeolides laboris.

linquenda tellus et domus et placens

uxor, neque harum quas colis arborum

te praeter inuisas cupressos

ulla breuem dominum sequetur.

absumet heres Caecuba dignior

seruata centum clauibus et mero

tinguet pauimentum superbo

pontificum potiore cenis.

3 HORACE, Odes n xv

lAM pauca aratro iugera regiae

moles relinquent, undique latius

extenta uisentur Lucrino

stagna lacu platanusque caelebs

euincet ulmos ; turn uiolaria et