TRANSLATION AND TRANSLATIONS

36 TRANSLATION

renderings is substantial, the translator is free no

longer. Here unless other considerations, such as the

necessities of Verse, imperatively forbid, he must select

the translation that is nearest to the original. Otherwise

he effaces something that he might have preserved

and sinks into an '

adapter.' He will not be moved by

the insinuation in Sir T. H. Warren's phrase, p. 106:

'A really good translation should be not so much

exact as faithful'; and he will share P. Cauer's opinion

without sharing his surprise that 'often it is just the

literal rendering that is also the most natural 1

.' So

even a grammatical difference must be preserved if it

is significant. On Aen. IV 625

' exoriare aliquis nostris

ex ossibus ultor' Professor Tolman correctly remarks:

1 how vivid the transfer from the third to the second

person' but in other respects his rendering: 'Arise some

avenger from our bones '

is too literal to be approved.

Not so that of Mr Rhoades, perhaps the best of our

translators of Vergil :

' O arise

| Unknown avenger

from my tomb.' I should prefer

' Rise from my dust,

Avenger!,' thinking 'dust' better than 'tomb' and that

aliquis is somewhat overtranslated by ' Unknown.'

Fidelity to the spirit, it is said, has the highest

claim on the translator. True ; but the spirit may be

lost if the letter is disregarded. Nothing is more

characteristic of Lucretius than the intensity of his

purpose that about his meaning there shall never be

mistake. Hence he doubles and redoubles an expression,

wholly fearless of tautology. I know no passage

where this is clearer than vi 557sqq., which describes

how earthquakes make buildings unstable, or where it

1 ' Manchmal ergibt es zur Uberraschung dass gerade die wortliche

Wiedergabe zugleich die natUrlichste ist' (op. cit. p. n).

ADVANTAGES OF LITERALITY 37

is more obscured by the lax renderings which the

assailants of accurate translation would approve.

I print the Latin text with a prose rendering, not

offered as a model translation but as a faithful reproduction

of the original, and a collation of the differences

in two standard prose translations, Munro's 1886

and Mr C. Bailey's 1910. I may add that if we were

in search of a lesson on the meaning of four Latin

prepositions pro ' forward,' in '

over,' ex ' up

' and re-

'

back,' no better one could be found.

Praeterea uentus cum per loca subcaua terrae

collectus parte ex una procumbit et urget

obnixus magnis speluncas uiribus altas,

incumbit tellus quo uenti prona premit uis. 560

turn supera terram quae sunt extructa domorum

ad caelumque magis quanto sunt edita quaeque,

inclinata minent in eandem prodita pattern

protractaeque trabes inpendent ire paratae.

et metuunt magni naturam credere mundi 565

exitiale aliquod tempus clademque manere,

cum uideant tantam terrarum incumbere molem !

quod nisi respirent uenti, uis nulla refrenet

res neque ab exitio possit reprehendere euntis.

nunc quia respirant alternis inque grauescunt 570

et quasi collecti redeunt ceduntque repulsi,

saepius hanc ob rem minitatur terra ruinas

quarn facit ; inclinatur enim retroque recellit

et recipit prolapsa suas in pondera sedes.

hac igitur ratione uacillant omnia tecta, 575

summa magis mediis, media imis, ima perhilum.

LUCRETIUS vi 557 576.

Furthermore when wind gathering from some one quarter

through hollow places below the earth presses forward and,

pushing with great force, bears hard on the deep caverns, the

earth leans over towards the side to which the head-forward

force of the wind is pressing it. Then the parts of houses which

38 TRANSLATION

are built up above the earth and each all the more as it rises uf

towards heaven, jut and lean over, setting forward to the same

quarter, and the rafters draggedforward hang over ready to go.

And (yet) men shrink from believing that a time of destruction

and disaster awaits the constitution of the great Universe, although

the earth's huge mass is leaning over in their sight.

Now should the winds not blow backwards, no force could hold

things back or pull them back from their passage to destruction.

As it is, since they do blow back by turns and their troops now

'

rally

' and come back and now are driven back and retire, for

this reason the earth threatens to fall more often than it falls.

It bends over and then sways back to the rear and the weights

that have slipped forward it takes back into their places. On

account of this therefore the whole of a building rocks, the top

more than the middle, the middle than the bottom, the bottom

but a very little.

Pro- 558 bears down Af., blows strong B. 560 headlong M.,

swooping B. 563 yielding M., tottering B. 564 wrenched

from their supports M., driven forward B. 574 after trembling

(falling B.} forward M., B.

ex- 561 all buildings M., houses that are built up B. 562

tower up M., raised B.

in- 560 leans over M., B. 593 lean over M., bend over B.

564 hang over M., hang out B. 567 hang ready to fall M.,

bowing to its fall B. 573 leans over M., B.

re- 568 (did not) abate their blowing M., breathed in again B.;

rein in M., put a curb on B. 569 hold up M., pull back B.

570 do abate M., breathe in B. 571 return to the charge M.,

charge again B., are defeated M., are driven back B. 573 sways

back M., swings back again B., recovers M., B.

The consideration of these details can never be

declined by the translator. To translate, I do not say

well but even passably, he must indeed first absorb

the spirit of his original ; but, this achieved, he must

not flatter himself that he is at once inspired, holding

a high commission to deal with the letter as choice or

mood may dictate. His is not the scene-painter's art,

ADVANTAGES OF LITERALITY 39

content with distant resemblance ; but the etcher's

which spares no touch or stroke that brings the copy

nearer to the exemplar. To him the whole is the sum

of its parts ; and he will agree with M. Maeterlinck in

his Translation of Macbeth :

' On aurait tort de croire

qu'il s'agit la de details infiniment petits. Ce sont ces

infiniment petits qui reconstituent,dans une traduction,

I'atmosphere de 1'original

1

.' Drudge's work this, it may

be said ; but not to a translator whose heart is in his

work.

(I append an example which speaks for itself from a

translation of an Italian song whose authorship is unknown

to me:

Voi che sapete You who have knowledge

che cosa e amor, what love may be,

Donne, vedete Ladies, pray tell me,

s' io 1' ho nel cor. dwells it in me ?

Quello ch' io provo Let me discover

vi ridir6, mypassion intense,

e per me nuovo strange visitation

capir nol so. that baffles my sense.

Sento un affetto With a strange longing

pien di desir my heart is fain,

ch' ora e diletto, now 'tis all pleasure,

ch' ora e martir. now 'tis all pain.

Gelo, e poi sento First I am freezing,

1' alma avvampar, then my soul burns,

e in un momento and the next instant,

torno a gelar. to ice it turns.

It seems somewhat strange that the same translator

who has given us a close and sufficient rendering of

stanzas I, 3 and 4, should in stanza 2 have slipped into

something not merely unfaithful to the original but out

of keeping with the rest of his work.)

1 Ritchie and Moore, op. cit., p. 17.

There is no more glaring fault in English writing than

its loose and arbitrary employment of 'synonyms'

and its almost morbid repugnance ravra Xeyeiv irepl

TWV avT&v, due in part no doubt to certain deficiencies,

as for example in pronouns, but chiefly to a sensuous

dislike of recurrences of the same word or rather of

the same sounds. As Latin and Greek are almost untouched

by this vice, the English translator of the

classics needs more than any other the cautions of

Professor Tolman :

' As a general rule the same English

word should be used to represent the same foreign

word '

(p. 45).

' When a foreign writer repeatedly uses

the same word, the translator has no right to attempt

the so-called refinement of his style by seeking to

avoid repetition. The superb diction of Matthew Arnold

is a standing contradiction to the old theory that the

same word or phrase must not recur in too close

connexion '

(p. 47).

The danger of this habit to philosophical or scientific

reasoning one example will suffice to show. In Cicero,

in the Tusculan Disputations, I 9, is a clear and simple

piece of reasoning on the question whether the dead

are miserable. The argument is futile unless the same

word is used to render miser throughout. Yet, though

I have seen scores of translations of the extract by

students, I can recall none in which it was not varied.

The difficulties of Translation increase when we pass

from structure and idiom to Vocabulary. All translation

presupposes the existence of similarities of thought

and of correspondences in expression between the

languages concerned. Without these the translator's

task is impossible, as indeed von Humboldt roundly

DEGREES OF CORRESPONDENCE 41

declared it to be 1

: 'All translation seems to me to be

simply an attempt to solve an insoluble problem.' And

with them, even under the most favourable conditions,

he will be perpetually vexed by Non-correspondences

or Imperfect Correspondences on the one hand and

Illusory Correspondences on the other. To illustrate

the correspondences, the non-correspondences and

the partial correspondences in speech, the figures of

Schopenhauer ('Parerga und Paralipomena,' Kap. 25),

adopted by Cauer and Tolman, may conveniently be

employed. If words be represented by circles, then

words of exactly the same meaning or contents may

be represented by circles of the same superficies, which

may be concentrically superposed above each other,

those of meaning partly the same by intersecting

circles, while those whose meanings have nothing in

common may be figured as circles that lie wholly outside

each other.

1 In a letter to A. W. v. Schlegel, the translator of Shakespeare,

July 23, 1796. And no less strongly Moritz Haupt: 'Translation is the

death of Understanding.'

42 TRANSLATION

In a few cases, as in scientific and technical terms

(above p. 30), the circles will coincide completely (A),

and there will be exact 'equivalents' to gladden

the translator's heart. Thus with the English

'

triangle

'

and the German 'Dreieck.'

Many times there will be No Correspondence

(B). Thus English has nothing corresponding to the

Latin augur nor Latin to the English premier. Such

words, whatever their grammatical category, are

strictly untranslatable. Translators must either take

over the Foreign word or provide a representative

by some Coinage of their own. Languages

have used both these methods in their borrowings,

most preferring to adopt, but German to re-coin. So

tricycle French and English from the Greek, Dreirad

in German.

Certain other devices, Periphrasis or Explanation

and Substitution of an Imperfect Correspondent,

whether of a General Term for a specific or of

one which is in some degree analogous to the original,

are confessedly mere approximations, and can be

employed but occasionally. Conington touches on two

of these in the Preface to his translation of Horace,

1870, pp. xviii sq. In dealing with 'local and temporary

allusions, proverbs, etc., a translator,' he says,

' has

three courses open to him, to translate more or less

verbally, so as to run the risk of being unintelligible

to a reader unacquainted with the original, to generalise

what is special, and to borrow something of the

imitator's licence, introducing a modern speciality in

place of an ancient. The last of the three methods

no doubt requires to be used very sparingly indeed,

or our great object of translating a classic will be

NON-CORRESPONDENCE 43

wantonly sacrificed.' Of his two methods, Generalisation

is feasible only where the specific reference is

so faint or so unimportant that it may be dropped

without much loss, as not seldom in Latin names for

the winds, e.g. the 'Eurus' of Horace, Odes II 16. 24

(no. 4). On Modernisations he has himself passed

judgment. Such arbitrary alterations of an original

are not tolerated in translations of contemporary

writing, and are still more objectionable when the

original is both foreign and ancient. Professor G. G.

Ramsay, in the Preface to his Histories of Tacitus,

pp. xiii sq., rightly protests against rendering adulter

by 'co-respondent' and diuus Augustus by the 'sainted

Augustus.' Propertius had written of his Cynthia that

such was the charm of her talk that it would make

the King of Heaven himself her slave, 'ilia suis uerbis

cogat amare louem,' I xiii 32, for which a modernising

translation (Mr Tremenheere's) gives 'She'd coax the

devil to her feet,' thus stripping the stately diction of

all its dignity and dragging the arch enemy of the

Christian into the imagery of a pagan poet who wrote

many years before Christ was born. The /e/aea icriXov

'A</>/>oSiVa9 of Pindar, Pythians, 2. 17, is not easy to

render, but 'domestic chaplain,' despite its claim to correspondence

in fact, and W. G. Headlam's 'Aphrodite's

pallid monk' must be rejected, as false in spirit to the

original.

In a few cases a non-corresponding term may

be used as a pis aller if a reader is warned by

inverted commas, italics or some similar device

that it is not to be understood in its accustomed

sense.

The use of Periphrasis or any other form of

44 TRANSLATION

explanation is defended on the ground that a translation

ought to be intelligible in itself. But to procure

this intelligibility reader and translator must cooperate.

Some equipment the reader must be presumed to

have translations are not for the complete ignoramus

and some effort he must make to understand.

The translator, on his side, is bound to provide all

necessary explanations, but these in notes, appendices

or indices, and not in the text. To Horace, Bassareus

or Thyoneus were by no means interchangeable with

Bacchus ; nor should they be replaced by it unless dire

necessity compel. In Hesiod turn a^oo-reo? 'Boneless'

(Op. 524) and (frepeoiicos 'House-carrier' (571) into

'cuttlefish' and 'snail,' and all the flavour of the diction

is gone. Translate sonipes 'Clatter-hoof (Phaedrus IV

4. 3) and barbatus 'Beardy' (iv 9. 10) by 'horse' and

'he-goat,' and all whiff of fable is lost.

In the vast majority of apparent correspondences

the circles will only partially coincide

(C) or the contents of the words will be partly the

same and partly different. Thus patruus and ' uncle '

have in common the notion of '

parent's brother.'

But ' uncle '

includes, whereas patruus excludes, the

maternal kinsman. Even where the central or principal

meaning of two words is the same, the accessory senses

or associations are often so different that the one word

is many times no fit representative of the other. Ass

and asinus denote the same animal ; but while the

dominant connotation in the English word is 'stupidity,'

the Latin word may merely connote '

insensibility

' or

'slowness.' Hence in Plautus, Pseudolus, 136, 'neque

ego homines mag-is asinos umquam uidi : ita plagis

costae callent,' we must see our students do not transPARTIAL

CORRESPONDENCE 45

late 'greater asses' but 'more like asses.' In Terence,

Eunuchus, 598,

' flabellulum tenere te asinum tantumj

the sense is not ' what an ass you must have looked '

(the Loeb translation) but ' to think of a big clumsy

creature like you holding a bit of a fan.' Senex is undoubtedly

' an old man,' as anus is

' an old woman.'

But, as I have shown elsewhere 1 for amis, such a translation

may arouse inappropriate associations, and at

Vergil, Georg. II 13454. 'animas et olentia Medi |

ora

fouent illo et senibus medicantur anhelis,' Dr Mackail

and Mr J. Jackson in prose translations give 'cure the

pantings of old age/ 'heals the gaspings of age.' I

expect Dr Way will some time regret that external

likeness beguiled him into rendering 0v\atcoi, Euripides,

Cycl. 1 82, by the vulgar English

'

bags

'

; but I do not

think that any translator on this side of the Atlantic

needs Professor Tolman's caution (op. cit. p. 46) that

tfaXepo? 7^09, 6a\epov Sdicpv should not be englished

by '

blooming marriage

' or '

blooming tear.'

In dealing with Connotations, the adherent as distinguished

from the inherent significations of words,

the translator has a delicate task before him. Into the

tissue of impressions which articulate speech excites

in its hearer or written in its reader, enter strands from

the senses and the emotions which, in justice to his

original, no translator may ignore. His duty here is a

negative one. He is not bound to provide instruction

or allurement ; but he is bound to avoid offence.

Diction unsuitable in any language for a particular

kind of composition is unsuitable also to translate

a similar composition in another. With warrant from

1 ' Flaws in Classical Research,' Proceedings of the British Academy,

Oct. 1908, vol. 3, pp. 164 sq.

46 TRANSLATION

the author indeed, as we have seen, and in the due

proportion, a translation may be coarse, bizarre, or uncouth

; but never otherwise. Where direct expression

is required, a translator has no option. The English

'snout' and 'hen/ as Miss I. M. Pagan ('Peer Gynt,' p. 8)

remarks, hideous monosyllables as they are, are all

that he can give for the musical disyllables of the

Norwegian. But in metaphorical expression there is

more freedom. Literally, no doubt, spuere is 'spit'

and uomere 'vomit.' But translators of Vergil avoid

'vomit' at Aen. IX 349, 'purpuream uomit ille animam,'

and at Lucretius II 1041 Mr Bailey uses the Biblical

' spew

' to render '

expuere ex animo rationem,' which

Munro had paraphrased by 'reject with loathing';

and it is significant that in criticising the line of

Furius Bibaculus which Horace S. II 5. 41 has parodied,

Quintilian (vill 6. 17) does not censure its phrasing as

'

ugly or offensive but only as 'harsh' or 'far-fetched':

sunt et durae, id est a longinqua similitudine ductae ut

'capitis niues' et '

luppiter hibernas cana niue conspuit

Alpes,'

'

'sputters his hoar snows o'er the wintry Alps.'

The translator is still further embarrassed

when aesthetic considerations are entangled

with ethical, as in matters where euphemism and

expurgation are not infrequently employed. Their inconsistent

treatment in the Loeb Library shows the

difficulty of the problem. Let it be enough here to

note that English still observes more reticence than

French or Italian; and that in its handling of the same

subject Latin is coarser than Greek, as may be seen by

comparing Martial with Aristophanes and Apuleius

with Lucian.

But translators are not merely balked by the nonILLUSORY

CORRESPONDENCE 47

correspondences and harassed by the partial correspondences

of languages ; for if not most wary, they

will be the sport of Illusory correspondences. The

fallacy that the expressions of two different languages

must correspond is most harmful when the sounds or

the letters of the words are alike or when there is a

superficial resemblance in the composition of phrases.

Most potent is it when the words have, or are believed

to have, a common origin. For it then unites two of

the most persistent delusions of mankind, that things

must be what they seem to be, and that change is not

at work in human affairs. I take the following illustration

from an unpublished popular lecture on Translation.

"Lurid and luridus (D) are two words whose

superficies coincide to the smallest possible extent.

For both are colour names ; and this is all. Let us now

see how that curious complex, the human ego, may be

expected to deal with them. The Eye reports to the

Brain,

'

I have seen the letters with which this pair of

words is written and printed. They are the same

lurid.' The Ear confirms. ' I have heard the sounds of

the words pronounced. In all essentials they are the

same lurid: luridus.' The Memory says, 'When

I was at school, or when I last consulted an etymological

dictionary, I noted that the English word was

the Latin word, borrowed.' Now comes the reasoning

Brain, or whatever you like to call the controller of

our mental processes. It says,

' These two words are

composed of the same letters so the Eye instructs

me and of the same sounds so the Ear informs me.

The resemblance, I see, is obvious, and things are what

they seem to be. Further, as the Memory assures me,

48 TRANSLATION

the English word and the Latin word were once the

same and things once the same are always the same.

Accordingly, using all the lights of my nature and

education, I pronounce luridus and hirid to be the

same ; and being, as I am, more or less unacquainted

with Latin but acquainted more or less with English,

I slide the circle of English over the circle of Latin;

and, see, my task is done !

'

Yes, done Mr Brain. And

this is why so many of your victims when they are

learning Latin are tempted to render lurida sulpura

as 'lurid sulphur' and luridi denies 1 as 'lurid teeth,'

instead of '

yellow brimstone ' and '

yellow teeth.' Yet

the facts, look you, are simple and plain.

' Yellow '

is

the English for luridus ; for ' lurid ' there is no single

word in Latin. And for the ascertainment of the

meaning of any word in any language its outward

form is of no moment whatever. For example, had

Fortune so decreed, the idea of yellow might have been

expressed in Latin by bonus. Then no one would have

dreamed that the Latin for 'yellow' meant 'lurid,'

but many would have toyed with the fancy that the

Romans thought a 'yellow face' to be a '&?#nie' one."

Messrs Ritchie and Moore (p. 6) have told us how

'on one occasion the mild expression demander une

explication, used in a French diplomatic note, gave dire

offence to the Government of the United States because

it looked like 'to demand an explanation,'

' while the

official English Translation of the Allied Note answering

Germany's first offer of peace in January, 1917,

renders prttendu as '

pretended,' where it clearly means,

1 Horace, Odes IV 13. 10. I leave it to the curious to consider why

C. S. Calverley has rendered this 'teeth of ghastly blue.'

1 ' Flew '

(Hor.

transuolat) demanded its rhyme, but why 'blue ' rather than ' hue '?

ILLUSORY CORRESPONDENCES 49

as generally, 'alleged,' ib. p. 4. A few years before the

War a German contributed to the daily press a series

of articles, written on the whole in excellent English,

in the course of which he had occasion to observe that

Germany was a formidable enemy (Ger. Feind). What

he wrote and the newspaper published was that the

German was a very dangerous fiend. On July 9, 1915,

the following advertisement appeared in the Times:

'Jack E. G. If you are not in khaki by the 29th, I shall

cut you dead. Ethel M.' The Berlin correspondent of

the Cologne Gazette telegraphed the following version

to his journal: 'If you are not in khaki by the 29th,

I shall hack [i.e. 'cut'] you to death' ('hacke ich dich

zu Tode '). It is commonly reported that in the Great

War an Anglican prelate concluded an address to

French soldiers with the prayer

' Que Dieu vous blesse \

'

These errors are gross and palpable ; but where the

matter is subtler, even the elect are not immune. Two

skilled translators of Martial, Mr W. C. Ker (Loeb

Library) in prose, and Mr W. T. Webb in verse (Select

Epigrams from Martial 1879) have put upon III 61 :

Esse nihil dicis quicquid petis, improbe Cinna,

si nil, Cinna, petis, nil tibi, Cinna, nego,

an interpretation which may be represented fairly by

the following version:

To unconscionable Cinna.

Whate'er you ask, 'tis asking nothing, cry you.

If you ask nothing, nothing I deny you.

This seems to have point enough, 'If what you ask is

nothing, I will give you all you ask.' But it is not the

point of Martial which is this: 'If what you ask is

nothing, then I am denying you nothing,' i.e. 'you

cannot complain of a refusal.'

P. 4

50 TRANSLATION

It may be urged that in dealing with these Illusory

Correspondences we are concerned with the equipment

of a translator rather than the inherent difficulties of

his task. In strictness this is true. But we must not

blink the fact that a translator has to work with an

imperfect instrument, that with every one the pull of

the native speech is steady, strong, and, worst of all,

unseen. ftlq \

eX/cet 7rpo9 avrrjv rrjv itc/iaSa TJ?? <f>povrt'So?

Aristophanes Clouds 233.

Even originals are not safe, and writers on the

Theory of Translation succumb. The excellent Tytler

gives as an example of a '

perfect translation of a

colloquial phrase by a corresponding idiom Mihi isthic

neque seritur neque repitur* Plaut. 'That's no bread

and butter for me.' His phrase is from Epidicus 205

'mihi istic nee seritur nee metitur.' re(a)p-itur is a

'suggestion' of the translation imp.

Further proof may be gathered from what our most

reputable Latin dictionaries present under the headings,

exempli gratia, of appareo, desidero, discipulus, fatalis,

humilis, instrumentum, malignus, posstdo, purus and

tiena. For more on this subject and on the urgent need

of a new Latin-English dictionary, I may refer to

what is said in 'Dead Language and Dead Languages'

(Murray, 1910) pp. 19 sq.

1 and for observations on

modern misunderstandings of the structure and ordering

of ancient sentences to 'Flaws in Classical Research'

in the Proceedings of the British Academy vol. 3,

pp. 165 sqq.

In the grapple between the foreign language

and the native, the native is most times the

1 Compare also Professor A. Souter's remarks and examples 'Hints

on Translation from Latin into English,' p. 10.

TRANSLATORS' ENGLISH 51

stronger. But the converse also is found. Prof.

Ramsay in the Preface to his Annals of Tacitus

vol. II, p. x, speaks of the possibility of the sense of

what is clear and idiomatic in English being overborne

by a knowledge of the original, and Professor Souter,

op. cit. p. 6, after saying 'An English translator into

English must of necessity know his mother-tongue

better than any other; but he must be distinguished

from the majority of his fellow-countrymen by a special

knowledge of the language to be translated' continues

'It will be strange, however, if, in the process of

obtaining his special knowledge, his sense of his own

language is not blunted.' A real danger is here. In

'translator's English,' as it has been called, the original

is not indeed falsified ; but the rendering is over-literal

and servile and its results are clumsy and harsh. Such

translations excite an involuntary repugnance in the

reader, the chief cause of the disparagement of 'accurate'

translation and of the prejudice against 'literal'

translation as such.

That translation may be feasible there must

be a minimum of coincidences and similarities

between the Foreign and the Native languages.

The task is easier with Cognate languages;

it is easier too with languages in the Same Stage

of Development. Further, the native character and

genius of a language may make it more or less suited

for the task of translation. For this work two

qualities are requisite, Copiousness and Flexibility,

the latter covering the colourlessness desiderated

by Professor Gildersleeve below, ^63. English

has the first in a very high degree. For the extent and

variety of its vocabulary it is comparable to Greek.

42

52 TRANSLATION

Furthermore, its bilingual character (for Latin or polysyllabic

English and native or monosyllabic English

are practically two tongues, now unhappily breaking

apart) gives it a reach and compass of expression that

few, if any, other languages can show. With such

resources at his command a composer in English can

vary his tone as it pleases him and can be familiar or

courtly, blunt or guarded, vivid or colourless at will.

Unhappily this is not all. 'Synonyms' are a source

of wealth to a language, so long as they are not used

as synonymous. And in its use of these our language

has from of old been lavish and undiscerning; and

to-day the riot is worse than ever. To a translator of

the classics again the double diction of English is often

as much of a hindrance as a help, and classical scholars

(the late Dr Verrall for example) not infrequently

prefer the rendering of English into Greek or Latin

to the converse. Commenting on this I said in a paper

read before the Classical Association 1

: 'English is a

composite language with two vocabularies, but Latin

is a simple language having but one.. . . The translator

from Latin into English is continually harassed by

having to make choice between these two strata in

our language, but the translator into Latin is free from

all such embarrassment.' The same trouble arises in

translating into English from French; as Messrs Ritchie

and Moore have noted, 'The wealth of synonym

largely due to the double element, Teutonic and

Romanic which allows English to reserve one synonym

for poetry and the other for prose, is not

paralleled in French '

(p. 23).

A flexible language English is not. Through

1

Proceedings ofthe Classical Association, vol. 8 (1911)1 p. 122.

ENGLISH AND GERMAN 53

the loss long ago of its inflexions, the order of words

in its sentences is little more than a vehicle of grammar

instead of an appliance for the grouping and emphasizing

of thoughts. Its marked and unaccommodating

individuality is another obstacle to the translator.

Italian, and German in spite of its clumsiness, will

come more readily to his hand.

The capabilities of German may be seen from a

felicitous version by Strodtmann of a stanza of Tennyson

which Professor Tolman quotes (p. n):

The splendour falls on castle walls

And snowy summits old in story ;

The long light shakes across the lakes

And the wild cataract leaps in glory.

Blow, bugle, blow ; set the wild echoes flying.

Blow, bugle ; answer, echoes, dying, dying, dying.

Es fallt der Strahl auf Berg und Thai

Und schneeige Gippel, reich an Sagen ;

Viel' Lichter wehn auf blauen Seen,

Bergab die Wasserstiirze jagen !

Bias, Hiifthorn, bias, in Wiederhall erschallend,

Bias, Horn Antwortet, Echos, hallend, hallend, hallend.

On German see also the opinion of Professor Gildersleeve

quoted below, p. 63.

We need hardly set out to show that some stages

in the development of a language are more

propitious to translation than others. The

Elizabethan period was such. What of the English

of to-day ? Of the alarming degeneration visible in

current writing I have spoken elsewhere 1 and here will

content myself with quoting the words of a critic and

writer of distinction, A. E. H., who says

2

:

' Written

1 Times Literary Supplement, August 15, 1918.

2 In a notice of the 'Cambridge History of Literature' in the Cambridge

Review of May 23, 1917.

54 TRANSLATION

English is now inert and inorganic: not stem and leaf

and flower, not even trim and well-joined masonry,

but a daub of untempered mortar.' No wonder that

the Editors of the Loeb Library should turn their eyes

to the Tudor Translations. More wonder that anyone

should find comfort in the reflexion which Mr Phillimore

took over from Mr Bevan in the Preface to his

Propertius(i9O9) :

'

I think that Mr Bevan in the Preface

to his excellent Prometheus Bound was right when

he argued that the present stage of the language is

peculiarly favourable to translators. The incipient

senile ataxy of English restores us something of the

receptiveness which in the Elizabethans was an effect

of juvenal elasticity.' For when did incapacity to

express your own thoughts make you a better exponent

of another's ?

I have spoken elsewhere of the defects of English

in details (Proceedings of tlie Classical Association, I.e.

p. 52). One of the worst is its confusion of the singular

and plural of the second personal pronoun. The

mischief goes far enough back; but some of the early

translators were aware of it and governed themselves

accordingly. In the latest verse translation of Plautus

Rudens, Sir R. Allison, 1914, lines 706 sq. 'Exi e

fano natum quantumst hominum sacrilegissume :

| uos

in aram abite sessum' are given as 'Out of the temple

wickedest of men! | And you sit by the altar' where

the meaning of Plautus is lost to the reader. The

translator of 1769 Bonnell Thornton avoids the pitfall

'Come forth, thou worst of sacrilegious villains. (To the

women} You seat you by the altar there.' So the Loeb

translation of Terence Eunuchus 1064 gives uobis and

tibi both as 'you.' Sallust lugurtha no 6 8

A CHECK TO LINGUISTIC DECLINE 55

deliberately varies between tu (Sulla) and uos (the

Roman people). But this variation is neglected or

obscured in all the English versions I have seen. The

French and Italian respect it. This confusion has had

a strange progeny. A fine poem of Longfellow springs

from nothing better. Plutarch Marius 12 has preserved

the mot of lugurtha when thrust into the dank Tullianum

: 'Hpa/cXet?, et><? tyvxpov VJAWV TO fta\avelov,

that is 'Hercules! how cold is your bath (i.e. this

Roman bath)!' But the poet, misconstruing and misremembering,

has transformed the exclamation as

follows:

How cold are thy baths, Apollo,

Cried the African monarch, the splendid,

As down to his death in the hollow,

Dark dungeons of death he descended

Uncrowned, unthroned, unattended.

How cold are thy baths, Apollo 1

.

It is not inopportune to remind ourselves of the

value of translation in arresting linguistic

decline. The close study of a model and the critical

attitude towards language which it implies are powerful

antidotes to the vices of expression so rampant to-day.

And in this regard the considered judgments of Quintilian

X 5. 2 sq. and of Pliny Epist. VII 9. 2 sq.,

'Quaeris quemadmodum in secessu, quo iam diu frueris,

putem te studere oportere. utile in primis et multi

praecipiunt, uel ex Graeco in Latinum uel ex Latino

uertere in Graecum; quo genere exercitationis proprietas

splendorque uerborum, copia figurarum, uis

explicandi, praeterea imitatione optimorum similia

1 I pointed this out over forty years ago, in the Academy of Oct. 30,

1880 ; but I fear that the readers and editors of Longfellow have paid

no heed.

56 TRANSLATION

inueniendi facultas paratur: simul quae legentem

fefellissent, transferentem fugere non possunt. intellegentia

ex hoc et iudicium adquiritur/ as well as the

words of Messrs Ritchie and Moore (p. 13): 'In France

it' (that is 'careful and deliberate translation') 'is a

regular part of language-study, especially in Latin,

and goes a long way to account for the high average

of prose style found among educated Frenchmen.

M. Vannier in his excellent book "La clart6 franchise"

remarks in this connexion "La version [i.e. such careful

rendering into the mother-tongue] est 1'exercice du

style par excellence'" may be commended to the notice

of those writers of to-day whose ignorance sloth and

perversity are debasing the English tongue 1

.

For the historian of a language or a literature the

question of the Best Age for translations may

have a certain importance. It has none for the translator

of the day. He will be wise, it is true, if before

beginning his task he inquire how far the current form

of the language is adequate for his purpose. But its

merits or demerits in a bygone age are nothing to

him. Time, the sworn foe of what is, does not spare

the best translations, whether they are in contemporary

or in archaizing speech: 'The best of translations as

I have said elsewhere2 are even from the first but poor

unsatisfying reproductions, and from the hour of their

making they steadily decline. As the words employed

1 Compare what Messrs Ritchie and Moore say on p. 3 1 :

' That

(i.e. translation from a foreign tongue) is the course which our great

writers have followed, that is the time-hallowed method of acquiring

skill in the use of English. Milton, Gray, Shelley, Byron, Tennyson

learned the secret of their style in translating from the Classics.'

2 'Dead Language and Dead Languages,' p. 17.

OBSOLESCENCE OF TRANSLATIONS 57

in them change their meaning, they become first inadequate,

next misleading and finally unintelligible.'

This Obsolescence of translations is recognized

in the common opinion that each age must have

its own translations, or that there is no such thing

as a 'final' translation. It might be expected to keep

pace with the change in the language. But in fact its

progress is more rapid, and translations age more

quickly than original works. For this the translators

are to blame. Their procedure may be gathered from

Sir T. H. Warren, op. cit. p. 100: 'Different ages have

different sympathies. The Romanticist finds Romanticism

in the Classics, the Impressionist Impressionism,

the Realist Realism.' Here for 'finds in' read 'imports

into,' and all is plain. The vision of a translator whose

eye is not singly on his task is caught by the beckoning

spirit of his age. Subtly and unconsciously he falsifies,

and what he is led to falsify is just that which at all

costs he should have preserved the spirit and character

of a foreign folk and bygone days. He wins the plaudits

of contemporaries, pleased with the discovery of ancient

worthies so like themselves, and unav/are how that

likeness has been produced. But the age passes,

posterity refuses the deceit, and the translation passes

too. The New Generation, unlessoned by experience,

returns to the original, and then, such are human

ways, works up another amalgam of its own. For, as

Messrs Butcher and Lang have said, 'the taste and

the literary habits of each age demand different

qualities in poetry and therefore a different sort of

rendering of Homer/ Preface to their translation of the

Odyssey. And so, each superseding its predecessor,

the Elizabethan conceits of a Chapman are followed

58 TRANSLATION

by the Augustan manner of a Pope, the ballad manner

of a Maguire and the Romantic manner of a Worsley.

All translation sooner or later must pass into oblivion.

But the translator of classics who desires for his

work the utmost life that mortality concedes must

exorcise the spirit of his age. The handiwork

must be of the present, but the inspiration come from

the past.

We can see from this how ill-judged are the attempts

to retain or restore as 'translations' renderings now

obsolete or obsolescent. The high, indeed matchless

claims of the Authorised Version of the Bible to

respect and attention rest now on its merits as English

writing, not on its merits as representing the Hebrew

and Greek originals. I yield to no man in admiration

of the Tudor Translations ; but I cannot gainsay that

when, for example, Philemon Holland renders in his

inimitable style Livy praef. 8, 9 as follows :

Sed haec et his similia ut- But these and such like

cumque animaduersa aut ex- matters howsoeuer they shall

istimata erunt haud in magno hereafter be censured or

equidem ponam discrimine : esteemed I will not greatly

ad ilia mihi pro se quisque weigh or regard. This would

acriter intendat animum quae I haue everie man rather to

uita, qui mores fuerint per thinke upon in good earnest,

quos uiros quibusque artibus and consider with me, what

domi militiaeque et partum et their life, and what their carauctum

imperium sit ; labente riage was : by what men and

deinde paulatim disciplina means both in war and peace,

uelut desidentes primo mores their dominion was atcheeved

sequatur animo, deinde ut and enlarged. Afterward, as

magis magisque lapsi sint, turn their discipline began by little

ire coeperint praecipites donee and little to shrinke, let him

ad haec tempora quibus nee marke how at the very first

uitia nostra nee remedia pati their behaviour and manners

ARCHAIC TRANSLATION 59

possumus peruentum sit, sunke withall ; and how still

they fell more and more to

decay and ruine and began

soone after to tumble downe

right even untill these our

dales wherein we can neither

endure our owne sores, nor

salves for the cure,

his Elizabethan English will not, except to experts in

past English literature with the faculty of historical

imagination, reproduce the effect of the original (which

is better rendered by Mr Naylor's translation in the

book already cited), and that to an ordinary reader

his Livy will appear an antiquated author.

In translating from the Ancients it may be asked

how far a translator should take into account

the Age of the Original. When in any literature

a work is primitive, as the Homeric poems in Greek,

or deliberately antique, as the writings of Sallust in

Latin, archaisms are in place. Hence, so far as our

principle is concerned, the contention of Butcher and

Lang, Preface to their translation of the Odyssey,

' The Biblical English seems as nearly analogous to

Epic Greek as anything our tongue can offer,' seems

sound. But we must not thus deal with Demosthenes

or Cicero. Works written in what was contemporary

speech must be translated into what is contemporary

speech 1

.

Ancient features must of course be preserved. But

they must not be exaggerated by the use of words or

phrases which suggest archaism to the reader where

there was none to the writer. Inattention to this has

sometimes injured Mr H. J. Edwards's otherwise good

1

'Contemporary,' however, must not be confused with colloquial.

60 TRANSLATION

translation of the 'Gallic War,' as at B. G. II 25 where

the words in italics have not the true Caesarian ring.

Scuto ab nouissimis militi- Taking therefore a shield

bus detracto, quod ipse eo sine from a soldier of the nearest

scuto uenerat, in primam aciem ranks, as he himself was come

processit centurionibusque no- thither without a shield, he

minatim appellatis reliquos went forward into the first line

cohortatus milites signa inferre and calling on the centurions

et manipulos laxare iussit quo by name and cheering on the

facilius gladiis uti possent. rank and file he bade them

advance and extend the companies,

that they might Ply

swords more easily.

If Fidelity requires translators to be true, as far as

may be, to the age and nationality of an author it

requires them no less to respect his Individuality,

or, as Dryden puts it in his Preface to Sylvae or the

Second Part of Poetic Miscellanies (Ker, p. 254), 'to

maintain the character of an author which distinguishes

him from all others, and makes him appear that

individual poet whom you would interpret.' He continues:

'For example not only the thoughts but the

style and versification of Virgil and Ovid are very

different ; yet I see even in our best poets who have

translated some parts of them that they have confounded

their several talents... if I did not know the

originals, I should never be able to judge by the

copies which was Virgil and which was Ovid.' The

translator, to use a happy phrase of Matthew Arnold

op. cit. p. 41, must be penetrated by a sense of his

author's qualities. He must sink himself if his author

is to emerge 1

.

1 In this regard how important is attention to 'mere details,' I have

already said, p. 39. If a reader is enabled to distinguish in the Loeb

IMPROVEMENTS ON AN AUTHOR 61

A translator so penetrated will find fidelity no

burden. It will be easy for him to follow the precept

rather than the practice of Pope (pp. 3, 6 above) and to

carry out the injunction of Roscommon:

Your author always will the best advise ;

Fall when he falls, and when he rises rise.

He will resent a compliment such as Sir George

Young paid to Professor Murray that he much prefers

Mr Murray to Euripides ; nor will he be envious

of the excellence of FitzGerald's work, holding with

Mr G. K. Chesterton of Omar Khayyam that it is

much too good to be a good translation.

Conscious Improvement upon his author, as we

have already seen, he will not attempt. He will leave

Hesiod his homeliness and Cicero his diffuseness.

Against Unconscious Improvement too he will

be upon the watch, though he will not flinch from

it in the one case where it is unavoidable. This is

when the language into which he is translating has

itself a higher literary character than that from which

he translates. Hence might be defended ' Frazer's

mistake if you can call it such in his monumental

work on Pausanias ' Tolman op. cit. p. 30.

But it is no mistake of Frazer's. His English is on

a higher plane than the decadent Greek of Pausanias ;

and there is no reason why he should spend pains

to bring it down to that level. And if in Professor

Wilamowitz's rendering of Goethe's Bin Gleiches, op.

cit, p. 17:

Translation of the 'Corpus Tibullianum' between the work of Tibullus

and the work of Lygdamus, long attributed to him, he owes it solely to

this attention.

62 TRANSLATION

Ueber alien Gipfeln xopixpcus fi

1st Ruh ;

In alien Wipfeln tirl

Spiirest du o-i'yato-' aqrat

Kaum ein Hauch. opvtw 8f 6p6os KOT' v-

Die Vogelein schweigen im \cu> tvdti av fie Paiov

Walde.

Warte nur, balde pLtwov, o8<ara, KOI trv

Ruhest du auch.

we feel that the translation is finer than the original,

the Professor is neither to be blamed nor to be commended

for this. It is Greek that is superior to German

as a medium of literary expression.

It has been held that translators should seek

in their native literature for analogues to the

works that they are translating and model their

translations thereupon. The notion will not bear examination.

Suggestions the translator should welcome

from whatever source they come, and those who have

been under the influence of his author are likely to

supply him with helpful ones. But in following up a

likeness too closely we may lose sight of the original.

True that Swinburne was deeply penetrated with the

spirit of Greece. But that is no reason why our vision

of Greek drama should pass through a Swinburnian

film. And, while agreeing with Professor Platt 1 as to

the kinship between Aeschylus and the Hebrew

prophets, one may at the same time doubt with him

' whether, despite their kinship, it is fair to load and

overload an Athenian dramatist with a style which

after all is alien to his.'

Professor B. L. Gildersleeve, who has devoted

1 Preface to 'The Agamemnon of Aeschylus freely translated.'

REMINISCENTIAL TRANSLATIONS 63

particular attention to the ethics and aesthetics of

translation, has protested against 'reminiscentiaP

translations in a passage already referred to from

which extracts may be given here.

In the hands of a master the German language, as is well

known, lends itself to translation much more readily than

English, not simply because of its various virtues upon which

I need not expatiate, but because of its comparative freedom

from reminiscential phraseology. Into the text of our literary

language have been woven threads from five hundred years of

continuous tissue and...no one can write English like a native

without enriching his discourse with the filaments of earlier

fabrics, distinctly the products of individual looms. Now a

language that is stiffened with such embroidery is hard to

translate from because so much is lost ; it is hard to translate into,

because it cannot wrap itself so closely round a foreign original

as a language which, if one excepts Luther's Bible to which

our Authorised Version is more than an offset has only a

century and a half of phrase-makers to supply the fibre....

Theoretically the translation ought to be achromatic. It may

be nothing but an etching ; but in the Muses' name do not

color an etching.... Vergil and Tennyson are near akin, and

when the eagle 'clasps the crag with hooked hands' there is a

certain satisfaction in recalling Palinurus, 'prensantem uncis

manibus capita aspera montis '

; but it ought not to work the

other way, and yet when Professor Tyrrell translates Ennius'

famous line :

Moribus antiquis stat res Romana uirisque

by

Broad-based upon her men and principles

Standeth the state of Rome,

Professor Tolman applauds the Tennysonianism 1

.

We may distinguish. That a rendering is

' reminiscential'

is no reason for choosing it. But is it a

reason for eschewing it? If it is the one that gives

the original best, may we not take it and let the

1 American Journal of Philology, vol. 22, pp. 105 sq.

64 TRANSLATION

reminiscence go hang ?

'

If,' as Mr Phillimore asks,

'there is a natural personal sympathy between Tibullus

and Andre" Chenier, why is it vicious for a French

rendering of Tibullus to remind the reader of his

modern analogue ?

' The only answer I can find is

this. If the reminiscent phrase is so intimately associated

with its alien origin that in the new setting it

will be but a purpureus pannus, then it must be

eschewed. Otherwise we are entitled and indeed

bound to make use of it. And the proper criticism to

pass on 'broad-based,' for which 'firm' would have

served, is not that as a translation it is

' reminiscent '

but that it is overdone. I see however no defence for

Dr Way's rendering of Euripides Cyclops 179 sq.

oSKOvv firei&r) TTJV vtaviv ti\(Tf

anavTts avrrjv SifKponjcraT' tv pepfi;

Well when you caught the naughty little jade

Didn't each man whip out his vorpal blade

And thrust her through, one after another, then ?

(Did the translator intend verpall)

On the other hand we may with Professor Tyrrell,

op. cit. p. 304, notwithstanding Milton's anticipation

commend Thornhill's rendering of Aen. IV 285 :

atque animum nunc hue celerem, nunc diuidit illuc,

This way and that dividing the swift mind,

and at Aen. IV 530 'aut pectore noctem | accipit' we

may 'take advantage of Lord Tennyson's musical

echo'

She ever failed to draw

The quiet night into her blood.

Nor should it be brought against Robert Whitelaw's

translation of Aeschylus Prometheus Vinctus 31

avB^ S)v aTfpjrjj rf)t>8f (ppovpT)<r(is irerpav.

So shalt thou sentinel this joyless rock,

PRINCIPLE OF COMMENSURATENESS 65

that it may be 'reminiscent' of a line in Scott's 'Lady

of the Lake' I 14. 15 sq. :

And mountains that like giants stand

To sentinel enchanted land.

COMMENSURATENESS

A translation must be true to its original in

Quantity as well as Quality1

. The two are not

independent, and inattention to the former cannot fail

to affect the latter. Professor G. G. Ramsay in a good

preface to a good translation of the 'Histories' of

Tacitus rightly insists on the impossibility of reproducing

all the mannerisms of his author. But terseness