Letters on Poetry and Art

Appendix II: An Answer to a Criticism

Milford accepts, (incidentally, with special regard to the word frosty in Clough's line about the Cairngorm¹), the rule that two consonants after a short vowel make the short vowel long, even if they are outside the word and come in another word following it. To my mind this rule accepted and generally applied would amount in practice to an absurdity; it would result, not indeed in ordinary verse where quantity by itself has no metrical value, but in any attempt at quantitative metre, in eccentricities like the scansions of Bridges. I shall go on pronouncing the y of frōstў as short whether it has two consonants after it or only one or none; it remains frōstў whether it is a frosty scalp or frosty top or a frosty anything. In no case does the second syllable assume a length of sound equivalent to that of two long vowels. My hexameters are intended to be read naturally as one would read any English sentence; stress is given its full metrical value, long syllables also are given their full metrical value and not flattened out so as to assume a fictitious metrical brevity; short vowels even with two consonants after them are treated as short, because they have that value in any natural reading. But if you admit a short syllable to be long whenever there are two consonants after it, then Bridges' scansions are perfectly justified. Milford does not accept that conclusion; he says Bridges' scansions are an absurdity and I agree with him there. But he bases this on his idea that quantitative length does not count in English verse. It is intonation that makes the metre, he says, high tones or low tones—not longs and shorts; obviously, stress is a high tone of the greatest importance and to ignore it is fatal to any metrical theory or metrical treatment of the language—

Page 745

and so far I agree. But on that ground he refuses to discuss my idea of weight or dwelling of the voice or admit quantity or anything else but tone as determinative of the metre; he even declares that there can be no such thing as metrical length; the very idea is an error. Perhaps also that is the reason why he counts frosty as a spondee before scalp; he thinks that it causes it to be intoned in a different way. I don't see how it does that; for my part, I intone it just the same before top as before scalp. The ordinary theory is, I believe, that the sc of scalp acts as a sort of stile (because of the opposition of the two consonants to rapid motion) which you take time to cross, so that ty must be considered as long because of this delay of the voice, while the t of top is merely a line across the path which gives no trouble. I don't see it like that; the delay of motion, such as it is and it is very slight, is not caused by any dwelling on the last syllable of the preceding word, it is in the word scalp itself that the delay is made; one takes longer to pronounce scalp, scalp is a slightly longer sound than top and there is too a slight initial impediment to the voice which is absent in the lighter vocable and this may have an effect for the rhythm of the line but it cannot change the metre; it cannot lengthen the preceding syllable so as to turn a trochee into a spondee. Sanskrit quantitation is irrelevant here (it is the same as Latin or Greek in respect to this rule); but both of us agree that the Classical quantitative conventions are not reproducible in English metre and it is for that reason that we reject Bridges' eccentric scansions. Where we disagree is that I treat stress as equivalent to length and give quantity as well as stress a metrical and not merely a rhythmic value.

This answers also your question as to what Milford means by "fundamental confusion" regarding aridity. He refuses to accept the idea of metrical length. But I am concerned with natural metrical as well as natural vowel (and consonantal) quantities. My theory is that natural length in English depends on the dwelling of the voice giving a high or strong sound value or weight of voice to the syllable; in quantitative verse one has to take account of all such dwelling or weight of the voice, both weight or sharp dwelling by ictus (= stress) and weight

Page 746

by prolongation or long dwelling of the voice (ordinary syllabic length); the two are different, but at any rate for metrical purposes in a quantitative verse can rank as of equal value. I do not say that stress turns a short vowel into a long one, but that it gives a strong sound value (= metrical length) to the syllable it falls upon, even if that syllable has a short vowel and no extra consonants to support it. There is a heavier voice incidence on the first i of aridity than on the second: this incidence I call weight; the voice dwells more on it, sharply, and that dwelling gives it what I call metrical length and equates it to the long syllable, gives it an equal value.

Milford does not take the trouble to understand the details of my theory—he ignores the importance I give to modulations and treats cretics and antibacchii and molossi as if they were dactyls, whereas I regard them as only substitutes for dactyls; he ignores my objection to stressing short insignificant words like and, with, but, the—and thinks that I do that everywhere, which would be to ignore my theory. In fact I have scrupulously applied my theory in every detail of my practice. Take for instance

Here art is long by natural quantity though unstressed,² which disproves Milford's criticism that in practice I never put an unstressed long as the first syllable of a dactylic foot or spondee, as I should do by my theory. I don't do it often because normally in English rhythm stress bears the foot—a fact on which I have laid emphasis in my theory as well as in my practice. That is the reason why I condemn the Bridgean disregard of stress in the rhythm,—still whenever it can come in quite naturally, this variation can occasionally be made. It is a question of the relations possible between stress value and unstressed quantitative

Page 747

values in a quantitative metrical system, which is not the same as their relations in accentual or stress verse. My quantitative system, as I have shown at great length, is based on the natural movement of the English tongue, the same in prose and poetry, not on any artificial theory.

In stress hexameter only dactyls, spondees and trochees doing duty for spondees are counted; but in quantitative verse all feet have to take their natural value and to act as modulations of the dactyl and spondee while both in the opening foot and the body of the line amphibrachii and cretics abound, even molossi come in at times. Opening tribrachs are very frequent in my hexameter

Milford seems to think I have stressed the first short syllable in what would be naturally tribrachs and anapaests to make them into dactyls—a thing I abhor. Cf. also in Ahana initial anapaests:

or

Yĕt sŭrvīves | bliss in the | rhythm of our | heart-beats, | yĕt ĭs thēre | wonder

or again

Ănd wĕ gō | stumbling, maddened and thrilled to his dreadful embraces

or in my poem Ilion

There are even opening amphibrachs here and there

24 December 1942

Page 748