NOTE UPON THE SATURNIAN METRE

This metre is illustrated by Nos. 1-4 (?), 5-6, 8, 10, 12-13 in this selection. Three views have been taken of its character.

1. It was at one time supposed to be purely quantitative. This view had the support of Bentley, who in the Phalaris (226-8) identified the Saturnian with a metre of Archilochus.[11] 'There's no difference at all', he says blithely. In more recent times the quantitative theory, in one form or another, has numbered among its adherents scholars of repute: e.g. Ritschl, Lucian Mueller, Christ, Havet. To-day it may be said to be a dead superstition. Its place has been taken by what may be called the 'semi-quantitative' theory.

2. The 'semi-quantitative' theory was popularized in this country by H. Nettleship[12] and J. Wordsworth[13]. It enjoyed the vogue which commonly attends a compromise; and it still has its adherents, as, for example, E.V. Arnold[14] (who follows the Plautine scholar F. Leo). But the more it is examined the more it tends, I think, to melt into a 'pure-accentual' theory. 'It allows the shortening of a long syllable when unaccented (dĕvictis)', says Nettleship[15]. Surely to say that dĕvictis is 'allowed' for dēvictis is to abandon the cause outright. But it is considerations of a more general character which seem likely to render untenable both the 'quantitative' and the 'semi-quantitative' theories. The recent researches of Sievers[16] and others into the earliest metrical forms tend to shew that this metre is an 'Indo-European' heritage, and that it must be judged in the light of its Eastern and Germanic cognates.

3. The best opinion, therefore, in recent years has been strongly on the side of the view which makes the principle of the Saturnian metre purely accentual. At the moment this view may, in fact, be said to hold the field. Unhappily those who agree in regarding the metre as purely accentual agree in little else. We may distinguish two schools:

(a) There is, first, what I may perhaps be allowed to call the Queen-and-Parlour school. 'There cannot be a more perfect Saturnian line', says Macaulay, 'than one which is sung in every English nursery—

Place beside this English line the Latin line which has come to be regarded as the typical Saturnian—

If we accent these five words as Naevius and the Metelli would in ordinary speech have accented them, we shall have to place our accents thus:—

since by what is known as the Law of the Penultimate the accent in Latin always falls on the penultimate syllable save in those words of three (or more) syllables which have a short penultimate and take the accent consequently on the ante-penultimate syllable. But those who accommodate the Latin saturnian to the rhythm of 'The queen was in her parlour ...' have to postulate an anomalous accentuation:—

The Saturnian line is, they hold, a verse falling into two cola, each colon containing three accented (and an undefined number of unaccented) syllables—word-accent and verse-accent (i. e. metrical ictus) corresponding necessarily only at the last accented syllable in each colon (as Metélli ... poétae above).

Now here there are at least four serious difficulties:

1. While the principle of the verse is accentual half the words in any given line may be accented as they were never accented anywhere else.

2. Sometimes verse-accent and word-accent do not correspond even at the last accent in a colon. There is, for example, no better authenticated Saturnian than

Cornelius Lucius Scipio Barbatus:

and it is incredible that at any period in the history of the Latin language the word-accent ever fell on the middle syllable of Lucius[17].

3. The incidence of word-accent is left unfixed save so far as the incidence of verse-accent enables us to fix it. But the incidence of the verse-accent is itself hopelessly uncertain. In a very large percentage of saturnian lines we abandon the natural word-accent and have at the same time no possible means of determining upon what syllable of what word we are to put the verse-accent.

is simple enough: but when we come to

or

we come, to speak frankly, to chaos.

4. A large number of well-attested saturnians yield only two accents in the second colon.

(b) Beside the 'Queen-and-Parlour' theory there is what I may call the Normal Accent Theory. It originated with two papers by W.M. Lindsay in the American Journal of Philology vol. xiv—papers which furnish a more thorough and penetrating treatment of the whole subject than is to be found anywhere else. Lindsay's view is in substance this:

1. The saturnian line falls into two cola of which the first (a) contains three, the second (b) two accented syllables.

2. a contains seven syllables in all, b contains six (occasionally five), save when ᵕᵕ takes the place of one accented syllable.

3. The accent is always the normal Latin accent, according to the Law of the Penultimate.

(A tetrasyllabic word has two accents when it stands at the beginning of a line, and a pentasyllabic word always.)

4. Each line begins with an accented syllable.

These are the essential rules. In addition Lindsay has been at pains to determine carefully the accentuation of 'word-groups'. Each word in a Latin sentence has not necessarily an accent of its own. Thus apud uos is accented apúd-uos; so again in-grémium, quei-númquam, ís hic-sítus. No part of Lindsay's papers throws so much light on the scansion of the saturnian verses as that which deals with these word-groups: but it is impossible here to deal with the subject in detail. I will give here the first two Scipio Epitaphs (5. i, ii) as they are scanned and accented by Lindsay:—

i.

ii.

But is it certain, after all, that the accent-law in Saturnian verse is the Law of the Penultimate? There was, as is well known, a period in the history of the Latin language when this Law did not obtain, but all Latin words were alike accented on the first syllable. When this period ended we cannot precisely determine. But, as Lindsay himself points out, the influence of the old protosyllabic accentuation was not quite dead even in the time of Plautus.[18] Now the saturnian verse undoubtedly reaches back to a very remote antiquity: even of our extant specimens some are very likely as old as the eighth century. It is probable enough, therefore, that the accent-law known at any rate to the first saturnian poets was the old protosyllabic law. And when we remember the hieratic character of the earliest poetry, when we take into account the conservatism of any priestly ritual or rule, may we not suppose it possible that saturnian verse retained the ancient law of accentuation long after the Law of the Penultimate had asserted itself in ordinary speech and in other forms of literature? Accented, as Lindsay accents it, according to the Law of the Penultimate, the saturnian loses the lilt and swing which it has under the old 'Queen-and-Parlour' system.

is not a music to pray to or dance to or die to. A much easier and more lively movement would be

that is, the movement given by the old protosyllabic accentuation.

The suggestion that the protosyllabic accent survived as a conscious archaism in saturnian verse right down to the time of the Scipios is, I think, at any rate worth considering. It carries us into speculations far wider than the particular problem with which it is immediately concerned. For if the protosyllabic law did actually survive in this way we can the more easily explain the swift and decisive victory which the Hellenizing Latin poetry won over the old native verse. What was conquered was an archaism, something purely artificial. The conquering force was not merely Hellenism but Hellenism plus a complete and radical change in Latin speech.

If anyone cares to analyse the extant remains of saturnian verse in the light of this suggestion, I would formulate three rules which can, I think, be deduced:

1. Each line has five feet, and each foot contains one accented syllable plus either one or two unaccented syllables.[19] The first foot, however, may consist of a monosyllable.

2. The third foot must consist of a trisyllabic word or 'word-group'[20]: save that occasionally the second and third feet together may be formed of a quadrisyllabic (or pentasyllabic) word with secondary accent.

3. The first and second, and again the fourth and fifth, feet may be either disyllabic or trisyllabic: but (a) two trisyllables may not follow one another in the first two feet, and (b) if the fifth foot (usually trisyllabic) is a disyllable the fourth must be trisyllabic.

The normal type is

─́─ ── │ ─́─ ── │ ─́─ ── ──││ ─́─ ──│ ─́─ ── ──

││ ─́─ ── ──

A common variation in the first two feet is either ─́─ ── ── │ ─́─ ──, or ─́─ ── │ ─́─ ── ──. A somewhat rare variation in the last two is ─́─ ── ── │ ─́─ ──. In the first foot ─́─ sometimes replaces ─́─ ── (or ─́─ ── ──), no doubt owing to the greater stress at the opening of the verse.

Some exceptions (or apparent exceptions) to these rules will no doubt be found. But the rules cover most of the extant examples of saturnian verse: and it must be remembered that the text of our fragments is often not at all certain. The system outlined has, however, the merit—which it shares with Lindsay—that it dispenses with most of the alterations of the text in which other systems involve us.

The Hymn of the Arval Brotherhood.

I have given the text of this celebrated piece according to what may be called the Vulgate; and in the sub-title, in the Glossary and in my Introduction p. 1 I have followed the ordinary interpretation. I may perhaps be allowed here to suggest a different view of the poem.

It begins with an appeal to the Lares. These are apparently the Lares Consitivi, gods of sowing. Then comes an appeal to Marmar, then to Mars. Then the Semones are invoked, who, like the Lares, are gods of sowing. There follows a final appeal to Marmar.

It is pretty clear that the Mars, Marmar, or Marmor, invoked in such iteration is not the war-god, but Mars in his more ancient character of a god of agriculture. But if this be so, what are we to make of lines 7-9,

'Be thou glutted, fierce Mars, leap the threshold, stay thy scourge',—or, as Buecheler takes it, 'stand, wild god'? This sort of language is appropriate enough to Mars as god of war, but utterly inappropriate to the farmer's god[21].

Now it so happens that for

the monumental stone to which we owe this inscription offers at one point

Now, when we remember the Lares Consitivi and the Semones, does it not look very much as though satur stood for sator, as though fere were a blunder for sere, as though saii were the vocative of Saius, 'sower' (cf. Seia a goddess of sowing, and Greek σάω σήθω), as though sia were the imperative of the verb sio (moisten)[22], and as though, finally, berber were to be connected with the Greek βόρβορυς and meant 'loam'? (I would give much the same sense, 'fat soil' to limen: (from the root lib-: cf. Gk. λείβω λειμών).)

We get, then,

'Be thou the sower: sower Mars, sow the soil, moisten the loam'. And this suggests what ought to be the meaning of enos iuuate. enos ought to mean harvests, or at any rate something in that kind. And why should it not? Hesychius knew a word ἔνος which he glosses by ἐνιαυτός, ἐπέτειος καρπός. See Suidas s.v. and Herwerden Lexicon Suppletorium.

The Hymn is a hymn for Seedtime. We know, however, that the festival at which it was sung fell in the month of May. The explanation of this has been hinted at by Henzen.[23] Henzen points out that the Arval Brothers entered on their duties at the Saturnalia, and that their worship is probably connected in its origin with Saturn, the god of sowing. (See Varro L.L. 5, 57, and apud Aug. C.D. 7. 13 p. 290, 28, Festus s.v. Saturnus.) We must suppose, therefore, that at some date when the meaning of its words had been already lost this hymn was transferred from a seedtime festival to a harvest festival.