ENGLISH PROSE NUMBERS

'ALL things', says Aristotle, ' are determined by number.' He is speaking (Rhetoric, iii. 8) of Greek prose, which he says should have rhythm, but should not be metrical. The rhythm, however, should not be strict; it should only go a certain length. He then discusses the feet which please the ear in Greek prose, and their diverse effects. In Mr. A. C. Clark's invaluable Fontes Prosae Numerosae (1909), a chain of Greek and Latin texts bearing on and illustrating the topic of prose rhythm, from Isocrates to Petrarch, is provided. The short preface anticipates Mr. Clark's tract, The Cursus in Mediaeval and Vulgar Latin (1910), in which he sums up and carries further the inquiries, hitherto little known in England, of Zielinski and other foreign scholars. The cursus, it will presently be seen, signifies certain special cadences, or sequences of feet, which are favoured at the ending, or at certain places only less emphatic than the ending, of a sentence. These cadences, to be defined below, were transferred from quantitative to accentual Latin, and thence to early Modern English. In the Church Quarterly Review for April 1912, Mr. John Shelly broke fresh ground, by investigating this transference in the case of the Collects and of parts of the New Testament. His article is entitled Rhythmical Prose in Latin and English. These studies came to the notice of Professor Saintsbury while his History of English Prose Rhythm (1912) was in the press, and he refers to them in various notes then inserted. His book is the first history of the theme; it furnishes a multitude of scanned and commented examples, and of nice analyses and judgements. It is a history and a body of criticism rather than a theory; but the frequent remarks on theory are highly suggestive,

though often avowedly provisional. I find myself in general agreement with them, though with some reserves, and use a rather different notation. On June 6 of this year Mr. Clark gave a lecture, now published, on Prose Rhythm in English. All these students together have opened new horizons; the whole question has assumed another colour during the last few years, since I first began to study it. I make free use of their results, acknowledging details where it is possible, and trying to re-state the problems that now seem to arise.

The present essay, for clearness, is divided into three sections. Section I treats of the separate and single feet that constitute English prose, of their comparative frequency and import, and of their differing value to the ear. The sequences of these feet are almost disregarded in Section I. Section II treats of the sequences of feet, that are found at the endings of the English sentence, or at other emphatic pauses in it; the cursus are a particular case of such sequences. It will then seem that the results of Section I and Section II are decidedly at cross purposes. Section III is an attempt to explain this apparent difference, and to reconsider the rhythmical relationship between verse and prose in English.

Some kindred questions are not discussed here at all. It will be assumed that accent' in English means energy or stress in the utterance; without considering its relations with pitch or quantity. Also I shall assume, what is not the case, that all accents are of equal energy and importance; marking, however, in many instances a subsidiary stress, by a grave accent. Some phoneticians use symbols to denote half a dozen degrees of emphasis, but this is hopeless in actual prosody, if only because it enlarges the chances of disagreement as to a particular scansion. Professor Saintsbury, like many others, uses quantity-marks in scanning English, though he carefully avoids dogmatizing

on the physics or physiology of the question. Accent-marks will be used here; but I must not be taken to think that accent alone constitutes a foot. I feel sure (with Mr. T. S. Omond) that though English syllables cannot have a strict time-value, English feet cannot be defined, in the last resort, without regard to time. But this question concerns verse more than prose. Yet the names of the classical feet are here used, without prejudice, meaning groups defined by the place of the accent and the number of the syllables. Thus 'decay' and 'distinct' and 'My Lórd' are all 'iambs', 'rápidly' is a 'dactyl', and so on: this usage is now familiar, and will, it is hoped, become clear to any non-classical reader as we proceed. Lastly, I shall not here discuss what is bad prose rhythm,—a large subject; the examples will all be of the other kind; but some of the conditions of excellent rhythm may become clearer on the way.

I. SINGLE FEET.

1. An English prose sentence, for the present purpose, consists of a number of feet, or groups of words, containing at least one accent in each foot, but sometimes two or even three accents. In verse the feet may and constantly do cut across the words (My bós | om's lórd'). In prose the foot, as here defined, must begin with the beginning and end with the end of a word, though not necessarily of the same word; the prose feet would be 'my bósom's | lórd', amphibrach and monosyllable. I see that Professor Saintsbury sometimes cuts the words in scanning prose. Cut they really may be in the utterance; but in Section III, I shall plead that this is due, not to the true prose-rhythm, but to the verse-rhythm intruding, and justly intruding. Meantime let it pass that the foot-unit in prose is made up of one or more entire words. Most of the feet that are common in our prose are to be heard in this sentence from Coleridge :

'What is Greéce at this présent | móment |? It is the country of the héroes | from Códrus | to Philopóemen ; and só | it would be, though all the sands of África should cover its córnfields | and ólive-gardens, | and not a flówer | were left | on Hyméttus | for a beé to múrmur in.'

Here there are five iambs (is Greéce'); four amphibrachs, -feet which Mr. Bridges, owing to their frequency, calls 'britannics' ('from Códrus'; 'its córnfièlds', which is almost an antibacchic', like 'and táll córn'); and one anapaest (for a beé'). The 'paeon' is a four-syllabled foot with one accent; three types are represented, namely two 'second paeons', accented on the second (' of África'); three 'third paeons' ('at this présent', 'on Hyméttus'). There is one fourth paeon (and not a flower'). There are three five-syllabled feet; these (called by Professor Saintsbury 'dochmiacs') may conveniently be thought of as cases of one or other paeon with a light syllable for preface ('It is the country', 'to Philopoémen'.1 There is but one trochee (móment'), and one monosyllabic foot, the first.

Now some of these features, such as the frequency of iamb and amphibrach, will be found to be features of English prose at large, whilst others, such as the relative abundance of paeons and five-syllabled groups, belong more to a special type of prose, the numerous, the periodic, the Latinized. The chief feet not figuring in this short passage of twenty-one groups are: the spondee (Jóhn Jónes', 'whíte róse'); the rarer molossus' ('Jóhn Jámes Jónes', púre white róse'); the somewhat rare first paeon ('édifices') and the commoner dactyl ('glórious'); and some less usual combinations of four- and five-syllabled feet, of which examples will appear.

2. There are four types of rhythm, irrespective of the number of syllables in the foot. These are:

6

1 'Epitrites', or four-syllabled feet with three accents, seem nearly always to break up into two, and Jóhn Jónes díed', 'bléss his deár heart'; I have therefore seldom regarded them.

(a) Rising rhythm. This, as here defined, begins on an unaccented syllable, ends on an accented one, and only contains one accent. Such feet are iamb, anapaest, fourth paeon. More light syllables are possible (and if it were só'). Even here a second stress begins to be set up on 'if'. But, in this and in all the types, the genius of English is against having more than three unaccented syllables running.

(b) Falling rhythm. This begins on an accent, and has one or more unaccented syllables following, as in trochee, dactyl, first paeon. After this limit, the tendency to set up new stresses is heard, e. g. in such a hideous word as 'éxtratèrritóriálity'.

(c) Waved rhythm, consisting of three syllables at least, begins and also ends on an unaccented syllable; accent occurs somewhere between. The amphibrach ('británnic'), the second and third paeons, and their extensions, are in waved rhythm. This is the most distinctive type of rhythm in prose as compared with verse, where its occurrence is doubtful. Also it admits of longer feet, for the reason given, than the two preceding types; for the accent coming midway allows of prolongation at either end. Burke has 'as if it were in the gámbols | of a bóyish | unlúckiness', in his hurried eloquence. But this must almost mark the limit.

:

(d) Level rhythm. Here the foot is wholly made up of accents, as in the monosyllabic foot, spondee, and molossus. The limit of length is less than in the other three types, for the ear will not bear more than three consecutive accents without forming a new foot. Level and waved rhythm combine in the bacchic' ('foúr ónly '), and in the 'antibacchic' (the Lord Gód'). These tend to split into two feet, but they need not do so; it is a question for the particular reader. The same remark is of course true of many other combinations. People cut up sentences in

1622

с