Prose Rhythm in Medieval Latin from the 9th to the 13th Century PDF

Preview

Summary

ACTA U N IV E R S IT A T IS S T O C K H O L M IE N S IS Studia Latina Stockholmiensia --------------------------------------------------- X X ------------------------------------------------------ Prose Rhythm in Medieval Latin from the 9th to the 13th Century By T O R E JANSON ACTA U N IV E R SIT A...

Description

ACTA U N IV E R S IT A T IS S T O C K H O L M IE N S IS Studia Latina Stockholmiensia --------------------------------------------------- X X ------------------------------------------------------

Prose Rhythm in Medieval Latin from the 9th to the 13th Century

By T O R E JANSON

ACTA U N IV E R SIT A T IS ST O C K H O L M IE N SIS The series includes theses and other studies by members of the University. Subscriptions to the series and orders for single volumes should be addressed to any international bookseller or directly to the publishers: Almqvist & Wiksell International Postbox 62, S-101 20 Stockholm, Sweden

Universities, libraries, learned societies, and publishers of learned periodi­ cals may obtain the volumes of the series and other publications of the U ni­ versity of Stockholm in exchange for their own publications. Inquiries should be addressed to Kungliga Biblioteket, Box 5039, S-102 41 Stockholm, Sweden, or to Stockholms universitets bibliotek, avd. Odengatan 59, Box 6404, S-113 82 Stockholm, Sweden. Romanica Stockholmiensia Stockholm Contributions in Geology Stockholm Economic Studies. New Series Stockholm Economic Studies. Pamphlet Series Stockholm Oriental Studies Stockholm Slavic Studies Stockholm Studies in Classical Archaeology Stockholm Studies in Comparative Religion Stockholm Studies in Educational Psychology Stockholm Studies in English Stockholm Studies in Histoiry Stockholm Studies in History of Art Stockholm Studies in History of Literature Stockholm Studies in Linguistics Stockholm Studies in M odern Philology. New Series Stockholm Studies in Philosophy Stockholm Studies in Psychology Stockholm Studies in Russian Literature Stockholm Studies in Scandinavian Philology. New Series Stockholm Studies in Sociology Stockholm Studies in Theatrical Histoiry Stockholmer Germanistische Forschungen Studia Graeca Stockholmiensia Studia Hungarica Stockholmiensia Studia Juridica Stockholmiensia Studia Latina Stockholmiensia Studies in North-European Archaeology

T O R E JA N SO N

PROSE R H Y T H M IN M EDIEVAL LA TIN

ACTA U N IV E R S IT A T IS S T O C K H O L M IE N SIS Studia Latina Stockholmiensia --------------------------------------------

X X ----------------------------------------------

Prose Rhythm in Medieval Latin from the 9th to the 13 th Century

By T O R E JANSON

Berlingska Boktryckeriet Lund 1975

Table of Contents

1 Introduction

7

2 Questions of Method 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9

10

The problem 10 A system of notation 13 The method of external comparison 15 The method of internal comparison 19 The method of internal comparison in practice 22 A limitation on internal comparison 26 Consillabicatioj accent, and related problems 28 Principles of sampling 32 Arrangement of tables 33

3 Survival and Revival: From the 9th through the 11th Century 3.1 3.2 3.3 3.4 3.5

Preliminaries 35 The main tradition: Hincmar, Anastasius, and contemporaries Northern Italy 40 The Roman chancery 45 Germany and France 50

4 Prevalence of Cursus: The 12th Century 4.1 4.2 4.3 4.4 4.5 4.6

5 The Medieval Theories of Cursus 5.1 5.2 5.3 5.4 5.5 5.6

60

The papal chancery in 1088 60 Leoninus cursus and southern Italy 63 Further development in the chancery 69 France 72 The early artes dictandi 77 The school of Florence 79

80

The problems 80 The theory of the French school 81 The practice of the theorists 86 Gregory V III and Peter of Blois 92 Bernhard of Meung and the school of Orléans Transmundus and Bene of Florence 101

98

36

35

6 Conclusion

104

Appendix 1. Tables

107

Appendix 2. Procedure of Calculation Appendix 3. Texts Bibliography

118

128

Index of Latin Authors

132

116

Introduction

Questions of medieval Latin prose style have attracted the attention of few investigators. And it is symptomatic that two great scholars who have written large works entering into the field were both really engaged in other pursuits: Norden (1918) followed the fate of classical prose style to the bitter end, and Curtius (1954) traced the kinds and themes of modern European literature backwards^ through the Latin Middle Ages, to antiq­ uity. O n the whole, little has been done in the field, and Norberg (1968) has stressed that not only the study of prose style, but of medieval Latin in general, has just begun: indeed, his book is the first full-scale survey of the subject.^ Still, there are very good reasons for trying to learn more about the prose style and its history. Apart from the obvious interest for students of literature, different medieval Latin prose texts are of crucial importance for scholars in many areas: general history, the history of art, religion, philosophy, and so on. If medieval Latin philologists could tell just a little more about the literary ideals and traditions that helped to form all those works, it would contribute significantly to the correct appreciation of them. So much, then, about the field in general. This study concerns the practice of ending sentences and clauses in certain accentual rhythms, or cursus.^ It was discovered by Valois (1881) that this technique was not only described in medieval texts but was also used in papal bulls from the 12th century on. About a decade later, it was proved by several scholars, viz. Couture (1892), Havet (1892), and Meyer (1893), that the rhythmical clauses existed in late antiquity and that the origin of the accentual patterns ^ Cf. especially Norberg (1 9 6 8 :9 2 ). For further information about Latin prose style in general, the reader is referred to Norberg’s study, and to the discussion of Schaller (1958), who gives ample references. ^ I use the word cursus in the sense it has acquired in modern scholarly literature. For a discussion of the term, see below, section 4.2.

was a system of metrical patterns. Further work showed that the metrical clausulae had a long and im portant history both in Latin and in Greek prose; and even before that, Meyer had shown (1893) that accentual rhythms similar to the Latin ones were used in medieval Greek prose. In the first decades after the original discovery, several studies were published on the cursus. Later on, however, interest was centered on the classical system of metrical clausulae: several im portant works were written in the first decades of this century, and after a lull, new studies have been published recently, e.g. by Drexler (1967) and Primmer (1968). The transitional period in late antiquity has been very competently investigated by Nicolau (1930) and Hagendahl (1937, 1952). But the medieval cursus was more or less abandoned for about half a century, until the publication of Lindholm’s thesis (1963).® The primary aim of that work is to investigate the decline and disappearance of cursus in the wake of the Renaissance. By the introduction of the rigorous methods of investigation developed by Hagendahl (and, before him, de Groot) into the study of medieval cursus, Lindholm has also greatly furthered all future study in the field. U nfortu­ nately, not much has been published after her book.* This study will be concerned with the history of cursus from the latter part of the 9th century to the first part of the 13th. The starting point is chosen so as to avoid the period of transition between metrical and accen­ tual systems: a special study with a methodology of its own would be needed to bridge the gap between the cursus mixtus studied by Hagendahl and the purely accentual cursus of the Middle Ages. The study ends soon after 1200, because at that time one dominant system of cursus had been established, that of the Roman curia. W hat happens after that time is mainly that this system gradually spreads over Germany and England, and in due time reaches the borders of the Latin-speaking world, Poland and Scandinavia. The decline and disappearance of the system have been studied by Lindholm. There are four main parts to the study. Chapter 2 is devoted to questions of method. For the purpose of this investigation, Lindholm’s methods were ® In spite of its title, the three-volume work by di Capua (1937— 1946) on It ritmo prosaico nelle lettere dei papi e nei documenti della cancelleria romana dal IV al X IV secolo does not reach beyond the sixth century, except for some very general considerations. The same author’s Fonti ed esempi per lo studio dello ‘^stilus curiae romanae” medioevale (1941) is a useful collection of texts but gives almost nothing new. The long article by Rajna (1932) is the only important contribution to the study of cursus between the tens and the sixties. ^ Vidmanová (1965) was able to reassess an earlier work on cursus in the works of Hus by applying Lindholm’s methods. The article by Fletcher (1969) is worthless.

not sufficient. Another method of comparison had to be devised: it is presented and discussed in this first part. C hapter 3 describes, at some length, the little-known history of cursus in the 10th and 11th centuries. In Chapter 4, it is shown how the usage of cursus developed in the course of the 12th century. The subject of Chapter 5 is the medieval theory of cursus, as it is laid down in the earlier texts on the subject, from about 1180 on for about a century. It is also demonstrated that a very particular usage of accentual rhythm described in some of the theoretical texts can actually be found in practice. Finally, there is a brief summary, with a discussion of the possible importance of the new findings for studies in medieval Latin prose.

Questions of Method

2.1 The problem The subject of this investigation is primarily the rhythmical cadences of clauses and sentences found in medieval prose texts. In practice, interest will be focused on ends of sentences: only for one period will we have to pay particular attention to ends of clauses within the sentences (cf. below, Chapter 4). Furthermore, there is a little to be said about the rhythm of the beginning of a sentence (below. Chapter 5). For the main investigation, it is necessary to find a method whereby the rhythms at the end of sentences used by different authors can be described, studied and compared correctly and efficiently. The purpose of this chapter is to present such a method. Before we can embark upon the question of methods of study, however, we have to be perfectly clear about what should be studied. Unfortunately, there is no common agreement about the nature of the rhythmical cadences. The principal viewpoints are that represented by medieval theorists and by the rediscoverer of cursus, Valois (1881), and that represented by Meyer (1893) and most modern scholars after him. The easiest way to see the differences between these views is to start with that on which all agree. It is generally accepted that authors who cared for the rhythmical cadences preferred certain patterns, among which were those illustrated by the examples hominem recepistis (called cursus velox), illum deduxit {cursus planus) and ire tentaverit [cursus tardus)} Further, all assert that if the cadence consists of two words (which is definitely the normal case) what matters about the first word is its accent, while the number of syllables before the accent is immaterial. Thus, attentissime commendamus is as good a velox as hominem recepistis, and resilire tenta­ verit is a tardus just like ire tentaverit.

^ The Latin examples are taken from medieval authors, mostly Peter of Celle and Peter of Blois. For the medieval (rather late) use of the terms velox, planus, tardus, see Lindholm (1963: 15) and Valois (1881: 193).

But here agreement ends. For medieval theorists, the cadence consists of words, normally two (we shall discuss the other cases later on). In the first word, as we saw, only accent is relevant. In the second word, accent and length are relevant. Thus, the cursus velox, {hominem recepistis) is described (although with varying terminology) as a tetrasyllable paroxytone word, preceded by a proparoxytone word. Similarly, the planus {ilium dedúxit) is seen as a trisyllabic paroxytone word, preceded by a paroxytone, and the tardus {ire tentaverit) as a tetrasyllable proparoxytone preceded by a paroxytone. However, Meyer (1893) introduced a completely different analysis. For him, the decisive characteristic of the cadence is not the length of the last word in itself but rather the num ber of syllables between the two last accents and after the last accent. The planus, for example, is seen as a cadence with two unaccentuated syllables (“Senkungen” ) between the two last accents, and one syllable after the last accent. This amounts to a contention that the place of the word-limit is immaterial, or at least not of decisive importance. Thus, Meyer asserts that ends like vineam nostram are by-forms to the usual planus {ilium dedúxit). Further, Meyer also proposed that there is a fourth preferred form of cursus, in addition to the three named above, viz. the one characterized by three unaccentuated syllables between the last two accentuated syllables and one after the last one, as in dgnos admittatis (named trispondiacus by modern scholars).^ Meyer’s view on the structure of rhythmical cadences has won wide acceptance, so much, in fact, that it can be said to be the view of modern scholarship. It is presented without any reservations in handbooks like Leumann, Hofmann & Szantyr (1965: 716). Lindholm accepts it wholly, and she is the first to carry out Meyer’s principle completely in practice. For in her very thorough statistical study, she indicates exactly which types of cadences she judges to belong to each form of cursus. Faithful to Meyer’s principle, she admits as forms of planus not only ilium dedúxit and vineam nostram but also nec impetravi, and, assuming a secondary accent, exaudie­ tis (cases with monosyllables are to be discussed later). However, it cannot be said that any very convincing piece of evidence has been presented which would corroborate Meyer’s hypothesis. For it is clear that Meyer’s suggestion is not a mere rearrangement of known facts: it is a real hypothesis in that it predicts that certain forms of cadences, such

^ The name trispondiacus seems to have been first used by Vacandard (1905). It should be noted that the term spondaicus is actually found in the works of the 14th century theorist Robert of Basevorn: see di Capua (1941: 77).

as vineam nostram, can be shown to have been preferred by medieval authors, although none of them says so. As a m atter of fact, almost no evidence at all has been adduced for or against the hypothesis. Meyer has simply been believed on the sheer attractiveness of his theory. Almost the only facts presented so far that are of relevance for the question are those found in Lindholm’s work. Through her diligence and care, we have a large body of reliable data on the usage in the late Middle Ages. O n the whole, these data are not very favorable to the hypothesis. In the general survey of the forms of cursus (Lindholm 1963: 39— 54) she notes herself that the types exaudietis, пес impetravit, and vineam nostram for planus, as well as vestrarum largitione for velox, are rare or avoided in the rhythmical authors. More evidence pointing in the same direction can be found in her tables. This, of course, in no way controverts Meyer’s hypothesis. In the first place, Meyer himself suggested, in his very short sketch of the development of prose-rhythm, that forms like vineam nostram occurred rather before the 12th century than after that time. In the second place, there is an im portant question of method that has to be solved before we are able to verify or refute any statements whatsoever on the use of cursus. T hat question is: how do we know whether a certain type of end of sentence is desired by the author or not? At first glance, the answer may seem obvious. If a certain cadence is desired, it will be used more often. Thus, one may find the answer by counting the number of different cadences in the author’s text. This is correct, of course, as far as it goes: but what is “more often” ? It is obvious that all possible end rhythms are not equally frequent even with writers who do not care about prose rhythm. For example, trisyllabic words are common, but words of more than six syllables are very uncommon in Latin. Thus, a cadence like ilium dedúxit will always be rather usual, but one like Petrum excommunicaveram is bound to be infrequent in most texts. Suppose, now, that someone proposes that just these two cadences are in fact desired by a certain author. How can that be proved or disproved? If we find that out of 1000 cases, the writer has used the type ilium deduxit in 240 cases and the type Petrum excommunicaveram in 10 cases, does that prove or disprove the hypothesis that these cadences are desired? The answer must be, of course, that the figures in themselves cannot prove or disprove anything as long as we have nothing with which to compare them. And what we need for comparison are the numbers for the different cadences to be expected in a text by the author, if he did not particularly care for them. In short, we need a method by which it is possible to

compare the actually observed frequency of any cadence with the frequency to be expected if the author is indifferent to the use of that cadence. It is the aim of this chapter to give a solution to that problem of method.

2.2 A system of notation To begin with, it is necessary to list and denote, in some practical way, all the possible rhythmical cadences. In this section, we shall limit ourselves to the cases where the cadence consists of two words: the others are dis­ cussed a little later. There, something will also be said about the concept of word. Provisionally we regard the word as simply a graphical unit. Thus, what we have to list here are all the possible two-word combinations that can occur at the end of sentences. As we have seen above, it is not necessary to take into account the full shape of the first word, but only its accent. In view of the Latin accentual laws, there are three possibilities. If the first word is a monosyllable, the accent must of course fall upon it, and we have an oxytone word. All other words must have the accent on the second or the third syllable from the end, i.e. they must be paroxytone or proparoxytone. A convenient way to notate these possibilities is to write pp for a proparoxytone word, p for a paroxytone, and 1 for a monosyllabic. As for the last word, it is necessary to account both for its length, in number of syllables, and for its accent. To denote the different possibilities, I have adopted the method of simply writing the figure for the number of syllables, followed by p or pp to determine the accent. Thus, 3p stands for words like deduxit, 4pp for words like consilio, 5pp for words like effica­ citer, and so on.® Monosyllabic and bisyllabic words are denoted as 1 and 2, respectively, without any marking for accent, as their accent does not vary.^ Now, we are able to describe any end of a sentence with regard to its rhythmical structure. The type ilium deduxit, for example, is denoted as p 3p. The regular cursus velox, such as hominem recepistis (or attentissime commendamus) is written as pp 4p, and the tardus, such as resilire tenta-

® As can easily be seen, this system is simply an adaption to a new purpose of the ingenious and practical method for describing rhythmical Latin verse that was introduced by Norberg (1958). When this had been written, I discovered that Paladini and De Marco (1970) used a similar notation. ^ I have encountered only a few instances of the bisyllabic words said by gram­ marians to be oxytone, such as illic, viden, etc. These have been excluded from the samples as cases where the accentuation is uncertain (cf. section 2.8).

verit, is written p 4pp. The jus...