A Descriptive Syntax of the Peterborough Chronicle from 1122-1154 [Reprint 2011 ed.] 9027916136, 9789027916136, 9783110806410

Citation preview

A DESCRIPTIVE SYNTAX OF THE PETERBOROUGH CHRONICLE FROM 1122 TO 1154

JANUA LINGUARUM STUDIA MEMORIAE NICOLAI VAN WIJK DEDICATA

edenda curat

C. H. VAN SCHOONEVELD INDIANA UNIVERSITY

SERIES

PRACTICA 103

1971

MOUTON THE HAGUE • PARIS

A DESCRIPTIVE SYNTAX OF THE PETERBOROUGH C H R O N I C L E FROM 1122 TO 1154

by

D A V I D L. S H O R E S OLD DOMINION UNIVERSITY

1971

MOUTON THE HAGUE · PARIS

© Copyright 1971 in The Netherlands. Mouton & Co. N.V., Publishers, The Hague. No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publishers.

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 70-111620

Printed in The Netherlands by Mouton & Co., Printers, The Hague.

For Betty my wife

PREFACE

In the last few years there has been a steadily growing interest in the study of early English syntax. Up until the last thirty or forty years most linguistic investigations were based on morphology and only incidentally on syntax. A brief glance at any Old or Middle English grammar will at once verify this statement. This earlier dependence on morphology led scholars to describe English up to the eleventh century as a language in which grammatical functions and relationships depended on inflection alone — or at least, to a very large degree — and in which word order had very little significance. This and other recent studies show that this was not exactly true. If we are ever to have a complete description of early English grammar and a sound and accurate description of the nature, growth, and development of the English language, it is essential to have descriptive analyses of the various extant prose and poetic works in Old and Middle English. This study, I believe, is a valuable part of the work that is now accumulating descriptive analyses of the syntax of prose and poetic works which survived the Middle Ages. I am happy to take this opportunity to express my gratitude to Professor William J. Griffin of George Peabody College not only for suggesting valuable improvements in this study but also for introducing me to language study. I would be remiss if I did not thank Professor Sherman M. Kuhn of the University of Michigan and Professor Charles R. Carlton of San Fernando Valley State College for encouraging me to go on with this study and for supplying me with helpful information. I am indebted, furthermore, to Charles Carlton, Robert Palmatier, Ann Shannon, and William Brown for ideas from their works which I used in the presentation of my data. I wish to thank Miss Anna Loe Russell for various kinds of cooperative and helpful assistance in locating and obtaining research materials and Louise DeVere for reading the proofs and making helpful suggestions. I would like to acknowledge the Old Dominion University Education Foundation with gratitude for supporting in part the production of this book. And, more especially, I wish to thank David, Jr., and Keith, my sons, for being patient even though they must have felt that their father was more interested in 'that

δ

PREFACE

book' than in them, and Betty, my wife, who merits deep gratitude for being both patient and understanding. Very special mention should be made of Betty for sharing her intelligence and for laboring devotedly with the typing from the rough draft through the final copy. In short, the book could not have been finished without her. David L. Shores Norfolk, Virginia December 14,1969

CONTENTS

Preface

7

List of Tables

13

1. Introduction 1.1. The Problem 1.2. The Text and Corpus 1.3. Importance of the Study 1.4. Related Research 1.5. Previous Studies of The Peterborough Chronicle 1.6. Edition of the Text 1.7. General Procedure of the Analysis 1.8. Explanation of Basic Concepts used in the Description 1.9. The Presentation of the Data

15 16 16 18 19 21 23 23 27 30

2. Primary Clause-Level Tagmemes 32 2.1. The Subject Tagmeme 32 2.1.1. The Pronoun or Modified-Pronoun Phrase 33 2.1.2. The Noun or Modified-Noun Phrase 34 2.1.3. The Nominal Clause 34 2.1.4. The Pronominal Adjective or Modified-Pronominal Adjective Phrase 35 2.1.5. The Double Subject 35 2.1.6. The Case of the Filler of the Subject Slot 35 2.1.7. Agreement with the Filler of the Predicator Slot 37 2.2. The Predicator Tagmeme 40 2.2.1. Person 40 2.2.2. Number 41 2.2.3. Tense 41 2.2.4. Passive 42

10

CONTENTS

2.3.

2.4. 2.5.

2.6. 2.7.

2.2.5. Mood 2.2.6. The Complex Verb 2.2.7. Special Patterns The Direct Object Tagmeme 2.3.1. The Noun or Modified-Noun Phrase 2.3.2. The Pronoun or Modified-Pronoun Phrase 2.3.3. The Pronominal Adjective 2.3.4. The Nominal Clause 2.3.5. The To-Infinitive The Object Complement Tagmeme 2.4.1. Case of the Filler of the Direct Object Slot The Indirect Object Tagmeme 2.5.1. Pronoun and Pronoun 2.5.2. Noun and Noun 2.5.3. Pronoun and Noun 2.5.4. Pronoun and Clause 2.5.5. Noun and Clause 2.5.6. Pronominal Adjective and Pronoun 2.5.7. Pronominal Adjective and Noun 2.5.8. The Dative-Accusative Combination 2.5.9. The Accusative-Accusative Combination 2.5.10. Case of Direct Object and Indirect Object The Subject Complement Tagmeme General Conclusions

44 46 49 50 51 52 52 53 54 54 55 55 56 57 57 59 59 60 60 60 62 62 64 66

3. The Order of Primary Clause-Level Tagmemes 3.1. Group 1 (Subject and Predicator) 3.1.1. Inversion of Subject and Predicator 3.2. Group 2 (Subject, Predicator, and Direct Object) 3.2.1. Nouns and Pronouns as Fillers of the Subject and Direct Object Slots 3.3 Group 3 (Subject, Predicator, and Subject Complement) 3.4 Group 4 (Subject, Predicator, Indirect Object, and Direct Object) . . . 3.5. Subjectless Clause Patterns 3.6. The Coordinated Predication Patterns 3.7. The Relative Order of Primary Tagmemes in Combinations of Two . . 3.8. General Conclusions

68 82 84 88 91 96 98 103 104 106 108

4. Secondary Clause-Level Tagmemes 4.1. Description of the Clause-Level Adverbial Tagmemes 4.1.1. The Introductory Formula Tagmeme 4.1.2. The Time Tagmeme

110 110 113 114

CONTENTS

4.1.3. The Location Tagmeme 4.1.4. The Manner Tagmeme 4.1.5. The Purpose Tagmeme 4.1.6. The Agent Tagmeme 4.1.7. The Relationship Tagmeme 4.2. Linear Ordering of Adverbial Tagmemes 4.2.1. Structures of Predication With One Clause-Level Adverbial Tagmeme 4.2.2. Structures of Predication With Two Clause-Level Adverbial Tagmemes 4.2.3. Conclusions Relating to Structures of Predication With Two Adverbial Tagmemes 4.2.4. Structures of Predication With Three Clause-Level Adverbial Tagmemes 4.2.5. Conclusions Relating to Structures of Predication With Three Adverbial Tagmemes 4.2.6. Structures of Predication With Four Clause-Level Adverbial Tagmemes 4.2.7. Structures of Predication With Five Clause-Level Adverbial Tagmemes 4.2.8. Structures of Predication With Six Clause-Level Adverbial Tagmemes 4.2.9. Conclusions Relating to Structures of Predication With Four, Five, and Six Adverbial Tagmemes 4.3. General Conclusions

11

115 116 116 117 117 117 118 120 129 129 138 138 141 143 143 143

5. Modified-Head Phrases 145 5.1. Part I: The Adjectival Modifiers 145 5.2. The Modified-Noun Phrase 146 5.2.1. Types of Structures Which Fill the Modifying Slots 146 5.2.2. The Tagmemes of the Modified-Noun Phrase 147 5.2.3. Linear Ordering of the Tagmemes of the Modified-Noun Phrase. 148 5.2.4. The Modified-Noun Phrase and Concord 151 5.2.5. Description of the Modifying Tagmemes 152 5.2.6. Patterns of Pre-Positional and Post-Positional Modifiers . . . . 1 5 9 5.3. Other Modified-Head Phrases 164 5.3.1. The Modified-Pronoun Phrase 164 5.3.2. The Modified-Pronominal Adjective Phrase 165 5.3.3. The Modified-Demonstrative Phrase 165 5.3.4. The Modified-Numeral Phrase 165 5.3.5. The Modified-Adjective Phrase 166 5.4. Special Modifying Relationships 166

12

CONTENTS

5.5. 5.6.

5.7.

5.8.

5.4.1. Genitive Constructions 5.4.2. Structures of Apposition 5.4.3. The Predicate Adjective Part II: The Adverbial Modifiers The Modified-Verb Phrase 5.6.1. The Negator 5.6.2. Multiple Negation 5.6.3. The Aspect Indicator 5.6.4. The Degree Indicator 5.6.5. The Direction Indicator 5.6.6. Other Adverbials Other Modified-Head Phrases 5.7.1. The Modified-Adjective Phrase 5.7.2. The Modified-Participle Phrase 5.7.3. The Modified-Pronominal Adjective Phrase 5.7.4. The Modified-Numeral Phrase 5.7.5. The Modified-Adverb Phrase 5.7.6. The Modified-Prepositional Group Phrase 5.7.7. The Modified-Clause Phrase General Conclusions

166 170 172 173 173 174 175 176 177 178 179 180 180 181 181 181 181 181 182 182

6. The Relationship and Joining of Clauses 6.1. Types and Levels of Dependent Clauses 6.2. Functions of Dependent Clauses 6.2.1. The Nominal Clause 6.2.2. The Adjectival Clause 6.2.3. The Adverbial Clause 6.3. The Parenthetical Clause 6.4. Coordination of T-units 6.5. Parataxis 6.6. Correlation 6.7. General Conclusions

184 186 187 187 190 194 202 204 208 210 212

7. Summary and Conclusions

214

Bibliography

222

Index

225

LIST OF TABLES

1. Inflectional Distinctiveness in Fillers of Subject Slots

37

II. Inflectional Distinctiveness in Fillers of Direct Object Slots

63

III. Inflectional Distinctiveness in Fillers of Indirect Object Slots

63

IV. Total Frequencies of Clause Groups and Patterns

73

V. Frequencies of Independent Clause Groups and Patterns VI. Frequencies of Dependent Clause Groups and Patterns

74 75

VII. Frequencies of Clause Groups and Patterns in First Continuation

76

VIII. Frequencies of Clause Groups and Patterns in Final Continuation

77

IX. Frequencies of Independent Clause Groups and Patterns in First Continuation. .

78

X. Frequencies of Independent Clause Groups and Patterns in Final Continuation. .

79

XI. Frequencies of Dependent Clause Groups and Patterns in First Continuation . .

80

XII. Frequencies of Dependent Clause Groups and Patterns in Final Continuation . .

81

XIII. The Relationship Between the Noun and Pronoun as Fillers of the Subject Slot and the SV and VS Orders in Independent Clauses

84

XIV. The Relationship Between the Presence of Adverbial and Connecting Tagmemes in Initial Position and the SV and VS Orders in Independent Clauses

86

XV. The Relationship Between the Presence of Adverbial and Connecting Tagmemes in the Initial Position and the SV and VS Orders with Nouns (N) and Pronouns (P) in the Subject Slot in Independent Clauses

88

XVI. The Relationship Between the Different Combinations of Nouns (N) and Pronouns (P) as Fillers of the Subject and Direct Object Slots and the Order of the Subject, Predicator, and Direct Object

92

XVII. The Relationship Between the Nouns and Pronouns as Fillers of the Subject and Direct Object Slots and the Order of the Subject, Predicator, and Direct Object in Independent Clauses

94

XVIII. The Relationship Between the Presence of Adverbials and Connecting Tagmemes in the Initial Position and the Order of Subject, Predicator, and Direct Object in Independent Clauses

95

14

LIST OP TABLES

XIX. The Relationship Between the Different Combinations of Nouns and Pronouns as Fillers of the Nominal Slots and the Order of the Subject, Predicator, Direct Object, and Indirect Object in Independent Clauses

100

XX. The Relationship Between the Presence of Adverbial and Connecting Tagmemes in the Initial Position and the Order of the Subject, Predicator, Direct Object, and Indirect Object in Independent Clauses

102

XXI. Frequencies of Occurrence of Subjectless Constructions

103

XXII. Frequencies of Occurrence of Coordinated Predications

105

XXIII. The Relative Order of Primary Tagmemes of Subjectful Predications in Combinations of Two

107

XXIV. The Relative Order of Primary Tagmemes of Both Subjectful and Subjectless Predications in Combinations of Two

107

XXV. Frequencies of Adverbial Slots and Their Fillers and Their Positions Relative to the Predicator

112

XXVI. Positions of Adverbial Slots and Their Fillers with Respect to the Predicator in Patterns with Only One Adverbial

119

XXVII. Linear Ordering of the Modifying Slots and Their Fillers of the Modified-Noun Phrase

149

XXVIII.

Linear Ordering and Co-Occurrence of the Modifier Tagmemes of Modified-Noun Phrases

150

XXIX. Patterns of Modified-Noun Phrases with Only Pre-Positional Modifiers (820) . .

160

XXX. Patterns of Modified-Noun Phrases with Post-Positional Modifiers (229) . . . .

161

1 INTRODUCTION

Over seventy years ago Otto Jespersen, the great Danish scholar, noted the neglect of syntax in historical studies of the English language.1 At just about the same time, C. Alphonso Smith in the opening sentence of his study of word order of Alfred's Orosius and jElfric's Homilies wrote that few subjects "have been so persistently slighted as that of the position of words and clauses".2 It would seem that after such urging, studies of syntax would abound, but Etsko Kruisinga as late as 1926 remarked that it was "strange that there should be no book of any type, whether advanced or elementary, on Old English syntax".3 Still later, George W. Small observed that syntactic research in Old English was so lacking that scholars had failed to produce an adequate method of conducting it.4 A chief reason for this neglect perhaps has been the belief that Old English was a highly inflectional and synthetic language which showed the various relationships within clauses and sentences predominantly by the variations of the forms of words. That is, inflection was the primary signal and almost autonomous, and syntax was based on it. In the last few decades, a lively interest has been shown in the study of early English syntax. It now appears that syntax is, if not more important than morphology, at least equally important in both diachronic and synchronic studies of early English. Of late, scholars have repeatedly revealed the significance of syntax in any account of the English language. This shift in interest may have been brought about because morphological studies have reached the saturation point, but a more likely reason is that several investigations6 have demonstrated that word order was functional as early as the ninth century. These studies have given firm support to Jespersen's notion that certain fixations of word order preceded the reduction of inflections. Some investigations have been concerned with limited word order patterns and their historical development, and others have been devoted to rather complete descriptive studies 1

Otto Jespersen, Progress in Language (New York, Macmillan and Co., 1894), p. 352. C. Alphonso Smith, "The Order of Words in Anglo-Saxon Prose", PMLA, VIII (1893), 210. s Etsko Kruisinga, "How to Study Old English Syntax", English Studies, VIII (1926), 44. 4 George W. Small, "On the Study of Old English Syntax", PMLA, LI (1936), 3. • These investigations will be listed and briefly described in the section on related research. 2

16

INTRODUCTION

of a single text, distinguishing and classifying the different types of word groups. Both of these approaches have yielded interesting and very significant results which have demanded the modification of some of the key notions about the historical development of the English language. These studies, furthermore, have accented a special need for a series of exhaustive descriptive synchronic studies of the existing examples of the language of all periods.

1.1. THE PROBLEM

The chief purpose of this study is to identify and describe the syntactic patterns in the latter part of the Peterborough Chronicle from the years 1122 to 1154, to tabulate their frequency, and to illustrate the descriptive statements with examples from the corpus. The approach is synchronic in that its concern is with the nature of the syntax of a corpus of a specific period, and it is quantitative in that its observations are based on frequency counts. The study, using a combination of traditional and more modern linguistic methods, will focus primarily on the analysis of clauses and the structure of larger syntactic units. Specifically, the study will center on the following problems: (1) (2) (3) (4)

the the the the

primary tagmemes of clause patterns and their order; secondary tagmemes of clause patterns and their order; nature of modification structures; relationship of clauses and the modes of joining them.

Attention, of course, will be given to the relevant relations between morphology and syntax. The purpose, however, is not to analyze morphology.

1.2. THE TEXT AND CORPUS

Of the different versions of the Anglo-Saxon Chronicle, the Peterborough Manuscript is the fullest and continues the longest, until 1154. Scholars now generally believe this manuscript, known as "E", was written by one scribe at one stretch up to 1121; they believe the same scribe then added at intervals annals 1122-1131 (the First Continuation). Later, early in 1155, the section dealing with events from 1132 to 1154 (the Final Continuation) appears to have been added by another scribe.® The corpus used in this study is the portion of the Peterborough Chronicle covering the ' For recent discussions of the paleographical features of the Peterborough Manuscript, see Dorothy Whitelock, The Peterborough Chronicle (Copenhagen, Rosenkilde and Bagger, 1954), pp. 13-35; and N. R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, The Clarendon Press, 1957), pp. 424-426. Cecily Clark mentions these aspects of the manuscript in her The Peterborough Chronicle 1070-1154 (Oxford, Oxford University Press, 1958), pp. xxx-lxvi and in "Gender in the Peterborough Chronicle, 1070-1154", English Studies, XXXVIII (1957), 109-110.

INTRODUCTION

17

years 1122 to 1154. This portion is regarded by scholars as an uncorrupted original document which was not only written but composed at Peterborough. Charles Plummer as early as 1899, in his revision of John Earle's Two of the Saxon Chronicles Parallel, stated that its origin was certain. "From end to end", he said, "it is unquestionably a Peterborough book." 7 The most recent study shows that Plummer's conclusion has been left unassailed, for Cecily Clark writes that the annals from 1122 on were composed at Peterborough and adds: And this makes them precious among medieval English texts, for ... there is hardly another of which it can be said both that it is an original, not a garbled copy, and also that its date and provenance are precisely known.8 She goes on to call attention to the vigorous and vivid language and notes its almost complete abandonment of the formulaic and annalistic patterns. 9 Dorothy Whitelock considers the Peterborough Chronicle important for the student of the English language because it is the earliest extensive example of the East Midland language, the dialect considered the chief ancestor of our modern standard English, and because it affords: ... the most substantial piece of English writing from the post-Conquest period. It was written at Peterborough, and thus the sections which were composed for it, not merely copied into it, ALLOW AN ESTIMATE TO BE FORMED OF TWELFTH CENTURY ENGLISH in a part of the Danelaw. [Emphasis mine.]10 The manuscript was among Archbishop Laud's collection of manuscripts and is now in the Bodleian Library in Oxford, where it is catalogued as MS. Bodley Laud Misc. 636. Charles Plummer's revised edition, the standard edition, has been the basis of most of the discussions of the Peterborough Chronicle and is the chief source for this study. Cecily Clark's book also includes an edition based on the original manuscript. Clark's edition is unlike Plummer's in that it does not closely follow the graphic devices used in the manuscript, but uses modern punctuation. It is, nevertheless, a reliable text for language study. With Plummer as the primary basis and Clark as the secondary, the sixteen annals were copied by the present investigator, after division into syntactic units, on analysis sheets, but only after a close comparison with a facsimile11 of the original manuscript. The most literal translation (and hence very useful for the present type of study) is S. I. Tucker's, 12 and the most literary is G. N. Garmonsway's. 13 7

John Earle and Charles Plummer, Two of the Saxon Chronicles Parallel (London, Oxford University Press, 1892-99), II, xxxv. 8 Clark, op. cit., p. xxx. * Ibid., p. xviii. 10 Whitelock, op. cit., p. 13. 11 This is included in Whitelock's edition, already cited in footnote 6. 12 S. I. Tucker, "The Anglo-Saxon Chronicle (1042-1154)", in English Historical Documents 10421189, edited by David C. Douglas and George W. Greenway (London, Eyre and Spottiswoode, 1953), pp. 102-203. 111 G. N. Garmonsway, The Anglo-Saxon Chronicle (London, Everyman, 1953).

18

INTRODUCTION 1.3. IMPORTANCE OF THE STUDY

If we are ever to have a sound and complete description of the nature, growth, and development of the English language, it is essential to have descriptive analyses of the various extant prose works in early English as well as of representatives of Modern English. This study, it is believed, will be doing a valuable part of the work that is now accumulating descriptive analyses of the syntax of prose and poetic works which have survived the Middle Ages. The earlier dependence on morphology, mentioned before, led scholars to describe Classical Old English as a language in which syntactic relationships depended on inflection alone, or at least predominantly, and in which word order had very little significance.14 Even Charles C. Fries reflected the traditional view in 1940: In Old English, however, the order of the words in such sentences (actor-action or subjectverb-object) has no bearing whatever upon the grammatical relationships involved. Taxemes of selection do the work, and word-order is non-distinctive and connotative.16 Recent studies show that this is not exactly true. However, before any conclusive generalizations can be made about the matter, there will have to be analyses of many more texts. Since 1930, a number of studies have been concerned with the historical development of word order patterns of English. Some of these studies,1® based on samplings from carefully selected manuscripts, centered on word order patterning of English from the Middle Ages to the present time. They stressed for the most part the relative position of subject, verb, and object. Some, however, were even concerned with the ordering of the finite and non-finite forms of the verb. For example, David Payne Harris in 1960 made a study of the positions of the verb and its auxiliary in twelfthcentury English.17 More complete and valuable studies have been made by Charles R. Carlton, Ann Shannon, Robert A. Palmatier and William H. Brown, Jr. 18 They are more complete and valuable in the sense that these scholars chose a single corpus and made a clause by clause analysis, describing the whole syntactic pattern, not just one selected relationship. The present study is a continuation of the Carlton-ShannonPalmatier-Brown work of a descriptive syntax of a single corpus.

14

This view is generally expressed in books on the history of the English language. Charles C. Fries, "On the Development of the Structural Use of Word Order in Modern English", Language, XVI (1940), 199. 11 For a brief discussion and bibliographical data on studies by Cassidy, Magers, Saitz, Harris, see the section on related research. 17 David Payne Harris, "Word-Order in Twelfth Century English", Studies in Language and Linguistics in Honor of Charles C. Fries, edited by Albert H. Marckwardt (Ann Arbor, The English Language Institute, University of Michigan, 1964), pp. 187-198. 18 A brief discussion of these works and bibliographical data will be presented in the section on related research. 16

INTRODUCTION

19

1.4. RELATED RESEARCH

It would be impractical to mention all the works which have had something to do with the study of early English syntax. Many of these have yielded some interesting information in spite of weaknesses and inadequacies. An excellent and thorough account of investigations before 1955 has been given by Robert L. Saitz.19 Charles R. Carlton, a few years later, in summary form noted the phases through which the study of early English syntax has gone, briefly discussed the treatment of syntax in some recent grammars, and mentioned the studies of limited patterns. 20 The present study is more closely concerned with those investigations which have been devoted to a synchronic description of a single dated corpus. It has, nevertheless, an affinity with some of the studies which have traced the development of limited patterns such as the subject-object relationship and the periphrastic genitive, that is, the studies which have contributed to an understanding of the relationship of inflections and word order as syntactic signals. Russell Thomas was perhaps the first to offer some confirmation that syntactic patterns were important signaling devices in early English when he found that the adnominal periphrastic genitive appeared in samples from all the periods and, more important, that the pre-positive genitive was established by the eleventh century, constituting 76.6 per cent of the genitive expressions that occurred in the samples studied.21 In his report of an investigation of the beginnings of the substitution of the periphrastic dative for the indirect object-direct object construction, Frederic G. Cassidy gave a complete historical account of the construction in a very extensive collection of texts of Old English prose. Although he was generally concerned with the problem of inflectional loss, his study specifically focused on the dative object-accusative object patterns. He found that periphrasis began to be used noticeably only after 1150. But what seems more important is the evidence that some of the word order patterns of Modern English were already well established around A.D. 900, when case distinction was still strong.22 He concluded that the pronominal dative object preceded the accusative object about 73 per cent of the time and that nominal dative objects followed the accusative object about 76 per cent of the time. By 1050, despite the inflectional distinctiveness of the dative and accusative objects, Cassidy concluded, case distinction was redundant, because the word order of Modern English was ocle

Robert L. Saitz, "Functional Word Order in Old English Subject-Object Patterns", unpublished Ph.D. dissertation, University of Wisconsin, Madison, 1955, pp. 4-41. 80 Charles R. Carlton, Descriptive Syntax of the Old English Charters (The Hague, Mouton, 1970), pp. 13-22. 21 Russell Thomas, "Syntactical Processes Involved in the Development of the Adnominal Periphrastic Genitive in the English Language", unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1931, pp. 111-113. 82 Frederic G. Cassidy, "The Backgrounds in Old English of the Modern English Substitutes for the Dative-Object in the Verb + Dative — Object + Accusative — Object", unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1938, pp. 86-87.

20

INTRODUCTION

curring 75 per cent of the time and by 1100, about 80 per cent. His corpus included over 10,000 examples from the early ninth century up to 1200. Mildred K. Magers in her study set out to demonstrate how a "syntactic order had replaced inflectional form as a grammatical vehicle". Her concern was the subjectverb-object order and the notion that this order appeared to fill the need of inflectional loss. She found that there was a definite trend by the tenth century toward the subjectverb-object pattern and concluded that word order as a grammatical device was well established and preceded the loss of inflection.23 Robert L. Saitz investigated word order and inflection in the subject-object relationship. After a detailed look at the possibility of inflectional distinctiveness in the Old English paradigms, he found that even in the ninth century, inflections could not have distinguished the subject and object in the subject-object pattern more than 41 per cent of the time and that if there had to be distinction between subject and object, there must have been some other signaling devices to mark the distinction 59 per cent of the time. Furthermore, he showed that in those patterns where case inflections did not distinguish subject and object, the subject stood before the object 94 per cent of the time.24 All these studies, which are quantitative in that the evidence is based on frequency counts, point to the fact that word order was functional in early stages of the English language. It remains now to look at the studies which are more closely similar to the present one. The first of these is Charles R. Carlton's study of the syntax of a carefully selected group of original charters written from A.D. 805 to 1066.25 Adapting to the study of Old English the method of Charles C. Fries' Structure of English, Carlton divided his corpus into basic syntactic units (sentences) as identified by formal and structural characteristics. Sentences were divided then into their functional parts. Once the primary and secondary elements were discussed and illustrated, Carlton moved on to the investigation of inflection and word order as grammatical devices. He concluded that inflection was the primary syntactic signal but because of the tendency of elements toward a certain pattern or patterns, that word order was also a functional syntactic signal. Ann Shannon's study presented a descriptive syntax of a portion of the Parker Manuscript of the Anglo-Saxon Chronicle. Instead of a sentence-level analysis, Shannon used a clause-level analysis. She described the clause (which she believes is the most convenient unit to employ in the analysis of Old English) as characterized by having only one essential element, the verb. In her analysis, then, there were as many 23 Mildred K, Magers, "The Development of the Grammatical Use of Word Order for Relationships Expressed by the Accusative with Special Reference to the Development in Subordinate Clauses", unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1943, pp. 44 and 89. M Saitz, op. cit., pp. 106-126. 86 Carlton, op. cit.

INTRODUCTION

21

clauses as finite verbs. She treated the order of major and minor elements, the combination of clauses, and morphosyntactic relations. One of her most important conclusions was that subjects and objects were not always distinctively marked by case endings and that morphological and syntactic signaling in many examples seemed to be redundant.26 William H. Brown, Jr., used an approach similar to that of Shannon in his study of King Alfred's Pastoral Care. He worked with a 22,000 word corpus from the Hatton Manuscript. The focal point of Brown's analysis was also the clause, and he also assumed that there were as many clauses as finite verbs. He recognized that nominal and verbal phrases had clearly established patterns and that the order of the subject, verb, and object fell into consistent, limited patterns, despite the fact that they occurred in all possible positions.27 Robert A. Palmatier described the syntax of the Ormulum, an early Middle English text in the East Midland dialect, dated about 1200. Palmatier, like Shannon and Brown, used clause-level analysis and assumed there were as many clauses as finite verbs. He concluded that syntactic relationships were signaled by inflectional agreement, order of occurrence, and semantic association, that syntactic patterns such as the noun-headed phrase were essentially like those of Modern English, and that the subject-verb, subject-verb-object, subject-verb-indirect object-direct object, and subject-verb-complement patterns were normal.28 These studies, all in all, show that early English did not depend entirely upon one kind of system for indicating relationships and that many of the syntactic patterns characteristic of Modern English have appeared and been operative in the earlier stages of the English language. Furthermore, it is evident that a descriptive linguistic method can be effectively applied to early stages of a language; scholars are now well on their way to providing the series of synchronic studies which are needed in giving an account of the nature and growth of the English language from generation to generation.

1.5. PREVIOUS STUDIES OF THE PETERBOROUGH CHRONICLE

The interest in the language of the latter part of the Peterborough Chronicle has been exceptional. The earlier description of the corpus has explained why this portion of the Chronicle has special attraction and importance for the students of the English language. However, much of the study of the document has dealt with morphology

26

Ann Shannon, A Descriptive Syntax of the Parker Manuscript of the Anglo-Saxon Chronicle from 734 to 891 (The Hague, Mouton and Company, 1964), p. 67. 27 William H. Brown, Jr., A Descriptive Syntax of King Alfred's "Pastoral Care" (The Hague, Mouton, 1970). 28 Robert A. Palmatier, A Descriptive Syntax of the "Ormulum" (The Hague, Mouton, 1970).

22

INTRODUCTION

and vocabulary. The earlier studies centered on the nature of the inflectional system of the language. The first was O. P. Behm's The Language of the Latter Part of the Peterborough Chronicle.29 Another similar study of the inflectional system was Heinrich Meyer's Zur Sprache der jüngeren Teile der Chronik von Peterborough.30 Neither of these went beyond morphology; they focused on what they generally described as a decaying and arbitrary inflectional system. In other studies, portions of the Peterborough Chronicle have been used as samples of twelfth-century English in diachronic studies which have given a historical account of a single form of expression or of limited patterns. Examples are found in the dissertations of Thomas and Cassidy, both of which were mentioned earlier. More recent investigations have been concerned with the vocabulary and certain grammatical forms relating to gender and the relative pronouns. Cecily Clark has provided a relatively complete study of the vocabulary31 and Angus Mcintosh published a short paper on the relatives jbe and pat in early Middle English.32 There are three further studies which have treated some aspects of syntax of the latter part of the Peterborough Chronicle. One is Wilhelm Roth's Die Wortstellung im Aussage-Hauptsatz Angelsächsischen Originalprosa, Annalen 800-990, 1066-1154.3S Roth's corpus included a very early portion of the Anglo-Saxon Chronicle and the latest, the portion under investigation here; however, his investigation seems to be mainly concerned with the relationship of verbal modifiers to the other parts of the sentence. The other study, an investigation of the word order of the complete Peterborough Chronicle, is Ewald Rothstein's "Die Wortstellung in der Peterborough Chronik". 34 Rothstein's study is more useful to the student of the language of the Peterborough Chronicle, yet it is far from being a descriptive syntax. Rothstein was preoccupied with the movement of the elements to the position they occupy in Modern English. The study noted also the influence of verbal modifiers on the order of the subject and verb. Both of these latter studies are somewhat dated. Rothstein's would be of greater use if he had drawn generalizations separately from the portion up to 1121 and the portion from 1122 to 1131. The third is Bruce Mitchell's article35 29

O. P. Behm, The Language of the Latter Part of the Peterborough Chronicle (Uppsala, [n.n.], 1884). Heinrich Meyer, Zur Sprache der jüngeren Teile der Chronik von Peterborough (Leipzig, Buchhandlung Gustav Fock, 1889). 31 Cecily Clark, "Studies in the Vocabulary of the Peterborough Chronicle, 1070-1154", English and Germanic Studies, V (1952-53), 67-89. 32 Angus Mcintosh, "The Relative Pronouns pe and pat in Early Middle English", English and Germanic Studies, I (1947-48), 73-90. 33 Wilhelm Roth, Die Wortstellung im Aussage-Hauptsatz Angelsächsischen Originalprosa, Annalen 800-900, 1066-1154 (Berlin, Mayor and Miller, 1914). 34 Ewald Rothstein, "Die Wortstellung in der Peterborough Chronik", Studien zur englischen Philologie, LXIV (1922), 1-108. 35 Bruce Mitchell, "Syntax and Word-Order in The Peterborough Chronicle 1122-1154", Neuphilologische Mitteilungen, LXV (1964), 113-144. What similarity there is between the conclusions of Mitchell's study and this one was not due to the investigator's use of Mitchell's study. In fact, the investigator, although he searched diligently for, and wrote to experts in the field about, such works, did not discover Mitchell's work until he had almost finished his study. 30

INTRODUCTION

23

"Syntax and Word-Order in The Peterborough Chronicle 1122-1154". Mitchell set out to show what was modern and what was archaic in the language of the latter part of the Peterborough Chronicle and concluded that the language was changing to Modern English, but that it was not as far advanced as earlier scholars had maintained. Thus far, Mitchell's study is the most accurate description of this corpus, and hence, the most useful; however, it lacks the detail and coverage of the present study. The present study made only limited use of these three works on word order and goes beyond them in its description of the syntax of the portion of the Chronicle studied. 1.6. EDITION OF THE TEXT

The basic edition used in the study was Charles Plummer's revison of John Earle's edition, Two of the Saxon Chronicles Parallel. Scholars seem to be unanimous in thinking that any linguistic account of the Peterborough Chronicle must be based on Plummer's revision of Earle.36 Plummer reproduced the manuscript very accurately, retaining most of its peculiarities. Even so, the investigator compared it with other editions and constantly referred to a facsimile reproduction of the original manuscript. In illustrating the descriptive statements in the analysis with examples from the corpus, all the features of Plummer's edition are kept, except that the following abbreviations in the original have been expanded to their full form: a66 aßßrice jerceb 6, 6isc Ihc S\ see b δ

abbot abbotrice aercebiscop biscop Iesus sancte t>et, t>at öet, öat 1.7. GENERAL PROCEDURE OF THE ANALYSIS

The general procedure of analysis used in this study is derived from descriptive techniques of modern linguistics. The investigator did not, in the analysis of the corpus, follow the method of any single linguist but rather adapted some of the ideas which were gathered from several books and articles.37 The presentation of the data, as well M

See, for example, Garmonsway, op. cit., p. xxxiii. " Benjamin Elson and Velma Pickett, An Introduction to Morphology and Syntax (Santa Ana, California, Summer Institute of Linguistics, 1964); W. Nelson Francis, The Structure of American English (New York, The Ronald Press Company, 1958); Charles C. Fries, The Structure of English (New York, Harcourt, Brace and Company, 1952); H. A. Gleason, An Introduction to Descriptive Linguistics (New York, Henry Holt and Company, 1955); Zellig Harris, Methods in Structural Linguistics (Chicago, The University of Chicago Press, 1951); idem, String Analysis of Sentence Structures (The Hague, Mouton and Company, 1962); Archibald Hill, An Introduction to Linguistic

24

INTRODUCTION

as the descriptive terminology, for the most part, follows the suggestions of Longacre and Elson and Pickett (see footnote 37) more closely than those of other scholars. The present investigation has as its central concern the syntax of a language that is no longer spoken. Basically, its aim is to determine how words are put together and how relationships are signaled in the language of the latter part of the Peterborough Chronicle. The language under investigation here is an earlier stage in the successive stages of the history of the English language; in many respects, it has its own distinct syntax. It is readily apparent that some adjustments must be made in the procedure of investigation of a language that is no longer recordable from a native informant. Since the present corpus is a written text from an earlier stage of the English language, and since there are no native informants, the analyst in interpreting the corpus had to rely on his own knowledge of the language, on the knowledge of the best scholars of the language, on the other available studies such as Behm's The Language of the Latter Part of the Peterborough Chronicle and Meyer's Zur Sprache der jüngeren Teile der Chronik von Peterborough, and finally, on the standard editions and their glossaries, dictionaries, and translations.38 The order of procedure of the analysis will now be explained. First of all, after the text was selected, several preliminary steps were required in preparation for the analysis of the corpus. It was necessary for the analyst to get a workable knowledge of the language by rereading the corpus closely, by studying the orthographic devices, and by studying carefully the available works on the language of the corpus. Other than the scholarly works mentioned in the section on previous studies of the Peterborough Chronicle, there are no grammars based specifically on the language of this period. For this reason, works describing the stages of English as close as possible to the language under investigation were read for any insights they might yield. In addition to several Old English grammars, the list included Fernand Mossi's A Handbook of Middle English·, Karl Brunner's Outline of Middle English Grammar·, Walerian Swieczkowski's Word Order Patterning in Middle English·, Samuel Moore and Albert Marckwardt's Historical Outlines of English Sounds and Inflections and Robert Palmatier's "A Descriptive Syntax of the Ormulum", a work mentioned earlier. Glossaries, notes, and discussions of the several editions as well as the Middle English Dictionary were also used.39 Structures (New York, Harcourt, Brace and Company, 1958); John P. Hughes, The Science of Language (New York, Random House, 1962); and Robert E. Longacre, "String Constituent Analysis", Language, XXXVI (1960), 63-88. i8 Some recent investigators have, it seems, prided themselves on not using translations as aids. It is within good linguistic practice to use translations as aids as long as the analyst remembers that he is analyzing the language under investigation and not the English translation. For an opinion about this from the scholarly tradition see Etsko Kruisinga, op. cit., p. 45; and from the modern linguistic tradition, see Elson and Pickett, op. cit., p. 66. " Fernand Mosse, A Handbook of Middle English (Baltimore, The Johns Hopkins Press, 1952); Karl Brunner, Outline of Middle English Grammar (Cambridge, Harvard University Press, 1963);

INTRODUCTION

25

A more thorough knowledge of general syntax was gained by studying works of the modern linguistic and earlier scholarly traditions. Works of modern linguists have already been listed. Works from the scholarly tradition included, to name just three, Jespersen's Essentials of English Grammar, Bryant's A Functional English Grammar, and Reinard Zandvoort's A Handbook of English Grammar.*0 Since the East Midland dialect can be considered a stage in the successive stages of the English language, the syntax of the investigator's own language was studied in detail from the viewpoints of both traditions; and because the modern stage of English is not as inflectional as the one under investigation, the inflectional and syntactical systems of Modern German were reviewed. After the preliminary steps were completed, a facsimile reproduction of the original manuscript was scrutinized to see if the graphic devices could be used to determine a structural unit smaller than the annal entries, of which there were sixteen in the corpus. Although it is diflicult to read these graphic devices as marking off what is known today as the sentence,41 they do give the reader the impression that the annal entry is not just a long continuum of words loosely joined together. The investigator felt that there was a sound enough basis for warranting his search for a provisional unit into which he could further segment the corpus and begin the analysis. Because of the nature of the text and the scribe's tendency to string groups of words together by y, the unit used in this investigation was defined as the smallest terminable syntactic unit, which will be referred to from here on as the T-unit.42 The T-unit is any segment that is grammatically isolatable without leaving residue. It, then, is a single independent predication, with complements or modifiers if any, and with any dependent or included predications which may be embedded in or attached to it. In this corpus, a T-unit may be simple or complex. If it has only a single independent predication without any dependent or included clauses embedded or attached, it is a simple T-unit. If a single independent predication has one or more dependent or included clauses embedded or attached, it is a complex T-unit. Either the simple or the complex unit may appear with or without an expressed Walerian Swieczkowski, Word Order Patterning in Middle English (The Hague, Mouton and Company, J 962); Samuel Moore and Albert H. Marckwardt, Historical Outlines of English Sounds and Inflections (Ann Arbor, George Wahr Publishing Co., 1960); and Hans Kurath and Sherman Kuhn (eds.), Middle English Dictionary (Ann Arbor, University of Michigan Press, 1952-). 40 Otto Jespersen, Essentials of English Grammär (New York, Henry Holt, 1933), reprinted, University of Alabama Press, 1964; Margaret M. Bryant, A Functional English Grammar (Boston, D. C. Heath and Company, 1959); and R. W. Zandvoort, A Handbook of English Grammar, ninth edition (Groningen, J. B. Wolters, 1964). 41 Whether these units are close to the concept of the sentence of Modern English (which, incidentally, is quite elusive if we consider the over two hundred definitions Fries alludes to in The Structure of English and those formulated since then) is hard to say. 42 The concept of the T-unit was borrowed from Kellog Hunt, who used it as a means of segmenting student compositions. See Kellog W. Hunt, Differences in Grammatical Structures Written at Three Grade Levels, the Structures to be Analyzed by Transformational Methods. Report to the U.S. Office of Education, Cooperative Research Project No. 1998 (Tallahassee, Florida, 1964).

26

INTRODUCTION

subject of the verb. An example of the simple T-unit with an expressed subject (subjectful) is as follows: On J)is geare waes se King Heanri on Cristes maessan on North wie. 'In this year King Henry was at Norwich on Christmas.' (250.1-2)48 An example of the subjectful complex T-unit is as follows: l>a hwile J>e jia munecas sungen J>aere messe, y se daecne hafde ongunnan {)one godspel Preteriens Iesus. J>a com se fir on ufen weard fione stepel. 'While the monks were singing the mass and the deacon had begun the gospel Preteriens Jesus, the fire came to the upper part of the tower.' (250.4-6) Examples of the simple and complex T-units without expressed subjects (subjectless) are as follows: y fylden {» land ful of castles. 'And (they) filled the land full of castles.' (264.2-3) y wuröon t>a alswa gode freond swa hi waeron aeror feond. 'And (they) became then just as good friends as they had before been enemies.' (259.29-30) These T-units, excluding the subjectless, resemble the simple and complex sentences of Modern English. As in Modern English, T-units can be and are joined by coordinating conjunctions to make compound structures, that is, two independent predications. This analysis, however, does not concern itself with structural units above the T-unit. In short, it does not recognize what is referred to in Modern English as COMPOUND SENTENCES. Nothing is lost, however, because this study is primarly concerned with T-unit-level, clause-level, and phrase-level analyses, specifically with how words, phrases, and clauses fill and pattern in these structural units. The T-unit was used in the analysis in the first place because of the scribe's tendency to string units together by the symbol 7. The possibility of compound T-units will be briefly mentioned in the discussion of the joining and relationship of clauses in Chapter 6. Once the T-units were segmented, they were then recorded individually on separate analysis worksheets. Before the T-units were broken down into their component parts, each was translated, first literally and then as idiomatically as possible. The purpose of the double translation was to reveal the differences of syntactic grouping in the two stages of the English language and to allow the analyst to see the distinctiveness of the syntax of the language under investigation. The next step was to break up the T-units into their clause-level slots, noting linear order and the classes of fillers and other relevant grammatical relationships, for example, concord, government, and case.44 If the T-unit was complex, the independent 49

Numbers are page and line references to Plummer's edition. The First Continuation runs from 250.1 to 262.19, and the Final Continuation, from 262.20 to 269.4. " The presentation of the data and the relevant terminology used are derived from the tagmemic method of description, a filler-and-slot technique, which will be described in the next section.

INTRODUCTION

27

predication was first analyzed, permitting the analyst to detect any function the dependent predication might have as part of the T-unit. After the T-unit level and clause-level analyses were carried out, attention was then focused on the structures of modification. The phrase rank and order of the modifying slots and their fillers were determined and classified. Then the forms of individual words were examined to determine the relations between morphology and syntax.45 And finally, the investigation was focused on the joining and relationships of clauses in T-units. During this process of analysis and classification, the frequency of occurrence and percentages of the linguistic units at the various levels were computed; later, the counts were made into tables showing frequency of occurrence as well as other relationships. The procedure then was to take the analysis from the whole to the part, beginning with the whole corpus, segmenting it into T-units, and analyzing them into primary and secondary functional slots for the purpose of determining syntactic patterns.

1.8. EXPLANATION OF BASIC CONCEPTS USED I N THE DESCRIPTION

It was noted earlier that the investigator in gathering, analyzing, recording, and reporting the significant facts of the corpus was guided by ideas derived from a number of linguists. Relatively speaking, however, the investigator followed more closely the tagmemic model exhibited in Benjamin Elson and Velma Pickett's An Introduction to Morphology and Syntax.*6 Let it be understood at the outset that he did not intend to use this model of analysis and description in all its detail and to follow it precisely from start to finish, but rather he sought to adapt some of the fundamental concepts which he believed would be useful in the description and presentation of the significant facts of the syntax of the Peterborough Chronicle. Elson and Pickett, after Pike and Longacre, advocate an approach to the analysis and description of language that does not look at the sentence as consisting of only two parts, a subject and a predicate, but of a string of functional parts; and they emphasize "the function of members in relation to each other in utterances". 47 They posit, also 44

Using a filler-and-slot method of analysis necessarily requires some attention to morphology in almost every step of the analysis; however, this study in its basic aspects is beyond morphology. " This work, already cited in footnote 36, owes much to Kenneth L. Pike's Language in Relation to a Unified Theory of the Structure of Human Behavior, Parts I, II, III (Glendale, Summer Institute of Linguistics, 1954-1960) and Robert E. Longacre's "String Constituent Analysis", Language, XXXVI (1960), 63-88. Works illustrating the use of the tagmemic model are Olive Shell, "Cashilio II", International Journal of American Linguistics, XXIII (1957), 179-218; and H. Hart, "Hierarchical Structuring of Amuzgo Grammar", International Journal of American Linguistics, XXIII (1957), 141-164. 47 Elson and Pickett, op. cit., p. 3.

28

INTRODUCTION

after Pike and Longacre, a unit of grammar called the TAGMEME, which has two aspects, a slot and a class filler. Earlier, Pike had defined the tagmeme as "a minimum functional segment of a sentence type in which the sentence type is divided according to grammatically functional parts...". 48 In the same article he spoke of the functional segment as a "functional slot with filler".49 Longacre, a few years later, wrote, "The tagmeme, thus defined [by Pike] is a substitution point along with the class found at that point", 50 and went on to add, "A substitution point [slot] does not exist apart from the occurrence of some linguistic item or sequence belonging to some class [filler].51 The important thing to keep in mind here is that the tagmeme must be understood in terms of both a SLOT and the FILLER of that slot. Neither can be said to exist without the other. For example, in the clause o s ν pet hi ealle abohton. 'They all paid for that.' (255.16) there are three slots (in which a variety of substitutions are possible): subject, direct object, and predicator (tagmemes are referred to by the name given to the slot). In this example, the subject slot (S) is filled by hi ealle; the direct object slot (O), by pet; and the predicator slot (V), by abohton. Basic to the understanding of the tagmeme is the recognition that 'slot' refers to GRAMMATICAL FUNCTION, not to linear position. A tagmeme may occur in different linear positions. For instance, consider these two examples: C S V Μ ο (1) y hi sxden openlice dat Xpist slep ... 'And they said openly that Christ slept ...' ~(2653) ~ c

s

ο

ν

Μ

(2) y se king hine underfeng mid micel wurtscipe. 'And the king received him with great honor.' (255.19) The direct object tagmeme in example 1 (slot filled by the nominal clause dat Xpist slep) stands last in the linear ordering of the tagmemes, whereas in example 2 it (slot filled by the personal pronoun hine) stands third in order, after the coordinating tagmeme y and the subject tagmeme. Notice, furthermore, that in example 1, the predicator tagmeme (slot filled by the verb sseden) and the manner tagmeme (slot filled by the adverb openlice) do not occur in the same positions as their counterparts do in example 2 SLOT THEN REFERS PRIMARILY TO FUNCTION. The tagmemes, of course, do occur, as the examples above obviously show, in a linear order, or as Longacre would say, in a 'structured matrix', which he calls a HYPERTAGMEME and defines as a "grammatical string [a sequence of words or word 48

Kenneth L. Pike, "Grammemic Theory", General Linguistics, II (1957), 35. *' Ibid. 50 Longacre, op. cit., p. 63. 81 Ibid.

INTRODUCTION

29

groups] with identifying-contrastive features that structurally distinguish it from other grammatical strings on the same levels".52 Since tagmemes, as defined above, are spoken of in reference to hypertagmemes, description can take place at various significant grammatical levels. Though there are others, the levels dealt with, for the most part, in this study are the clause level and phrase level and to a lesser degree, the T-unit level. Hence, it is proper to speak of CLAUSE-LEVEL TAGMEMES, PHRASE-LEVEL TAGMEMES, a n d T-UNIT-LEVEL TAGMEMES. All

the slots recognized in the previous examples were discovered at the clause level and thus may be considered clause-level tagmemes. Example 2 was treated as a string of four tagmemes. The tagmemes, in the order of their occurrence, were the subject tagmeme (S), the direct object tagmeme (O), the predicator tagmeme (V), and the manner tagmeme (M). Also spoken of in this study are phrase-level tagmemes. In the following example y te Lundenissce folc hire wolde taecen. 'And the people of London wanted to capture her.' (266.30) the modified-noun phrase, te Lundenissce folc, which fills the subject slot of the clause, is considered to manifest a clause-level subject tagmeme. "A morpheme sequence [here, words or word groups] which fills a grammatical slot is said to manifest a tagmeme at the level where it is discoverable."53 A corollary to this statement is that if a word or word-group sequence is recognized at one level, then it must be analyzed at another level. That is to say, since the modified-noun phrase te Lundenissce folc was recognized at clause-level, it will be analyzed at phrase-level. Analyzing the modified-noun phrase te Lundenissce folc shows that it consists of three words with folc serving as the nucleus (head slot) of the phrase and te and Lundenissce serving as satellites (modifying slots) of the phrase. At this level, then, are also functional slots with fillers or tagmemes. Whereas head slots are stable, the modifying slots of a modified-noun phrase admit of great variety. In the example above, there are the head tagmeme and two modifier tagmemes, one of which is a determiner filled by te and the other, a qualifier, filled by Lundenissce. Although this study is concerned to a large degree with clause-level tagmemes (Chapters 2, 3, 4) and phrase-level tagmemes (Chapter 5), reference is made to Tunit-level tagmemes in Chapter 6. Since T-units may be simple (having only one independent predication) or complex (having one independent predication and one or more dependent predications), they are said to have a main slot fillable by an independent clause and sometimes an included slot fillable by dependent clauses. Not all included slots, however, are discoverable or recognized at the T-unit level. The following three examples show how dependent clauses manifest slots at three different levels: " 53

Ibid., p. 64. Elson and Pickett, op. cit., p. 59.

30

INTRODUCTION

(1) for agenes him risen sona J>a riceman pe wseron swikes. 'For against him rose the powerful men who were traitors.' (263.17-18) (2) Air he waere wel ded. J>a waere {)ser coren twa papes. 'Before he was dead long, two popes were elected.' (260.17-18) (3) Gif twa men öfter Hi coman ridend to an tun. al f)e tunscipe flugaen for heom. 'If two or three men came riding to a village, all the villagers fled from them.' (265.3-4) , In the first example, pe waeron swikes, an adjectival clause, does not operate at either T-unit or clause level, but at phrase level, for it fills a post-positional modifying slot, the head of which is riceman. In the second example, xr he waere wel ded, an adverbial clause, fills a clause-level slot, the slot of the time tagmeme. It is important here to notice that both these clauses are replaceable by words or phrases. In the third example, gif twa men oper iii coman ridend to an tun, an adverbial clause, does not fill either a phrase-level or clause-level slot and is not replaceable by a word or phrase; it is said to manifest the included or dependent predication slot at the T-unit level. In the analysis of the clause-level tagmemes a distinction, for convenience of description and presentation, was made between those tagmemes which the investigator thought were obligatory to the construction and those that appeared to be optional to the construction. The first group, which included the subject, predicator, direct object, indirect object, and subject complement, were called PRIMARY TAGMEMES; the second group, which consisted of the various adverbials, were called SECONDARY TAGMEMES. The investigator is well aware of the dangers inherent in making such distinctions in the constituents of a clause of a language no longer spoken. These notions of primary and secondary are notions influenced, of course, by a knowledge of the structure of Modern English. The investigator, furthermore, is also aware of the problems of making such a division in Modern English; for in some constructions, adverbials are just as obligatory to a construction as a subject or predicator. Be that as it may, the distinction is made here solely for the purpose of showing which tagmemes, as a rule, recur in clauses as nuclear slots and those which occur in clauses as peripheral slots, that is, those which appear to be more necessary than others in given clause patterns. In this description and presentation, then, the emphasis is on meaningful grouping and functioning at different levels.

1.9. THE PRESENTATION OF THE DATA

The presentation of the data will begin, in Chapter 2, with the identification and description of the primary clause-level tagmemes and a brief discussion of some of the more important morphosyntactic relations.. Chapter 3 continues the discussion of primary tagmemes by focusing on their linear ordering. Chapter 4 treats the

INTRODUCTION

31

secondary clause-level tagmemes, with emphasis on their variety of functions and their positions relative to each other and with respect to the predicator. Chapter 5 treats the description of the modified-head construction. Chapter 6 presents a description of the relationships and joining of clauses. The last chapter, Chapter 7, summarizes the findings and presents the general conclusions of the study. Each chapter dealing with the findings of the study will include a statement of objectives, a brief description of method, and a statement of results.

2 PRIMARY CLAUSE-LEVEL TAGMEMES

In this corpus there were five primary clause-level tagmemes: (1) the subject tagmeme, (2) the predicator tagmeme, (3) the direct object tagmeme, (4) the indirect object tagmeme, and (5) the subject complement tagmeme. The predicator, it was explained earlier, was enough to constitute a clause and was, therefore, the simplest form of predication. Yet most of the clauses were predications consisting of at least a subject tagmeme and a predicator tagmeme. Very rarely did these two tagmemes occur alone; they were usually accompanied by one or more other clause-level tagmemes such as the three other tagmemes listed above and those discussed in Chapter 4. Since the next chapter will treat the word order of these clause-level tagmemes, very little will be said about it here. When linear ordering is mentioned, it will be in reference to the ability of case inflection to signal related meanings and syntactic functions and distinctions. The purpose of this chapter is to isolate the primary clause-level functional slots, to identify the form classes or larger structural units which fill them, to tabulate their frequency, to illustrate them with varied examples from the corpus, and finally, to discuss some of their fundamental features and grammatical relationships.

2.1. THE SUBJECT TAGMEME

The subject is the first primary clause-level tagmeme that will be considered here. Although it occurred in almost every possible position relative to other tagmemes in the clause, it had a relatively fixed position with respect to the predicator tagmeme. That is, the subject occurred in a clause where there was a predicator; and when it did, it more frequently stood before the predicator than after it. In this corpus, the filler of the subject slot may be singular or plural, simple or compound. As a rule, it agreed inflectionally with the number of the filler of the predicator slot and was in the nominative case. Of the 763 clauses in the corpus (484 independent and 279 dependent), 702 (92 per cent) appeared with an overt subject. The subject slot may be filled by a pronoun or

PRIMARY CLAUSE-LEVEL TAGMEMES

33

1

modified-pronoun phrase, a noun or modified-noun phrase, the pronominal adjective or modified-pronominal adjective phrase,2 and the nominal clause. 2.1.1. The Pronoun or Modified-Pronoun Phrase The most common filler of the subject slot was the pronoun or modified-pronoun phrase. Three hundred eighty-three clauses (55 per cent) occurred with pronouns or modified-pronoun phrases in the subject slot. Of these, only nine (2 per cent) were modified-pronoun phrases. One hundred ninety (50 per cent) of this group occurred in independent clauses and 193 (50 per cent) in dependent clauses. Examples of the pronoun and modified-pronoun phrase as fillers of the subject slot are presented below: PRONOUN:

(1) I>a tidde hit on an Wodnesdei ... f>et se king rad in his derfald .... 'Then it happened on a Wednesday ... that the king rode in his deer park ....' (251.4-5) (2) Bis waes eall ear gedon Öurh se biscop of Seresbyrig ... aer he waere dead. 'This was all done before by the bishop of Salisbury ... before he died.' (251.25-27) (3) E>aet fir hi seagon in öe daei rime. 'They saw the fire at daybreak.' (250.22) (4) I>is bebaed se asrcebiscop Willelm of Cantwarabyrig y ealle t>a leod biscopes da pa waeron on Englalande. 'This was ordered by William, archbishop of Canterbury, and all the other diocesan bishops who were in England.' (260.5-7) (5) Dauid his broöer pa waes eorl on Noröhamtune scire feng to rice .... 'David, his brother, who was earl in Northhampton, rose to power ....' (254.23-24) (6) y me laet hire dun on niht of J»e tur mid rapes. 'And one let her down from the tower at night with ropes.' (267.22-23) (7) y sex fleh y isde on fote to Walingford. 'And she fled and went on foot to Wallingford.' (267.23-24) MODIFIED-PRONOUN PHRASE:

(1) mid him helden da of Rome, y se due of Sicilie. 'With him held those of Rome and the duke of Sicily.' (260.20) (2) £>et hi ealle abohton. 'They all paid for that.'(255.16) (3) ... £>et her nouper sculde besuiken other. '... that neither of them should betray the other.' (267.10-11) (4) y hi of Normandi wenden alle fra Jje king. 'And they of Normandy all turned from from the king.' (267.25) (5) Alle he waeron forsworen y here treothes forloren. 'All of them were forsworn and their agreements destroyed.' (264.1) 1

Nominal adjectives are included in this group. This group includes such words as ealle, sume, nati, fela, xlc, an, oder, bade, etc. because they fill both nominal and adjectival slots. 8

34

PRIMARY CLAUSE-LEVEL TAGMEMES

2.1.2. The Noun or Modified-Noun Phrase Of the 702 subject slots, 305 (43 per cent) were filled by nouns or modified-noun phrases. Two hundred thirty-eight (34 per cent) of these occurred in independent clauses and sixty-seven (66 per cent) in dependent clauses. The noun and modified-noun phrase may be simple or compound. Some typical examples are as follows: NOUN:

(1) ... swa drane doö on hiue. '... as drones do in a hive.' (258.11) (2) y Waleram belaf mid |>one kyng. 'And Waleran remained with the king.' (259. 26-27) (3) oc Xpist it ne uuolde. 'But Christ did not desire it.' (262.27) (4) God geare his sawle. 'God save his soul.' (258.34) MODIFIED-NOUN PHRASE:

(1) Da sone l?£er after sende se kyng hise write ofer eall Englalande. 'Then soon after that the king sent his writs all over England.' (251.14-15) (2) oc ferden pe zrcebiscop y te wisemen betwux heom. 'But the archbishop and the wise men went between them.' (268.9) (3) On J)a ilea tyma ferden pes eorles sander men mid unsachte fram kyng. 'At the same time the count's messengers went from the king without agreement.' (252.5-6) (4) y weax }>a micel unfrid betwux him y hise J)eignas .... 'And then great strife grew between him and his thanes ....' (253.9) (5) 1 £>es ilees geares ferde se forensprecene abbot Henri ham to his agen minstre to Peitou be t>es kynges leue. 'And this same year the aforementioned abbot Henry went home to his own monastery in Poitou by the King's permission.' (259.3-4) (6) y te Lundeniscefolc him underfeng .... 'And the London people received him ....' (263.14) 2.1.3. The Nominal Clause In this corpus there were only four clauses (less than 1 per cent) which filled the subject slot, one of which was an impersonal. Not counted here were clauses such as: ΡΆ gelamp hit ... Jjet se eorl Waleram of Mellant ferde fram his an castel Belmunt het ....' 'Then it happened ... that the count of Meulan, Waleran, went from one castle of his called Beaumont ....' (253.24-25) It would not be going too far astray to consider hit merely as an expletive and the clause as the true subject. Several of these constructions occurred in the corpus. Be that as it may, examples of the clause regarded as filling the normal subject slot are as follows:

PRIMARY CLAUSE-LEVEL TAGMEMES

35

(1) Pe nan ne heafde staerf of hungor. 'He who had nothing died of hunger.' (255.5) (2) y se pe pet ne wolden done forgede his circe .... 'He who would not do that should give up his church ....' (260.3-4) (3) nu him behofedpet he crape in his mycele codde in xlc hyrne .... 'That he should creep into every corner of his big wallet was now required of h i m — ' (262.5) 2.1.4. The Pronominal Adjective or Modified-Pronominal Adjective Phrase There were twelve subject slots (2 per cent) which were filled by pronominal adjectives. Only one was in a dependent clause. Six different examples are presented here: (1) ac iedon ealle samodlice to Jjone kyng .... 'But they all went together to the king....' (251.22) (2) ...felaoöre wendan fram him .... 'Many others turned from him ....'(253.11-12) (3) Se an was gehaten Petrus. 'One was named Peter.' (260.18) (4) Se oder het Gregorius. 'The other was named Gregory.' (260.21) (5) for nan ne waes ο J>e land. 'For none was in the land.' (264.32) (6) sume flugen ut of lande. 'Some fled out of the land.' (264.34) 2.1.5. The Double Subject There were two clauses with what may be called a double (that is, redundant, in contrast to coordinated) subject: (1) oc ure Laford God selmihtig pa eall digelnesse sed y wat. he seoö fet man laet Jjet aerme folc mid ealle unrihte. 'But our Lord God Almighty who sees and knows all secrets, He sees that they treat the poor people with complete injustice.' (255.1-2) (2) ... se man de hafde an pund he ne mihte cysten aenne peni at anne market, '...the man who had a pound, he could not get the value of a penny at a market.' (255.9-10) 2.1.6. The Case of the Filler of the Subject Slot Some scholars (Jespersen is the first who comes to mind) feel that a morphological system can be considered completely reliable only if an inflectional ending is distinctive in a word in isolation. Though earlier Old English was highly inflectional, recent studies (for example, those of Saitz, Carlton, and Shannon which were mentioned in the Introduction of this study) have shown that nominative and accusative inflections of the form classes could not have distinguished subject and object in all subject + predicator + object patterns. Even in examples where the demonstrative pronoun and adjective, whose inflections more distinctively marked function, were attributes of the noun, this sometimes was true. The confusion of case was so great

36

PRIMARY CLAUSE-LEVEL TAGMEMES

in this portion of the Peterborough Chronicle that the pronoun must primarily be considered if anything is to be said with confidence about the case of the filler of the subject slot in this corpus. In the First Continuation, it is true that the nominative form se of the demonstrative pronoun was the regular form for the subject, but se occurred also in the oblique cases and in the Final Continuation gave way to pe, or its variants te, the, all of which were invariable, regardless of function. A few examples from the First Continuation will make this clear: (1) Dis waes eall ear gedon öurh se biscop of Seresbyrig. 'This had all been done before by the bishop of Salisbury.' (251.25-26) (2) y se eorl 7 ealle J>a heafed menn ... flemden se oöer abbot Heanri ut of l>a mynstre. 'And the count and all the chief men ... chased the other abbot, Henry, out of the monastery.' (262.1-2) (3) pe erthe ne bar nan corn. 'The earth bore no corn.' (265.8) In the first example se is used with a noun in the object slot of the preposition; in the second, with a noun in the direct object slot. The filler of the subject slot, fie, in the third example (from the Final Continuation, which incidentally contained only one occurrence ο he) was invariable for all numbers, cases, and genders. If it were not for word order and lexical association, it would be hard to say whether the subject was fie erthe or nan com. In one clause, even the accusative form fione was used with a noun which filled the subject slot: Pus earmlice waes pone abbotrice gifen betwix Cristesmesse y Candelmesse at Lundene. 'Thus wretchedly was the abbacy given between Christmas and Candlemas in London.' (258.8-10) However, this was a single occurrence, and it could represent a writer's mistake. Yet se and pe often occurred where early Old English would have had more distinctive forms for cases and numbers. Using se and the subject forms of the personal pronoun {he, ic, sex, we, and fiu) as a basis for establishing the case of the filler of the subject slot in the First Continuation and only the personal pronoun in the Final Continuation, it can be said that the filler of the subject slot occurred in the nominative case. Table I presents an analysis of the inflectional distinctiveness of fillers of the subject slot. The fillers marked 'ambiguous' (pronouns, nouns, pronominal adjectives) consist also of phrases in which these form classes function as heads. The table indicates that the filler of the subject slot was distinctively marked (nominative case) in about 49 per cent of the total number of occurrences and was non-distinctively marked in about 51 per cent of the examples. (The percentages in these and the subsequent tables are to the nearest whole per cent.)

PRIMARY CLAUSE-LEVEL TAGMEMES

37

TABLE I

Inflectional Distinctiveness in Fillers of Subject Slots

Nom. Pron. Noun w/Nom.Attrib. Noun w/Acc.Attrib. Pronouns (Ambig.) Nouns (Ambig.) Pron. Adj. (Ambig.) Clause

First Continuation No. %

Final Continuation No. %

Nr>

/o

122 120 1 94 86 7 4

96 7 0 73 88 4 0

36 3 0 27 33 1 0

218 127 1 167 174 11 4

31 18 -1 24 25 2 1

434

28 28 -1 22 20 2 1

268

Total

702

2.1.7. Agreement with the Filler of the Predicator Slot As a rule, the filler of the subject slot showed grammatical concord with the filler of the predicator slot. Whenever the filler of the subject slot was singular or plural, either before or after the predicator, the filler of the predicator slot generally agreed as in the following examples: (1) y on fone lententyde £>aer toforen forbearn se burch on Gleaweceastre. 'And then in the spring before that the borough at Gloucester burned.' (250.3) (2) Jja hwile J>e pa munecas sungen Jsaere messe .... 'While the monks were singing the mass....' (250.4) (3) Jiaer aefter comen feale tacne wide hwear on Englaland. 'After that many signs came far and wide in England.' (250.10-11) (4) hi swa diden. 'They did so.' (251.17) (5) y Jjanon he com to Burch. 'And thence, he came to Peterborough.' (258.10) (6) y micel ungerime folces waepmen y wimmen forburnen. Ά great host of people, men and women, burned to death.' (253.17-18) If the subject slot in dependent clauses was filled by a relative pronoun, the verb agreed with the idea contained in the antecedent of the pronoun. EXAMPLES:

(1) ... eorles castelas Walerampa wzron on Normandi.'... Count Waleran's castles which were in Normandy.' (254.4) (2) oc ure Laford God aelmihtigpa eall digelnesse sed y wat .... 'But our Lord God Almighty who sees and knows all secrets ....' (254.35) (3) Des ilce gasres he gaef J>one abbotrice of Burch an abbot ... se hzfde his abbo-

38

PRIMARY CLAUSE-LEVEL TAGMEMES

trice S'Johs of Angeli on hande. 'This same year he gave the abbacy of Peterborough to an abbot ... who had his abbacy of St. Jean d'Angely in hand.' (257.13-15) (4) for agenes him risen sona J>a ricemen pe wseron swikes. 'for against him soon rose the powerful men who were traitors.' (263.17-18) 2.1.7.1. The collective noun. - There were several uses of the collective noun in this corpus. Folc occurred five times; in every instance, it was used with a singular verb, as in: 1 te Lundenisce folc him underfeng. 'And the London people received him.' (263.14) The other two collective nouns were ungerime and tunscipe, each of which occurred once and with a plural verb: (1) y micel ungerime folces ... forburnen. Ά great host of people ... burned to death.' (253.17-18) (2) ... al te tunscipe flugxn for heom. '... all the township fled from him.' (265.4-5) 2.1.7.2. The coordinated filler of the subject slot. - Agreement of the filler of the subject slot and the filler of the predicator slot in this corpus was fairly normal except in clauses having subject slots filled by coordinated fillers. If a clause had two or more coordinated singular nouns or modified-noun phrases as filler of the subject slot, the verb of the predicator slot, if it came before the subject, was usually singular as in these examples: (1) y mid him ferde se biscop Bernard of Wales, y Sefred abbot of Gleastingbyrig y Anselm abbot of Sancte ALdmund. y Iohan xrcedxcne of Cantwarabyrig. y Gifard wxs pes hinges hird clerc. 'And with him went Bishop Bernard of Wales, and Sefred, abbot of Glastonbury, and Anselm, abbot of St. Edmond's, and John, archdeacon of Canterbury, and Giffard, (who) was the king's court clerk.' (252.18-21) (2) f a scyrte da flescmete y se ceose y se butere. 'Then the meat, cheese, and butter ran short.' (261.24-25) This was generally true also if the coordinated filler of the subject slot was separated and stood after the predicator. Two typical examples are as follows: (1) ι sone pe cosan abbot ferde 7 te muneces mid him to Oxenforde to J>e king. 'And soon the elected abbot and the monks went with him to Oxford to the king.' (268.33-34) (2) y Jjeonen he ferde to Wudestoke. y his biscopes y his hird eal mid him. 'And from that place he went to Woodstock, his bishops, and all his counts with him.' (251.3-4) In the inverted order (VS) the verb was, as a rule, singular even if one of the compound fillers of the subject slot was plural:

PRIMARY CLAUSE-LEVEL TAGMEMES

39

mid him ferde pes kinges stiward of France Amalri. y Hugo Gerueises sunu. y Hugo of Munford. y fela oöre godre cnihte. 'With him went the king of France's steward, Amaury, and Hugo, Gervase's son, and Hugo of Munford, and many other good knights.' (253.26-28) However, if the coordinated filler of the subject slot had plural forms in the inverted order, then the verb was likely to be plural. Here is a typical example: him wiö cwseden muneces y eorles y peignes ealle mest pe pier wseron. 'Monks, earls, and thanes, almost all who were there, spoke against him.' (252.3-4) Furthermore, if the subject slot with the coordinated fillers undivided occurred before the predicator, it generally took a plural verb, especially so when it had a plural element closest to the verb, as in the first two of the examples which follow: (1) and se prior y se munecas of Cantwarabyrig. y ealle pa oöre pe öser wseron munechadesmen hit wiö cwseden fülle twa dagas. 'And the prior and the monks of Canterbury and all the other monks who were there spoke against it two whole days.' (251.28-30) (2) ... swa f>et se eorl Walaram of Mellant. y Hamalri. y Hugo of Mundford, y Willelm of Romarc. and fela oöre wenden fram him y helden here castles him togeanes. '... so that Count Waleran of Meulan, and Amaury, and Hugo of Mundford and William of Roumare, and many others left him and held their castles against him.' (253.10-12) (3) ... J)aer da eorles sunu of Normandi y pes eorles dohter of Angeow wseron totwemde for sibreden. '... there where the Count of Normandy's son and the Count of Anjou's daughter were separated on account of consanguinity.' (258.7-8) The following three examples taken from the Final Continuation were different in that the verb was plural regardless of the position of the fillers of the subject slot: (1) oc ferden pe xrcebiscop y te wisemen betwux heom. 'But the archbishop and the wise men went between them.' (268.9) (2) J)is y te othre foruuardes Jjet hi makeden suoren to halden pe king y te eorl y te biscop & te eorles y rice men alle. 'This and the other agreements that they made, the king, and the count, and the bishop and the earls, and all powerful men swore to keep.' (268.13-15) (3) ... f>et his wif y his sune wseron bebyried set Fauresfeld. '... there where his wife and his son were buried at Faversham.' (268.22) 2.1.7.3. Subjectless clause. - Another interesting aspect of agreement was the occurrence of a subjectless clause just after two simple independent clauses, each with subject and predicator whose fillers were singular. The verb, because the content of the clause referred to the fillers of the subjects of the other clauses, was plural, as in the following examples:

40

PRIMARY CLAUSE-LEVEL TAGMEMES

Siööon t>a com se kyng to Englaland innon heruest. y se eorl com mid him. y wurdon fca alswa gode freond swa hi waeron s r o r feond. 'Afterwards then came the king to England in the fall, and the count came with him, and (they) became then just as good friends as they had been enemies before.' (259.28-30) Although there were exceptions, there was, as a rule, inflectional agreement between the filler of the subject slot and the verb. Singular nouns or pronouns took singular verbs, and plural nouns or pronouns took plural verbs. In the case of the subject with a coordinated filler, however, the verb was normally singular if one of the coordinated elements was singular and especially so if the coordinated fillers were divided.

2.2. THE PREDICATOR TAGMEME

The second primary clause-level tagmeme is the predicator. The predicator was either filled by a simple verb (finite verb) or a complex verb (finite verb with non-finite verb). Of all the primary clause-level tagmemes, the predicator was the only one found in all clauses and coordinated predications. In this corpus, there were 920 predicators, 743 (81 per cent) of which were simple, while 177 (19 per cent) were complex. In the discussion of the subject tagmeme, it was seen that the verb of the predicator slot had a grammatical relationship with the filler of the subject slot and, as a rule, showed agreement. The predicator, of course, was the one obligatory tagmeme in the clause, and the filler of the predicator slot was inflected to show contrasts in person, number, tense, and mood. Because of the nature of the corpus, an annalistic chronicle, not all these contrasts were shown with equal frequency. The prevailing form of the verb was third person singular, past tense, indicative mood: (1) On t>is geare wxs se king Heanri on Cristes maessan on Norht wie. 'In this year King Henry was at Norwich on Christmas.' (250.1-2) (2) Da hwile J>et se xrcebiscop wxs ut of lande, geaf se kyng öone biscoprice of Baöe t>es ewenes canceler Godefreiö waes gehaten. 'While the archbishop was out of the land, the king gave the bishopric of Bath to the queen's chancellor, who was named Godfrey.' (252.34-35) (3) he seide J>et fülle feoht was sett betwenen öa Cristene y J>a heöene. 'He said that a full fight was set between the Christians and heathens.' (259.20-21) 2.2.1. Person Although the third person was the usual form, others did occur: FIRST PERSON:

(1) Laferd kyng ic swelte. 'Lord King, I am dying.' (251.8)

PRIMARY CLAUSE-LEVEL TAGMEMES

41

(2) Ne ]since man na sellice J>et we soö seggen. O n e must not think it remarkable that we tell the truth.' (258.15-16) SECOND PERSON:

... wel pu mythes faren all a daeis fare, sculdest thu neure finden man in tune sittende. '... you could well travel a full day's journey and you were not to find a man sitting in a village.' (264,29-30) 2.2.2. Number Sufficient examples in both singular and plural number to demonstrate this contrast have been presented in a previous discussion on agreement. 2.2.3. Tense As in Old English, verbs in this corpus had only two tense inflections or forms: the present and the past. As a rule, time relationships other than the present and past were either indicated by context or by accompanying time tagmemes. Though the usual time reference expressed was past, there were examples of other time relationships : PRESENT TIME:

(1) oc ure Laford God aelmihtig jja eall digelnesse sed y wat. he seod Jjet man last J>et aerme folc mid ealle unrihte. 'But our Lord God Almighty who sees and knows all secrets (he) sees that one treats the poor people with complete injustice.' (255.1) (2) nu hem behofed Cristes helpe y eall Cristenes folces. 'Now they need the help of Christ and all Christian people.' (262.18-19) (3) Ne Jsince man na sellice t»et we soö seggen. O n e must not think it remarkable that we tell the truth.' (258.15-16) (4) God hit bete J)a his wille bed. 'May God amend it when he wills it.' (261.25) Past perfect time: (1) Enmang t>is was his nefe cumen to Englelande Stephne de Blais. y com to Lundene. 'Meanwhile, his nephew, Stephen of Blois, had come to England and (he) came to London.' (263.13-14) (2) Micel hadde Henri king gadered. 'King Henry had gathered much.' (263.27) (3) y se daecne hafde ongunnan J>one godspel Preteriens Iesus ....' 'And (when) the deacon had begun the gospel Preteriens Jesus ....' (250.5) There were no distinct forms to express future time. In the earlier part of the Chron-

42

PRIMARY CLAUSE-LEVEL TAGMEMES

icle future was expressed by using a present tense form, sometimes accompanied by an adverbial expression of time. Usually the sentence in context would give the sense of future time such as is found in these examples: (1) and gif hwa eow wiöstent we eow fultumiad pxt ge hit magon gegangen, 'and if anyone resists you, we will help you so that you may go to it.' (3.11-12) (2) y hwile abbot t>e bep t>aer coren of |>e munecan J) he beo gebletsad of t>an aercebiscop of Cantwarbyrig. 'Whichever abbot is chosen there by the monks, he will be blessed by the archbischop of Canterbury.* (36.15-16) Such examples did not exist in this corpus. However, there were examples using the present and past tense forms of sculan and willan (the ancestors of shall and will in Modern English). These forms, of course, carried other meanings: the first, volition and the second, desire and obligation. Some examples are as follows: (1) ... J>et hi scolden cumen to his gewitenemot on Candel messe deig to Gleawceastre .... '... that they should come to his council meeting on Candlemas at Gloucester ....' (251.16-17) (2) he hem hit wolde typian. 'He would grant it to them.' (251.20) (3) Nu we willen siegen sumdel wat belamp on Stephnes kinges time. 'Now we will tell partly about what happened in King Stephen's time.' (265.30-31)

2.2.4. Passive Except for the old interesting passive form hatte (from hatan), the notional passive was usually formed and expressed periphrastically with either beon/wesan or weorpan. There were five examples of hatte and het,z one of which is the following: y hatte he Sancte Willelm. 'And he was called St. William.' (266.2) The general use of hatan (in the past participial form) in this corpus was with the auxiliary beon/wesan as in the following example: ... l>a geaf he {Done biscoprice of Lincoln an clerc Alexander wxs gehaten. '... then he gave the bishopric to a clerk who was called Alexander.' (253.1-2) In Classical Old English, wesan and weorpan were used as auxiliaries to form the passive. Though scholars have suggested that wesan was normally used in durative constructions and that weordan was used in perfective constructions, there was no uniformity of such usage here. There were several examples of each, but beon/wesan was the usual form. In addition to the uses of the form hatte and the impersonal (to be discussed later), the passive was expressed in three periphrastic constructions: the present indicative passive, the past indicative passive, and the future indicative '

This f o r m (Aet) is the regular active past tense form.

PRIMARY CLAUSE-LEVEL TAGMEMES

43

passive. There was only one example of the first and two of the third. The past indicative passive had numerous examples. The examples are presented below along with a brief comment about each. 2.2.4.1. Present indicative passive. - The present indicative passive was made up of the present tense of beon/wesan and the past participle of hatan \ y te munekes innen daeis cusen oJ>er of heom saelf. Willelm de Walteuile is gehaten. 'And the monks within the day elected another from themselves who was called William of Walterville.' (268.30-31) 2.2.4.2. Past indicative passive. - The past indicative passive consisted of the past tense form of beon/wesan or weordan and the past participle form of the relevant verb. The first four examples used beon/wesan as the auxiliary verb. (1) y hine bebyrigde se biscop of Ceastre Rotbert Pecceö wxs gehaten. 'And the bishop of Chester, who was called Robert Pecceth, buried him.' (251.12-13) (2) Dis wses eall ear gedon öurh se biscop of Seresbyrig .... 'This had all been done before by the bishop of Salisbury ....' (251.25-26) (3) y wxran t>aer underfangen of J>one pape Honorius mid micel wuröscipe. 'And (they) were there received by Pope Honorius with great honor.' (255.34-35) (4) oc hi wseron siööen totweamde for sibreden. 'But they were afterwards divorced on account of consanguinity.' (257.9) (5) On öes ilces geares wearö swa micel flod ... {jet feola tunes y men weordan adrencte .... 'In this same year there was so great a flood that many towns and people were flooded ....' (256.1) (6) oc se eorl weard gewunded at an gefiht fram anne swein. 'But the count was wounded by a thane in a fight. (258.31-32) (7) £»us earmlice weard eall fet folc swengt. 'Thus miserably were all the people afflicted.' (259.22-23) 2.2.4.3. Future indicative passive. - This form was made up of the present tense of sculan, the infinitive form of beon, and the past participle of the relevant verb. The examples are as follows: (1) ... £et hit scolde beon underded into Clunni. '... so that it should be subjected to Cluny.' (261.11) (2) Wenden J>et it sculde ben forholen. '(They) expected that it would be concealed.' (265.34) 2.2.4.4. Impersonal man (me). - In addition to these expressions of the passive voice, English has historically used the impersonal man in such a construction as hine man ofsloh pa (55.10) to express the passive. This can be translated as someone killed

44

PRIMARY CLAUSE-LEVEL TAGMEMES

him then or he was killed then. The following examples are apparently related to this construction: (1) Man /erode hine to Lincolne mid micel wuröscipe. 'He was taken to Lincoln with great honor.' (251.11) (2) ... £>et man scolde selten clerc ofer muneces .... '... that a clerk should be placed over monks ....' (252.9-10) (3) Me dide cnotted strenges abuton here haeued .... 'Knotted strings were put around their heads ....' (264.12) 2.2.5. Mood In addition to the indicative mood (the usual mood of this corpus), there were also uses of the subjunctive. There were no verbs in the imperative in this corpus. Since most of the examples presented before were indicative, only those illustrating the subjunctive will be presented here. The subjunctive in independent clauses here was virtually confined to expressions of wishes and commands as: (1) Ne pince man na sellice J>et we soö seggen. 'One must not think it remarkable that we tell the truth.' (258.15-16) (2) God scawe fore. 'May God provide.' (258.28) (3) God geare his sawle. 'May God save his soul.' (258.34) (4) God slmihtig haue his milce ofer f>et wrecce stede. 'May God Almighty have mercy on that wretched place.' (259.9-10) The use of the subjunctive was clear in some temporal and concessive clauses; in others, especially the periphrastic subjunctive, it was ambiguous. In temporal clauses, it occurred when the conjunction was xr such as is shown in the following examples: (1) ... xr he wxre dead. '... before he was dead.' (251.26-27) (2) ... xr se aercebiscop of Cantwarabyrig come .... '... before the archbishop of Canterbury came ....' (252.23-24) (3) /Er he wsere wel ded. J>a waere coren twa papes. 'Before he was dead long, two popes were chosen.' (260.17-18) (4) y he ferde him sone to Lincol y was J)aer bletcaed to abbot xr he ham come .... 'And he went soon to Lincoln and was there consecrated as abbot before he came home....' (268.35-36) There were only two other temporal clauses with xr, one clearly not subjunctive and the other ambiguous in that the finite verb form is the same in both the indicative and subjunctive moods. The examples are these: (1) Des ilce geares xr se biscop of Lincolne com to his biscoprice forbearn eall meast se burh of Lincolne. 'This same year, before the bishop of Lincoln came to his bishopric, almost all of the borough of Lincoln burned.' (253.16-17)

PRIMARY CLAUSE-LEVEL TAGMEMES

45

(2) ... xr hi mihte cumen to £>es papes spraece. '... before he could come to the Pope's council.' (252.26) There were five concessive clauses in the corpus: three had pxh as a conjunction; one had pop; the other had no conjunction at all, but began with the verb wxr. (1) ... pxh hit wxre here unjmncas. '... although it was to their dislike.' (252.14) (2) ... pxh hit litel behelde. '... although it was of little avail.' (259.20) (3) y hi togaedere comen y wuröe saehte. pop it litel forstode. 'And they came together and agreed, although it amounted to little.' (263.22-23) (4) ... pxh hit litel forstode. '... although it amounted to little.' (255.30) (5) he hit dide foröi J>et he wolde Jsurh his micele wiles öxr beon wxr it tweolf monö oööe mare .... 'He did it because he wanted, through his great trickery, to be there (whether) it were twelve months or more ....' (259.7-8) There were two other dependent clauses which were clearly subjunctive. One was a noun clause in apposition to an object of the preposition. The other was a conditional clause introduced by gif. It was difficult to tell whether conditional clauses, as a rule, took the subjunctive in this corpus because the other examples were periphrastic verb constructions, which, of course, were ambiguous. The example is as follows: ... gif J)sr wxre hure an unwreste wrenc .... '... if there were at least one poor trick....' (262.6) Bare is a subjunctive form of beran and was used in the following clause, perhaps because of indefiniteness: Wua sua bare his byrthen gold y sylure. durste nan man sei to him naht bute god. 'Whoever bore his burden in gold and silver, no one dared to say anything but good to him.' (263.11-12) In Classical Old English, the subjunctive was generally used after verbs denoting conjecture, verbs of wishing and commanding, and in indirect discourse. In addition to these and other uses, it was often used in periphrastic constructions with the past tense forms of the modal auxiliaries.. These forms in the indicative and subjunctive were then identical. Similar uses in this corpus were ambiguous in the same way. Just a few examples will show this: (1) E»a beJ)ohte he him J>et gif he mihte ben rotfest on Engleland l>et he mihte habban eal his wille. 'Then he thought to himself that if he could be firmly established in England, he could have all that he desired.' (257.33-34) (2) Wenden J>et it sculde ben forholen. '(They) expected that it would be concealed.' (265.34) (3) ... se abbot of Clunni heafde him beboden t>et he scolde cumen to him .... '... the abbot of Clunny had ordered him that he should come to him ....' (261.30-31)

46

PRIMARY CLAUSE-LEVEL TAGMEMES

(4) Xpist ne wolde £>et he sculde lange rixan. 'Christ did not wish that he should rule for a long time.' (268.2) 2.2.6. The Complex Verb The complex verb, which was made up of one finite verb as auxiliary and at least one non-finite verb form, was indeed of great complexity and variety, simply because of the internal structure of the verb phrase itself, but most of all because of its relationship to the diversity and order of modifiers with which it patterned. The word order, as well as the nature of the modifiers of the predicator, deserves closer scrutiny. Thus it will be treated again in another chapter in a section on the modified-verb phrase. The primary concern of this section is the identification of the complex verb patterns and a description of the kinds of auxiliaries and how they introduce the non-finite forms of the verbs. Earlier, it was mentioned that the corpus contained 920 predicators. One hundred seventy-seven (19 per cent) of these were filled by complex verbs. This number does not include the coordinated non-finite forms of the predicator such as the ones in this sentence: ... for hi uueron al forcursaed y forsuoren y forloren. '... for they all were accursed, forsworn, and lost.' (265.6-7) These coordinated elliptical constructions number nineteen. Of the 177 complex verb patterns, ninety-five (53 per cent) were finite verb + infinitive patterns. In this group were included seventeen finite verb + to-infinitive patterns. Five of the toinfinitives were inflected. Seventy-nine (45 per cent) of the patterns consisted of the finite verb + past participle. The other three complex verb constructions were sequences of three verbs. Two of these consisted of finite verb + infinitive + past participle; the other consisted of finite verb + past participle + to-infinitive. Not counted in the number of 177 complex verbs were three patterns which will be briefly discussed once the others have been illustrated with examples from the corpus. It has already been shown that the patterns were separated according to the verbal elements they contained. This is sufficient except for the infinitive constructions, which, for convenience of presentation, need to be subdivided. They, therefore, were further categorized on the basis of the kind of finite verb with which they patterned. The finite verbs were divided into two groups: TRUE-AUXILIARIES and QUASIAUXILIARIES. The first are verbs whose primary function seems to be catenative in that they begin and form chains. In this group are traditional modal auxiliaries from the Classical Old English period and the auxiliaries beon/wesan, habban, and weordati, which were used to form the compound tenses. All the modals but one occurred more than once. Here is a list of the modals and the number of times each occurred: sculan 29, willan 15, magan 13, Ixten 9, motan 3, cunnan 3, durran 3, don 6, agan 1. These combined with the plain infinitive (without to) most of the time. Two, don and

PRIMARY CLAUSE-LEVEL TAGMEMES

47

agan, however, introduced to-infinitives, and sculan twice introduced a three verb sequence. Beonjwesan appeared sixty times, habbati twenty-one times, and weordan five times. These three were used primarily to form the past participle constructions, but to-infinitives occurred three times with habban and twice with beonjwesan. The latter was also one of the verbs in the two patterns with three verbs. All of these, since the corpus is a chronicle, of necessity occurred frequently as simple main verbs. The quasi-auxiliaires were finite verbs whose common function was as a main verb. In this corpus, they occurred only with infinitives. Four plain infinitives and eight to-infinitives appeared with quasi-auxiliaries. Of the nine quasi-auxiliaries, only two occurred more than once: herdon three times and suoren twice. The others were unnan, wendan, betxcan, gifan, lyhtan, myntan, and tacan. The complex verbs, then, consisted of either a finite verb and one non-finite form (a plain infinitive, to-infinitive, or past participle) or combinations of a finite verb and two non-finite forms (an infinitive and past participle or a past participle and to-infinitive). IT IS SIGNIFICANT THAT THE AUXILIARY NEVER OCCURRED AFTER THE MAIN VERB IN THIS CORPUS. The different combinations with varied examples are as follows: Patterns of two verbs: TRUE-AUXILIARY + PLAIN INFINITIVE :

(1) Da hi waeran t>aer gegaderod. t»a bed se cyng heom \>xt hi scolden cesen hem aercebiscop to Cantwarabyrig swa hwam swa swa hi woldon. 'When they were gathered there, then the king ordered them that they should choose from themselves an archbishop to Canterbury, whomsoever they wish.' (251.17-19) (2) y he hem hit wolde typian. 'And he would grant it to them.' (251.20) (3) t>et waes for se man öe hafde an pund. he ne mihte cysten aenne peni at anne market. 'That was because the man who had a pound, he could not get the value of a penny at the market.' (255.9-10) (4) y Jieonon he sende him ... y let hine don on harde bände. 'And thence, he sent him and had him put in hard bands.' (256.12-14) (5) of his utgang ne cunne we iett noht seggon. O f his leaving, we cannot yet say anything.' (258.28) (6) ... Jjet gif he moste Engleland secen .... '... that if he were permitted to visit England ....' (262.12) (7) Durste nan man misdon wiö oöer on his time. 'No one dared do wrong against another in his time.' (263.9-10) (8) I ne can ne i ne mai teilen alle J>e wunder ne alle J>e pines J>et hi diden wrecce men on J)is land. Ί do not know, nor am I able to tell of all the horror nor all the torture that they did to wretched men in this land.' (264.24-25) (9) ... y dide heom cumen Jiider. '... and (he) had them come to that place.' (266.36)

48

PRIMARY CLAUSE-LEVEL TAGMEMES

QUASI-AUXILIARY + PLAIN INFINITIVE:

(1) y t>a muneces herdon öa horn blawen J>et hi blewen on nihtes. 'And the monks heard the horn blowing that they blew at night.' (258.23-24) (2) I>a J>e king was ute Jja herde J)et siegen, y toe his feord y besäet hire in ]je tur. 'When the king was out, then (he) heard that said and took his property and besieged her in the tower.' (267.21-22) (3) Xpist him unne J)us enden. 'Christ grant him to end thus.' (269.3-4) TRUE-AUXILIARY + TO-INFINITIVE :

(1) ... ealle £>a £>et Cristendome hxfdon to begemen. y to locen .... '... all those who had to preserve and protect Christendom ....' (259.34-35) (2) He dide öone king to understanden Jjet he wolde mid alle forlseten Jjone minstre.... 'He gave the king to understand that he would give up the monastery completely ....' (259.4-5) (3) J>e eorl heold Lincol agsnes Jje king, y benam him al Jret he ahte to hauen. 'The count held Lincoln against the king and took from him all that he ought to have.' (266.19-20) QUASI-AUXILIARY + TO-INFINITIVE:

(1) Dauid king of Scotland toe to uuerrien him. 'David, King of Scotland, began to fight him.' (263.21) (2) y gif he leng moste liuen. alse he mint to don of pe horderwycan. 'And if he had lived longer, he meant to do the same for the office of treasurer.' (265.22-23) (3) ... me lihtede candles to seten bi. '... one lit candles to eat by.' (266.12) (4) wende to bigxton Normandi Jjier {wrh. '(He) expected to get Normandy through that.' (267.30-31) (5) y alle diden him manred. y suoren J)e pais to halden. 'All did him homage and swore to keep the peace.' (268.16-17) AUXILIARY + INFLECTED INFINITIVE :

(1) Jsa sone swa he haefde wind, swa ferde he ofer into Normandie. y betxhte {»a eall Engleland to geamene y to wealdan jDone biscop Roger of Searesbyrig. 'Then as soon as he had wind, he went over into Normandy and entrusted them all of England to the care and rule of Roger, bishop of Salisbury.' (253.6-8) (2) oc se kyng hit dide for to hauene sibbe of se eorl of Angeow. y for helpe to hauene togaenes his neue Willelm. 'But the king did it in order to have peace from the Count of Anjou and help against his nephew, William.' (256.34) (3) ... y forjii {jet he wxs an haefod öa aö to swerene C witnesse to berene .... '... and because he was a party to the swearing of the oath and to bearing witness ....' (258.6-7)

PRIMARY CLAUSE-LEVEL TAGMEMES

49

(4) se kyng hem geaf ealle leue ham to farene. 'The king gave them all permission to go home.' (260.7-8) TRUE-AUXILIARY - F PAST PARTICIPLE:

(1) ... y se daecne hafde ongunnan Jione godspel Preteriens Iesus .... '... and the deacon had begun the gospel Preteriens Jesus ....' (250.5) (2) he wses boren of Luuein. 'He was born in Louvain.' (252.36) (3) Des ilces gseres on Jjone lenten tide wses se eorl Karle of Flandres ofslagen .... 'This same year in the spring Count Karl of Flanders was slain ....' (257.3-4) (4) oc se eorl weard gewunded at an gefiht fram anne swein. 'But a count was wounded in a fight by a thane.' (258.31-32) (5) Enmang t>is was his nefe cumen to Englelande Stephne de Blais. y com to Lundene. 'Meanwhile his nephew, Stephen of Blois, had come to England, and (he) came to London.' (263.13-14) Patterns of three verbs: TRUE-AUXILIARY + INFINITIVE+PAST PARTICIPLE :

(1) ... t»t hit scolde beon underöed into Clunni. '... that it should be subjected to Cluny.' (261.11) (2) Wenden J>et it sculde ben forholen. '(They) expected that it should be concealed.' (265.34-35) TRUE-AUXILIARY + TO-INFINITIVE + PAST PARTICIPLE :

J>et wies fortan ]jet hit wzs don done pape to understand .... 'That was because it was given the pope to understand ....' (252.27) 2.2.7. Special Patterns There were four other patterns that demand separate attention: the finite verb with the present participle, the auxiliary with the verb of motion omitted, the verb-adverb sequence, and the habban ...on hande pattern. 2.2.7.1. The present participle. - The periphrastic forms such as was fighting, were fighting of Modern English were not used frequently in Old English. Their notions could be expressed by the past tense forms and usually were. Verb phrases made up of a form of beon and a present participle did not occur in the corpus studied here, though they did appear in the earlier part of the Chronicle. In the text here being analyzed, however, there were three instances of the use of present participles in association with finite verbs. Whether they were used adjectivally or verbally is uncertain. The examples are:

50

PRIMARY CLAUSE-LEVEL TAGMEMES

(1) y riden J>aer sprecende. 'And (they) rode there speaking.' (251.7) (2) ... on t>a tun jsa waes tennploges oöer gangende ....'... in a town which had ten or twelve plows going ....' (261.21) (3) Gif twa men o]?er iii coman ridendto an tun .... 'If two or three men came riding to a town ....' (265.3-4) These participles apparently were not used in tense formation (or predicatively); they seem to function more like non-contiguous adjectivals. Yet, one could argue the opposite for example three, since it comes after an intransitive verb. 2.2.7.2. The omitted infinitive. - After four modal auxiliaries, infinitives were omitted in clauses where the meaning (motion, for example) could be easily supplied from the context. Here are three examples: (1) y t>ser man him held Jset he ne mihte na east na west. 'And he was held there so that he could not (go) east or west.' (262.8-9) (2) ... J>et he ne my the nowiderwardes .... ' . . . s o that he could not (go) in any direction ....' (264.21) (3) hi scolden nedes. 'They had to, of necessity.' (262.3) 2.2.7.3. The verb-adverb-pattern. - Another special pattern was the finite verb and adverb sequence. There were several uses of gifan and up that seem to indicate that up was tied in with gifan. In the first two examples, the elements were contiguous. The other had a nominal separating gifan and up. Examples are as follows: (1) ... y dide him gyuen up Jjet abbotrice of Burch ....'... and (he) made him give up the abbacy of Peterborough ....' (262.29) (2) ... he alle his castles sculde iiuen up. '... he should give up all his castles.' (267.14) (3) ... he uuolde iiuen heom up Wincestre .... '... he would give up Winchester to them ....' (266.36) The habban ...on hande pattern. - There were three uses of the habban ...on hande pattern in the corpus, and all three occurred in the same passage under annal 1127. In each passage habban and on hande were separated by nominal and adverbial tagmemes. The pattern appeared to be idiomatic enough to consider a unit and to translate possessed or had (possession). One example should be enough: oc he wolde hauen baöe on hand. 'But he wanted to have (own, possess) both.' (257. 21-22) 2.3. THE DIRECT OBJECT TAGMEME

The direct object is the third primary clause-level tagmeme. It was filled by a nominal or nominal phrase which had a grammatical relationship with the accompanying

PRIMARY CLAUSE-LEVEL TAGMEMES

51

verb. That is, it occurred with transitive verbs and stood either before or after, but generally after them. Counting all the constructions which had direct objects, the direct object tagmeme occurred after the predicator 324 times (76 per cent). Excluding the objects which were clauses and infinitives and those which appeared in subjectless clauses and coordinated predications, the direct object stood after the predicator in 282 (56 per cent) of the examples. Unlike the subject, its filler did not agree in number with the verb, and it occurred in the oblique cases most of the time. It was an element in subjectless clauses and coordinated predications. Like the subject, the direct object may be simple or compound. At times it joined another object to form a pattern often referred to as the indirect object-direct object construction. In the 763 clauses and 157 coordinated predications, the direct object appeared 406 times (44 per cent). These figures do not reflect the counting of coordinated objects and objects of the infinitive. The direct object slot may be filled by a pronoun or modified-pronoun phrase, a noun or modified-noun phrase, a pronominal adjective, a clause, or a to-infinitive. 2.3.1. The Noun or Modified-Noun Phrase The most common filler of the direct object slot was the noun or modified-noun phrase, which occurred in 201 examples representing about 50 per cent of the total number of direct objects. Ninety-six (48 per cent) of these were in independent clauses, fifty-six (28 per cent) in dependent clauses, and forty-nine (24 per cent) in coordinated predications. Examples: NOUN

(1) Ac J>et ofercom Rome ]}et ofercumeö eall weoruld.... 'But that which overcomes all the world overcame Rome ....' (252.29-30) (2) t>a sone swa he hsfde wind .... 'Then as soon as he had wind ....' (253.6) MODIFIED-NOUN PHRASE:

(1) ... he dide mare yuel .... '... he did more evil ....' (267.32) (2) y he iaf him pat abbotrice. 'And he gave him the abbacy.' (268.34-35) (3) Da cusen hi an clerc Willelm of Curboil waes gehaten. 'Then they chose a clerk (who) was called William of Corbeil.' (251.32-33) (4) Hugo of Mundford he sende to Engleland .... 'Hugo of Mundford he sent to England ....' (253.34) (5) ... y namen pone eorl Waleram y Hugo Gerueises sunu. y Hugo of Munford. y fif y twenti oöre cnihtes .... '... and (they) took Count Waleran, Gervase's son, Hugo of Mundford, and twenty-five other knights ....' (253.30-31)

52

PRIMARY CLAUSE-LEVEL TAGMEMES

(6) y se kyng him geaf eall his land buton his castel ane. 'And the king gave him all his land but his one castle.' (259.27-28) (7) on fif 7 twenti wintre ne biden hi naefre an god dxi. 'In twenty-five winters, they had never experienced one good day.' (262.3-4) (8) ... he seoö t>et man laet pet serme folc mid ealle unrihte. '... he sees that they treated the poor people with complete injustice.' (255.1) 2.3.2. The Pronoun or Modified-Pronoun Phrase The pronoun or modified-pronoun phrase was the second most common filler of the direct object slot. Of the 406 direct object slots, 117 (27 per cent) were filled by this group. Only two of these were modified-pronoun phrases. Fifty-nine (50 per cent) of the 117 pronouns or modified-pronoun phrases occurred in independent clauses, thirty-one (26 per cent) in dependent clauses, and twenty-seven (24 per cent) in coordinated predications. Examples: PRONOUN:

(1) ... swa hwam swa swa hi wolden to aercebiscop. '... whomsoever they wanted as archbishop.' (251.24) (2) ... t»aer J)a ilce lagas J>a Anselm aercebiscop haefde aeror beboden .... '... there the same laws which Archbishop Anselm had ordered before ....' (255.29-30) (3) y man ferode hine to Lincolne mid micel wuröscipe .... 'And they took him to Lincoln with great honor ....' (251.11) (4) he hem hit wolde ty^ian. 'He would grant it to them.' (251.20) (5) ϊ ealle Jm biscopas him underfengen. 'And all the bishops received him.' (252.2-3) (6) pis he dyde eall for pes biscopes luuen. 'This he did all for the bishop's love.' (253.3-4) (7) and βφδοη man hem ofslaeö. 'And afterwards, they killed them.' (255.2-4) MODIFIED-PRONOUN PHRASE:

y se biscop Roger of Saeresbyrig sende ofer eall Englalande y bebead hi ealle J)et hi scolden cumen to Winceastre to Cristesmesse. 'And the bishop, Roger of Salisbury, sent all over England and ordered all of them that they were to come to Winchester at Christmas.' (255.11-13) 2.3.3. The Pronominal Adjective Another filler of the direct object slot was the pronominal adjective, which includes words that functioned both as adjectivals and as nominals. Of the 406 direct object slots, only forty-three (11 per cent) were filled by members of this group. Twenty-

PRIMARY CLAUSE-LEVEL TAGMEMES

53

three (54 per cent) occurred in independent clauses, and nine (20 per cent) occurred in coordinated predications. Here is a list with the number of times each occurred: nan, 3; mare, 6; ealle, 9; stime, 5; oder, 10; aw, 5; micel, 3; bade, 2; fela, 1. Examples of each follow: (1) öe nan ne heafde staerf of hungor. 'He who had nothing died of hunger.' (255.5) (2) l>a J)e uurecce men ne hadden nan more to gyuen .... 'When the wretched men had no more to give ....' (264.28-29) (3) Eall t»t jsa beon dragen toward swa frett J)a drane y dragaö fraward. 'The drones thus eat and drag out all that the bees drag in.' (258.12-13) (4) y sume ne iaf he noht. 'And some, he did not give up.' (267.15) (5) for suricman sone raeude oper J>e mihte. 'For everyone who could robbed the others.' (263.6-7) (6) I>a hi öider coman öa nam man an y an y benam aelc öone riht hand y t>a stanes beneöan. 'When they came to that place, they took them one by one and deprived each of his right hand and his testicles below.' (255.13-14) (7) y sex fleh y forles i>ar micel. 'And she fled and lost much there.' (266.30-31) (8) oc he wolde hauen bade on hand. 'But he wanted to have both in hand.' (257.2122) (9) y of Jm oöre swa fela swa him fmhte he sende norö y suö to hise casteles on heftnunge. 'And of the others, he sent as many as he pleased to prison north and south to his castles.' (254.1-3) Although swa was not counted here as a pronoun filler of the direct object slot, there may be good reason to do so. In example 3 above, illustrating ealle, swa could be considered the direct object offrett instead of an adverb as it has been regarded here. If so, eall plus its relative clause would be an appositive. Another example followed the one mentioned: eall t>et he mihte tacen wiöinnen y wiöuten of lsered y of lswed swa he sende ouer sae.... (258.14-15) In both these cases, it seems reasonable to translate swa as that. For example, the translation above would be 'that, he sent across the sea'. 2.3.4. The Nominal Clause The clause was the fourth most frequent filler of the direct object slot. It appeared in forty-two examples, 10 per cent of the total number of direct object slots. Twentyone of these occurred in independent clauses, five in dependent clauses, and sixteen in coordinated predications. All forty-two followed the verb and, with the exception of four, were introduced by pet, which had no nominal function in the clause it introduced. The sole function of pet was that of a subordinating tagmeme. Two of the clauses were direct discourse and, therefore, did not have connecting tagmemes.

54

PRIMARY CLAUSE-LEVEL TAGMEMES

One of the other two was joined by swa hwam swa swa, which functioned as the direct object within the boundaries of the clause it linked to the main clause. The other was joined by wat, which also served as a clause-level tagmeme (the subject) within its clause. Examples which follow will include those of the direct discourse, the clauses introduced by the conjunctions that act as primary tagmemes, and two of the other thirty-eight introduced by pet. (1) l>a aseh dune biscop of Lincolne y seide to Jsam kyng. Laferd kyng ic swelte. 'Then the bishop of Lincoln fell down and said to the king, "Lord King, I am dying.'" (251.7-8) (2) oc man seiö to biworde. hxge sitted pa aceres dxleth. 'But one says as a proverb, "The hedge remains that divides the fields.'" (261.12) (3) ... fet hi mosten cesen of clerc hades man swa hwam swa swa hi wolden to ercebiscop. '... that they could choose from the secular clerks whomsoever they wished as archbishop.' (251.23-24) (4) Nu we willen siegen sumdel wat belamp on Stephnes Kinges time. 'Now we will tell in part what happened in King Stephen's time.' (265.30-31) (5) ... y saedon pet hi ssegon on norö east fir micely brad wiö jx>ne eoröe. '... and (they) said that they saw in the northeast a great and broad fire near the earth.' (250.18-19) (6) t>a bejjohte he him pet gif he mihte ben rotfest on Engleland {jet he mihte habben eal his wille. 'Then he thought to himself that if he could get firmly rooted in England that he might have his will.' (257.33-34) 2.3.5. The To-Infinitive The to-infinitive, the last type of filler of the direct object slot, occurred only three times: twice in independent clauses and once in a coordinated predication. Instances are as follows: (1) wende to bigseton Normandi J>aer Jjurh. '(He) expected to get Normandy through that.' (267.30-31) (2) t>is y te othre foruuardes J^et hi makeden suoren to halden J>e king y te eorl y te biscop & te eorles ι rice men alle. 'The king, and the count and the bishop and the earls, and all powerful men swore to keep this and the other conditions that they made.' (268.13-15) (3) y alle diden him manred. y suoren pe pais to halden. 'And all did him homage and swore to keep the peace.' (268.16-17) 2.4. THE OBJECT COMPLEMENT TAGMEME

A special type of clause-level tagmeme recognized in the corpus is the object complement. The object complement slot was filled by a nominal which had the same referent

PRIMARY CLAUSE-LEVEL TAGMEMES

55

as the filler of the direct object slot. In this corpus there were two kinds of fillers. The first was a single noun which stood after the direct object as in the following examples: (1) y te king makede Teodbald xrcebiscop jje was abbot in the Bee. 'And the king made Theobald archbishop who had been an abbot at Bee.' (266.14-15) (2) Hi laeiden gaeildes on the tunes aeure um wile y clepeden it tenserie. 'They constantly laid taxes on the village and called it tallage.' (264.27-28) The other was a set phrase consisting of a noun and a preposition which introduced it. The object and its complement may be adjacent or non-adjacent. Typical examples of the first and then the other are as follows: (1) ... and saeden J>aet hi naefre mare ne wolden hafen munec hades man to xrcebiscop ofer hem. '... and said that they nevermore wanted to have a man in monastic orders as archbishop over them.' (251.21-22) (2) I>es ilee WilleLm hasfde aeror numen öes eorles dohter of Angeow to wife. 'This same William had taken earlier the Count of Anjou's daughter as wife.' (257.8-9) 2.4.1. Case of the Filler of the Direct Object Slot Because of the close relationship of the direct object to the indirect object in general and the confusion of case of the fillers of these two object tagmemes in particular, discussion of this matter will be delayed until after the indirect object has been presented as a clause-level tagmeme. 2.5. THE INDIRECT OBJECT TAGMEME

The indirect object, the fourth primary clause-level slot, was filled by a nominal or nominal phrase which had a grammatical relationship with a transitive verb. Other than having restrictions of appearing with predicators filled by transitive verbs and direct objects, the indirect object did not have any other compulsory grammatical relationship with these or the other two nominal tagmemes. That is, its filler did not agree in number or case with them; neither did it modify them in any way, except to expand the syntactic pattern by complementation. The most frequent position of the indirect object tagmeme with respect to the predicator and direct object was medial, that is, between them in the subject + predicator + indirect object -f- direct object group as in this clause: ... }jet he scolde beieton him Jjone mynstre of Burch ... . ' . . . that he would get them the monastery of Peterborough ....' (261.10-11) In this clause him is the filler of the indirect object slot in the dative case, and done mynstre is the filler of the direct object slot in the accusative case. This order, however,

56

PRIMARY CLAUSE-LEVEL TAGMEMES

was not absolutely fixed. In addition to the SVIO pattern just illustrated, there were eight other variations, seven of which occurred more than once. Along with the typical pattern are listed here the variations with the number of times each occurred: SVIO, 30; SIVO, 8; SVOI, 6; OSVI, 6; VSIO, 6; SOIV, 4; VSOI, 2; ISVO, 2; and SIOV, 1. This list shows that the indirect object occurred in all the possible positions. (More about word order of the indirect object-direct object construction will be presented in the next chapter.) Of the 406 clauses and coordinated predications together having direct object tagmemes, sixty-five (16 per cent) contained also an indirect object tagmeme. Of the sixty-five indirect objects, forty-seven (72 per cent) were filled by pronouns or modified-pronoun phrases, sixteen (25 per cent) by nouns or modified-noun phrases, and two (3 per cent) by pronominal adjectives. Forty-three (66 per cent) of the fillers of the indirect object slots were in the dative case, and all of these were pronouns. Nine (14 per cent) were in the accusative case; four of these were pronouns and five, nouns. Thirteen (20 per cent) of the fillers of indirect object slots were ambiguous as far as case was concerned. Eleven of these were ambiguous because the determiners and adjectives either lacked inflectional endings or had inflectional endings which were nondistinctive. The other two were the pronouns xlc and pe, the first of which was uninflected and the second, invariable in form. It may be of interest to note that the indirect object occurred in association with a direct object slot filled by a clause thirteen times (29 per cent), by a noun thirty-eight times (58 per cent), by a pronoun ten times (16 per cent), and by a pronominal adjective four times (6 per cent). In the indirect object-direct object construction, both object slots were filled by pronouns in six examples (9 per cent) and nouns in eight (12 per cent). The frequency of occurrence of other combinations are as follows: pronoun and noun, 31 (48 per cent); pronoun and clause, 9 (14 per cent); noun and clause, 4 (6 per cent); pronominal adjective and pronoun, 6 (9 per cent); and pronominal adjective and noun, 1 (2 per cent). Since all the types of indirect objects appeared in these combinations, examples will not be given solely according to types of slotfillers but according to the combinations of the slot-fillers just listed. The reason for this should be obvious. The first is the pattern with two pronouns. 2.5.1. Pronoun and Pronoun

With pronouns filling both object slots, the indirect object stood before the direct object in three of the six examples in these patterns: SVIO, SIOV, ISVO. In all these the clause was independent, with the filler of the indirect object in the dative case and with the filler of the direct object in the accusative. Here is an example of the two pronoun pattern with the indirect object before the direct object: he hem hit wolde ty£>ian. 'He would grant it to them.' (251.20) The other three examples had the direct object before the indirect object: twice in the

PRIMARY CLAUSE-LEVEL TAGMEMES

57

SOIV pattern and once in the OSVI pattern. In these the fillers of the indirect object and direct object slots were also dative and accusative respectively. An example of the two pronoun pattern with the direct object before the indirect object is as follows: se kyng hit hem tidde. 'The King granted it to them.' (251.25) 2.5.2. Noun and Noun Eight clauses, one of which was dependent, occurred with nouns filling the object slots. The indirect object came after the direct object six times and in all positions except the first. The patterns of these and the number of times each occurred are VSOI, 2 and SVOl, 4. The other two were SVIO and SIVO. None of the fillers of the indirect objects was distinctively dative. Only five accusative forms appeared. Three signaled the direct object and two the indirect. The others were ambiguous in form. Since there were no dative forms, a pattern with two accusatives will serve to illustrate the direct object-indirect object order, and one with an ambiguous form and a genitive direct object, the indirect object-direct object order: (1) y se kyng Henri geaf pone biscoprice aefter Micheles messe pone abbot Henri his nefe of Glastingbyri. 'And King Henry gave the bishopric after Michaelmas to his nephew, Henry, abbot of Glastonbury.' (260.12-14) (2) ... Jjet man scolde beniman ealle pa minitere t>e waeron on Englalande heora Liman .... '... that one should take the limbs from all minters who were in England ....' (255.7-8) 2.5.3. Pronoun and Noun Thirty-one examples appeared with pronoun and noun as fillers of the two object slots. Twenty of these appeared in independent clauses, six in dependent clauses, and five in coordinated predications. All except two occurred with the pronoun as the first of the two. Although the positional pattern of the two objects with the other tagmemes frequently varied (there were seven different orders of the four slots: SVIO, SIVO, SVOI, OSVI, VSIO, SOIV, ISVO), the examples showed that when a noun joined a pronoun in the indirect object-direct object construction, it usually stood in the last position. Even in the three patterns where the noun served as the indirect object, it came after the pronoun. This is significant, considering that the indirect object followed the direct object in only five of the thirty-one patterns. Twenty-eight (90 per cent) of the indirect objects in this combination were pronouns, and only three (10 per cent) were nouns. All except three (10 per cent) of the direct objects were nouns. Twenty-six (84 per cent) of the thirty-one patterns occurred with the indirect object before the direct object. This order occurred in the SVIO pattern fifteen times, the SIVO pattern seven times, the VSIO pattern three times, and the ISVO pattern

58

PRIMARY CLAUSE-LEVEL TAGMEMES

one time. The filler of the indirect object slot was in the dative case (him) twentythree times (89 per cent), the accusative two times (7 per cent), and the invariable pe once (4 per cent). The filler of the direct object slot was in the accusative case nine times (35 per cent); and in the other seventeen instances (65 per cent), forms were either uninfected or not clear. Some examples are as follows: (1) y se kyng him geaf done xrcebiscoprice. 'And the king gave him the archbishopric.' (252.2) (2) y alle diden him manred .... 'And all did him homage ....' (268.16) (3) y iaef hine pone eorldom. 'And (he) gave him the earldom.' (257.7) Only five (16 per cent) of the thirty-one patterns occurred with the direct object before the indirect object. The patterns with the number of times each occurred are SVOI2, SIOV 2, and OSVI 1. The filler of the indirect object slot was a noun ambiguously marked in one instance; in the others, it was a noun marked accusative and a dative pronoun twice each. The filler had ambiguous inflections in two examples and showed the accusative case in three. In the two examples with the dative pronouns, the nouns ambiguously marked came first. The accusative pronoun was the first element in the other three. Two of these were followed by nouns marked accusative, and one by an ambiguously marked noun. Of the following examples, the first three had the pronoun before the noun: (1) ... pet he sror hafde giuenpone kasere Heanri of Loherenge to wife. '... whom he had earlier given to the emperor, Henry of Lorraine, as wife.' (256.10) (2) ... y betahte hine his sune Rotbert eorl of Gleucaestre .... '... and (he) entrusted him to his son, Robert, of Gloucester ....' (256.20) (3) beteahte hine siööon pone sercebiscop Willelme of Cantwarabyrig. '(He) later entrusted him to the archbishop of Canterbury.' (255.20) (4) ... y micele gersumes him geaf on gold y on silure. '... and much treasure in gold and silver (he) gave him.' (259.13-14) The third example is interesting. Usually in both earlier and modern grammars of the English language, scholars have stated in their descriptions of the indirect objectdirect object construction that, of the two referents, the indirect object is the personal referent. This point does not have to be much stressed in the study of Classical Old English, for case could distinguish the two easily. In this corpus there were three examples with both objects having personal referents, and all three had accusative forms for both objects, as in the third example above. In speech, intonation contours would certainly have resolved any ambiguity even if the utterance was out of context, but that, of course, does not help here. That the indirect object occurred in all possible positions further confuses the matter because one cannot be sure that the second object is the indirect object even though there seems to be a preference for that order when it is a noun. In isolation, the objects cannot be distinguished with assurance.

PRIMARY CLAUSE-LEVEL TAGMEMES

59

In context, there is no problem: hine is the direct object and frone xrcebiscop is the indirect object. In all these examples the noun with pone as a marker is the indirect object. It is difficult to say what this means. In modern dative periphrastic constructions, and in some in this corpus, the position in the three examples is typical. For example: (1) ... y brohton hem to pone kinge. '... and brought them to the king.' (253.32) (2) ... cusen J>a muneces abbot of hem self.... '... the monks elected an abbot from themselves ....' (261.34-35) (3) y te munekes innen daeis cusen ojser of heom sxlf.... 'And the monks within a day elected another from themselves ....' (268.30-31) Is it possible that the indirect object, especially if it was a noun, moved to this position before the use of the periphrastic indirect object became widespread? It would seem that position would be more of a signal device than any other. Look at the first of the three examples. Not only is the direct object in the dative case, the word kinge has a dative ending (e) preceded by an accusative demonstrative pronoun and the preposition to. It is apparent that case was not the only syntactic signal. 2.5.4. Pronoun and Clause Nine (14 per cent) of the indirect object-direct object constructions occurred with pronouns as fillers of the indirect object slot and clauses as fillers of the direct object slot. All the direct objects followed both the predicator and the indirect object. In fact, only three examples deviated from the SVIO pattern, and these deviations involved only the subject and predicator. That is to say, there were three examples with inversion of subject and predicator. All were, of course, the VSIO pattern. In all but one case the pronoun was dative. The exception was in the accusative case. Some typical examples are presented below: (1) ... j?a bed se cyng heom pxt hi scolden cesen hem xrcebiscop .... '... then the king ordered them that they were to choose from themselves an archbishop ....' (251.18-19) (2) ... y bebead hi eallepet hi scolde cumen to Winceastre .... '... and (he) ordered all of them that they should go to Winchester ....' (255.12) (3) ... y saeide him pet he wses eald man .... '... and (he) said to him that he was an old man ....' (258.1-2) 2.5.5. Noun and Clause There were only four examples (6 per cent) of the noun and clause filling the indirect object-direct object construction. Like the pronoun and clause combination, this pattern, with the indirect object first, followed the predicator. In only one example

60

PRIMARY CLAUSE-LEVEL TAGMEMES

was the case form distinctive, and that one was accusative. Only the SVIO pattern was represented, and two of the examples follow. The first is the one with the accusative filler: (1) y he saedepone cyngpet hit wass togeanes riht .... 'And he said to the king that it was unjust ....' (252.9) (2) ... bead biscopes y abbotes ... pet hi scolden ealle cumett to Lundene at Michxles messe .... '... (he) ordered bishops and abbots ... that they all were to come to London at Michaelmas ....' (259.32-36) 2.5.6. Pronominal Adjective and Pronoun Six (9 per cent) of the indirect object-direct object constructions were filled by the pronominal adjective and pronoun combination, with the former occurring before the latter four times. In the three different patterns (OSVI, 3; SVIO, 2; SVOI, 1), the indirect object occurred only two times before the direct object. The other times it occurred last with the direct object in either the first or third positions. In all instances the filler of the indirect object slot was dative, and the filler of the direct object slot was non-distinctive in that it was uninfected. Examples of this combination follow (the first is the pronominal adjective before the pronoun): (1) ... eall he scolde hem betaecen. '... all of this he would commit to them.' (262. 15-16) (2) ... forJ)i Jjet he ne iaf him al.... '... because he did not give them everything ....' (266.17) 2.5.7. Pronominal Adjective and Noun There was one example of a pronominal adjective and noun combination, and that occurred in the SVIO pattern. The filler of the indirect object slot was the pronominal adjective which was non-distinctive in case. The direct object filler was an accusative noun. Here is the example: ... y benam selc done riht hand y J>a stanes beneöan. '... and (one) took from each his right hand and his testicles below.' (255.13) 2.5.8. The Dative-Accusative Combination Out of the sixty-five indirect object-direct object patterns, eleven (17 per cent) combined dative filler with accusative filler. The number of these along with the syntactic pattern in which they appeared are as follows: SVIO SOIV VSIO SIOV

Pronoun Pronoun Pronoun Pronoun

Dative Accusative Dative Dative

Noun Accusative Pronoun Dative Noun Accusative Pronoun Accusative

4 2 2 1

PRIMARY CLAUSE-LEVEL TAGMEMES

SIVO ISVO

Pronoun Dative Pronoun Dative

Noun Accusative Pronoun Accusative

61

1 1

No strong generalizations can be made from so few examples, yet the tendencies here are worth noting. In all these the dative filler was a pronoun, and the accusative filler was a pronoun in four cases and a noun in seven. In nine out of the eleven patterns, no other primary tagmeme (subject or predicator) separated the dative object (indirect) and accusative object (direct). Only twice did the direct object come before the indirect object. In nine of the eleven examples the dative object preceded the accusative object. In the two examples where it did not, the accusative was a pronoun in identical patterns (SOIV). From the different positions of the pronoun (1, 2, 3), in these few patterns there was, it is evident, an indication of some freedom of movement. Yet this movement can easily be exaggerated, for in the eleven cases, as was noted, the indirect object came before the direct object. And in only two of these was it separated by another primary tagmeme, once each by a subject and by a predicator. Just a moment ago, it was pointed out that whenever nouns filled the indirect object-direct object construction, the direct object usually occurred not only after the predicator and indirect object but also in the last position. These few examples are parallel to those on which that observation was based. If this sample is the least bit indicative, then, despite the different positions shown, it seems safe to say that there was at least a preference for the dative object to precede the accusative object and for the predicator to precede both. Furthermore, this is especially true if the dative object was filled by a pronoun and the accusative object by a noun. Some typical examples are as follows: SVIO: ... })et he scolde begeton hem done mynstre of Burch .... '... that he would get them the monastery of Peterborough ....' (262.12-13) SOIV: y se kyng hit him iaette .... 'And the king gave it to him ....' (258.5) SIOV: he hem hit wolde tyjiian. 'He would grant it to them.' (251.20) YSIO: ... Ipa. geaf se eorl him pone abbotrice of Sancte Iohns minstre of Angeli. '... then the count gave him the abbacy of the monastery of St. Jean d'Angely.' (257.26-27) ISOV: him me hit beraefode mid strange geoldes y mid strange motes. 'One deprived him of it by severe taxes and by severe courts.' (255.4)

62

PRIMARY CLAUSE-LEVEL TAGMEMES

2.5.9. The Accusative-Accusative Combination There were four examples of the accusative object-accusative object appearing in the same clause. In all but one example the combination was filled by a pronoun plus noun; the exception was noun plus noun. These combinations and patterns are as follows: SVIO SVOI SVOI OSVI

Pronoun Accusative Pronoun Accusative Noun Accusative Pronoun Accusative

Noun Noun Noun Noun

Accusative Accusative Accusative Accusative

Three of these were pronouns plus noun combinations and one, noun plus noun. The noun served as filler of the indirect object in all examples but one. Although just four examples are not enough for a trustworthy generalization, it is remarkable here that in examples where the noun filled the indirect object slot, the indirect object occupied the last position, thus giving some more evidence that the noun generally came last in the indirect object-direct object combination. For example: (1) beteahte hine siööon pone sercebiscop Willelme of Cantwarabyrig. '(He) entrusted him later to the archbishop of Canterbury.' (255.20) (2) j se kyng Heanri geaf pone biscoprice ©fter Micheles messe pone abbot Henri his nefe of Glastingbyri. 'And King Henry gave the bishopric after Michaelmas to the abbot, Henry of Glastonbury, his nephew.' (260.13-14) 2.5.10 Case of Direct Object and Indirect Object Where earlier Old English inflection has been characterized as full and as being the primary basis for distinguishing syntactic relationships, inflection in the language of this portion of the Peterborough Chronicle cannot easily and accurately perform this grammatical function alone, because contrasting inflectional morphemes have been greatly reduced, and in some cases, have just about disappeared. Even when an inflectional ending was present, it was a problem to determine exactly just what its force and function were. This, of course, was much truer of the Final Continuation. However, a strong tendency toward simplification was very apparent in the First Continuation; and in the Final Continuation, it had advanced almost to the stages of disintegration. It was stated before that the personal pronoun was the best basis for determining the case of the filler of the subject slot since distinctive forms in nouns and adjectives were hard to see. To establish the case of the fillers of direct and indirect objects was even a greater problem, for both dative and accusative forms interchange functions without any evident reasons for doing so. For example, the personal pronoun hine was used eighteen times as a direct object filler in the First Continuation and not at all in the Final Continuation. The noun plus the demonstrative pronoun done served

63

PRIMARY CLAUSE-LEVEL TAGMEMES

as a direct object filler thirty-three times. (See Tables II and III just below for these and other percentages, and also for those of the indirect object.) The form done did not occur in the Final Continuation either. The traditional dative form him (singular and plural) appeared twenty times as a direct object filler and twenty-nine times as an indirect object filler in the First Continuation. In the Final Continuation him was used as filler for both the direct and indirect objects. As filler of the direct object slot it was used thirty-seven times and as filler of the indirect object slot, fourteen times. This situation and the fact that the rest of the 406 direct objects and sixty-five indirect objects had either uninfected or ambiguous form classes and larger syntactic patterns (clauses and infinitives) filling their slots make the matter of precisely establishing the accusative as the case of the direct object filler and the dative as the case of the indirect object filler a ticklish one, whether the corpus is considered as a TABLE II

Inflectional Distinctiveness in Fillers of Direct Object Slots First Continuation No. % Acc. Pron. Dat. Pron. Noun w/Acc.Attrib. Pron. (Ambig.) Nouns (Ambig.) Pron. Adj. (Ambig.) Clause Infinitive

18 20 33 29 83 14 27 0

8 9 15 13 37 6 12 0

224

Final Continuation No. % 0 37 0 13 85 29 15 3

0 20 0 7 47 16 8 2

182

Total No.

%

18 57 33 42 168 43 42 3

4 14 8 10 41 11 10 1

406

TABLE III

Inflectional Distinctiveness in Fillers of Indirect Object Slots First Continuation No. % Acc. Pron. Dat. Pron Noun w/Acc. Attrib. Pron. (Ambig.) Nouns (Ambig.) Pron. Adj. (Ambig.) Clause Infinitive

2 29 5 0 8 2 0 0 46

4 63 11 0 18 4 0 0

Final Continuation No. % 0 14 0 2 3 0 0 0 19

0 73 0 11 16 0 0 0

Total No.

%

2 43 5 2 11 2 0 0

3 66 8 3 17 3 0 0

65

64

PRIMARY CLAUSE-LEVEL TAGMEMES

whole or in two parts. In the First Continuation where both forms were used, there was apparently no strict rule governing the use of one or the other. Out of thirty-eight uses of the personal pronoun as filler of the direct object slot, hine was used eighteen times and him twenty times. Out of thirty-one uses of the personal pronoun as filler of the indirect object slot, hine was used only twice, and him was used twenty-nine times. This does not, however, necessarily establish the dative case as the regular one for the indirect object filler, but rather points to the Final Continuation, where him is used exclusively for both the direct and indirect objects. In short, use of hine and him in the First Continuation was either haphazard, or hine was on the verge of extinction; in the Final Continuation it has become extinct. If this can be said, then hine and done can be considered as archaisms and him as an objective form for both object functions, as it is today. That the indirect object filler in the sixty-five examples was not expressed by a distinctively marked dative ending either in the nouns or the nouns with attributes tends to support this. An ending e, which may be considered dative, however, did appear in set phrases such as to wife, with nouns in other prepositional phrases, and with some nouns filling direct object slots. Since an example of the set phrase has already been presented, only examples of the other two are listed here: (1) y hi nan helpe ne haefden of t>e kinge. 'And they had no help from the king.' (267.28) (2) oc scae hedde litel blisse mid him. 'But she had little joy with him.' (268.1-2) Even so, trying to establish the inflectional ending e as dative may be difficult. For one thing, gender4 had disintegrated in this corpus and was no longer a grammatical force. In the First Continuation se was used with nouns of all genders. Seo, the feminine form, occurred once, and that was with a masculine noun: t>eos ilce geares foröferde Raulf seo aercebiscop of Cantwarbyrig. 'This same year Ralph, the archbishop of Canterbury, died.' (250.16-17) Another indication that e did not have any grammatical force was that when it was used, it normally occurred in a prepositional periphrase, which, of course, made it grammatically redundant.

2.6. THE SUBJECT COMPLEMENT TAGMEME

The subject complement is the fifth and last primary clause-level tagmeme of the clause. It was filled by a nominal or nominal phrase which had the same referent as the subject. It might also be a noun plus a preposition (which precedes the noun) as to wife. In this corpus the filler of the subject complement slot was not inflected to 4 Cecily Clark, "Gender in the Peterborough (1957), 109-15.

Chronicle 1070-1154", English Studies, XXXVIII

PRIMARY CLAUSE-LEVEL TAGMEMES

65

agree with the filler of the subject slot, except that both regularly had the same number. Generally, the order of the subject, predicator, and complement was SVC, although the complement occurred in every possible position. In dependent clauses the normal order with all three tagmemes present was also SVC. However, whenever the relative subject filler was omitted in a set clause such as y hine bebyrigde se biscop of Ceastre Rotbert Pecceö wses gehaten. 'And the bishop of Chester (who) was called Robert Pecceth, buried him.' (251.12-13) the order was (S)CV. The subject complement was linked to the subject by the predicator filled by one of the forms of hatan or of the equational verbs beonjwesan and weordan. When the set phrase occurred as the subject complement filler, the predicator fillers were haJgian and bledsian. Of the 920 structures of predication, sixty-eight (7 per cent) appeared with subject complements. Forty-six (68 per cent) occurred in patterns with an overt subject. The other twenty-two (32 per cent) were, therefore, in subjectless constructions, of which five were coordinated predications. A subject complement might be filled by a noun or modified-noun phrase or a set phrase. There were sixty-five of the first and only three of the second. Some typical examples are as follows: PROPER NOUN:

(1) Se an was gehaten Petrus. 'One was called Peter.' (260.18) (2) ... se pape on Rome Calistus waes gehaten. '... the pope in Rome (who) was called Calistus.' (254.27) (3) ... j aefter his daei ware Henri king. '... and after his day Henry would be king.' (268.11) COMMON NOUN:

(1) he was canonie of an mynstre Cicc hatte. 'He was canon of a monastery called Cicc.' (252.1) (2) siööan warö he munec on Clunni .... 'Later, he became a monk at Clunny ....' (257.23) (3) for agenes him risen sona Jsa ricemen t>e waeron swikes. 'For soon the powerful men who were traitors rose against him.' (263.17-18) MODIFIED-NOUN PHRASE:

(1) ... aer waes pes Caseres wif of Sexlande .... '... who was before the wife of the emperor of Saxony ....' (256.29) (2) Ful heui gier waes hit. 'It was a very bad year.' (255.3) (3) God wimman scae wses. 'She was a good woman.'(268.1)

66

PRIMARY CLAUSE-LEVEL TAGMEMES

Set phrase: (1) ... y waes J>aere sone gebletsod to biscop .... 'And was there consecrated as bishop ....' (252.14-15) (2) y he waes gehalgod to biscop .... 'And he was consecrated as bishop ....' (260.14) (3) ... y was J>aer bletcaed to abbot.... '... and was consecrated as abbot....' (268.35)

2.7. G E N E R A L CONCLUSIONS

In brief summary then, the facts especially worth noting here are the following: (1) Except for a few special features, the primary tagmemes of this corpus have not altered greatly from those of Classical Old English; and, for that matter, they do not vary greatly from those of Modern English. (2) As was true of Classical Old English and also of English today, the primary functional slots were manifested by a variety of form classes and larger structures. (3) Though the simplest form of predication was that with only a predicator tagmeme, most predications consisted of at least subject and predicator tagmemes. (4) A close examination of the corpus and a thorough comparison with the morphological studies of Behm, Meyer, and Clark (cited earlier) show that the inflectional system was in great confusion and was frequently incapable of signaling grammatical relationships without the aid of other devices. (5) As a general rule, it may be said that gender and case inflections (with regard to the distinction of the subject-direct object relationship and the indirect objectdirect object relationship) were not prominent here as in Classical Old English and, furthermore, were far advanced toward Modern English. (6) Usually, there was inflectional agreement in number between the fillers of the subject and predicator slots. (7) Heavy coordinated fillers of the subject slot tended to be split, with a part before and a part after the predicator. (8) The auxiliary verb never occurred after the main verb of the predicator, although it did not always occur adjacent to it. (9) As was true of Classical Old English, there were only two tenses for verbs in this corpus, but through adverbs of time and periphrastic constructions time relationships other than present and past times were expressed. (10) Except for a few uses of hatte, notional passive was expressed periphrastically and impersonally. (11) The indirect object usually stood before the direct object, occurring in the IO order 84 per cent of the time. (12) In the indirect object-direct object constructions when the accusative and dative inflections were distinctive, the dative usually preceded the accusative.

PRIMARY CLAUSE-LEVEL TAGMEMES

67

(13) When a noun joined a pronoun in an indirect object-direct object construction, the noun, as a rule, stood last. (14) When both object slots were filled by nouns, the indirect object generally stood last. (15) As a general rule, the fillers of the nominal slots were not consistently marked for case. (16) In the Final Continuation se gave way to the analogical form pe, and the accusative inflections gave way to the dative.

3

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

Works treating the historical development of word order in the English language have stated, more often than not, that since Old English was highly inflectional and had distinctive case endings for nouns, pronouns, and adjectives, the primary elements (subject, predicator, object, etc.) were free and could appear almost anywhere. Charles C. Fries remarked that the order of words in Old English had no bearing upon the grammatical relationships within clauses1 and implied that word order did not seem to be forming clear patterns until the fourteenth century and to be fully established until the middle of the fifteenth century.2 He said specifically: "Taxemes of selection [inflectional endings] do the work and word-order is non-distinctive ...." s Since Fries made his summary statement, a number of important studies have shown that to speak of Old English syntax as "highly inflectional, loosely ordered" was not to give a very accurate characterization of it. These studies proved, although the primary elements appeared in every position, that Old English had less syntactic freedom than has heretofore been acknowledged and that some analytic structures were well established as early as the tenth century.4 1

Charles C. Fries, "On the Development of the Structural Use of Word-Order in Modern English", Language, XVI (1940), 199. 2 Ibid., p. 204. s Ibid., p. 199. 4 See, for example, Robert L. Saitz, "Functional Word Order in Old English Subject-Object Patterns" unpublished Ph.D. dissertation, University of Wisconsin, Madison, 1955; Charles R. Carlton, Descriptive Syntax of the Old English Charters (The Hague, Mouton, 1970); and Ann Shannon, A Descriptive Syntax of the Parker Manuscript of the Anglo-Saxon Chronicle from 734 to 891 (The Hague, Mouton, 1964). Important also are the following earlier studies, the first two of which Fries even made extensive use of in his 1940 article: Russell Thomas, "Syntactical Processes Involved in the Development of the Adnominal Periphrastic Genitive in the English Language", unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1931; Frederic G. Cassidy, "The Backgrounds in Old English of the Modern English Substitutes for the DativeObject in the Verb + Dative — Object + Accusative — Object", unpublished Ph.D. dissertation. University of Michigan, Ann Arbor, 1938; and Mildred K. Magers, "The Development of the Grammatical Use of Word Order for Relationships Expressed by the Accusative with Special Reference to the Development in Subordinate Clauses", unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1943.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

69

In this corpus, clause patterns admit of great variety. At first sight, despite the apparent decay of the inflectional system, the word order in clauses with subject and predicator or subject, predicator, and object seems to be highly irregular. There were indeed various combinations of the primary tagmemes, and they occurred in different clause patterns. Even though the tagmemes appeared in practically every position, however, the corpus displayed tendencies to certain patterns and showed that some of them were distinctly preferred. In the immediately preceding chapter, the purpose was to isolate and identify the primary functional slots and to describe their fillers. The present chapter deals mainly with the linear ordering of primary tagmemes within clauses. In order to develop a coherent account of that ordering, clauses have been classified into the following eight groups, on the basis of the particular combinations of the primary tagmemes they contained: 6 Group Group Group Group Group Group Group Group

1 2 3 4 5 6 7 8

(Subject, Predicator) (Subject, Predicator, Direct Object) (Subject, Predicator, Subject Complement) (Subject, Predicator, Indirect Object, Direct Object) (Predicator) (Predicator, Direct Object) (Predicator, Subject Complement) (Predicator, Indirect Object, Direct Object)

They have been further classified according to the position of the primary tagmemes in linear order. A detailed report will be presented on these eight groups and the various order-patterns found in them. The discussion will also cover such matters as the influence of slot-fillers and accompanying secondary tagmemes on linear ordering. Frequency counts of the clause patterns will be reported, and examples from the corpus will be regularly presented. Preliminary to the presentation of the data, however, are several points about the procedure in classification that need explanation. The first concern about the procedure in classification has to do with the distinction made between subjectless clauses and coordinated predications. Synthetic languages which have a complete and distinctive set of inflectional endings for all persons, genders, numbers, tenses, and moods can actually do away with subject pronouns. That is, the subject is contained in the inflectional ending of the finite verb. The samples of early English that have survived show that the Old English verb could not perform this task in all instances, even in the ninth century when inflection was fullest. Consequently, the subject pronoun was used most of the time. Yet, since there were patterns without a subject (that were not coordinated predications or imperatives), 5

Note that the listing of tagmemes within parentheses must not be interpreted as linear order, but simply as membership. Groups 5 through 8 are subjectless clauses. The coordinated predication, which will be treated separately, is not included in these groups.

70

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

the scribes must have felt that some statements following other statements with or without means of connection were sufficiently clear without pronouns and, therefore, left them out, as in this example: Jjaet fir hi seagon in öe daei rime and Ixste swa lange pet hit wies liht ofer call. 'That fire they saw at daybreak and (it) lasted so long that it was light everywhere.' (250. 22-23) Here, the stretch of words beginning with and Ixste was analyzed as a subjectless clause. It is treated as such because there is disparity in its formal reference, since the action of the verb relates to the object, pxt fir, and not to the subject, hi. If the action of the verb of the stretch of words beginning with and had related to the subject, it would have been treated as a coordinated predication such as is found in this example: Da ferde se kyng Jjenen to Portesmuöe. y Ixi £>aere eall ofer Pentecoste wuce. 'Then the king went from there to Portsmouth and stayed there all of Pentecost week.' (253.5-6) In this example, the subject of hei is not within the boundaries of Ixi (as was true also of Ixste), but Ixi is in formal as well as logical concord with the subject of the preceding clause. Its subject reference is unmistakably clear. Therefore, all units visibly joined by a conjunction and clearly in both logical and formal concord with a subject in the preceding clause without disparity were in this study considered coordinated predications. Units with or without a coordinating conjunction and without a clear formal reference to an expressed subject in the preceding clause were treated as subjectless clauses. Another different example may be instructive: Siööon J)a com se kyng to Englaland innon heruest. y se eorl com mid him. y wurdon pa alswa gode freond swa hi wxron xror feond. 'Afterwards then came the king to England in the fall, and the count came with him, and (they) became then just as good friends as they had been enemies before.' (259.28-30) In this example the unit beginning with 7 wurdon was analyzed as a subjectless clause because it does not have an overt subject and there is disparity in number between the verb wurdon and the subjects in the two preceding clauses to which it relates.® Subjectless clauses like those above were included in the frequency count of independent clauses. There were, however, some dependent clauses that were subjectless. All these were adjectival clauses without a relative pronoun, but with a subject complement as shown in this example: ... t>a geaf he Jjone biscoprice of Lincolne an clerc Alexander wxs gehaten. '... then he gave the bishopric of Lincoln to a clerk (who) was called Alexander.' (253.1-2) Another point concerns predicators. It must be kept in mind that the predicator * This is not to say that this structure was not clear and natural enough to the listeners and readers of the period.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

71

(indicated by V in the pattern labels) refers to the finite verb form. This is important because in complex verbs the finite form and non-finite form were not always adjacent. As a matter of fact, not only did adverbials come between them, but also objects and subjects as found in the following examples : (1) ... y let hine beran ham to his inne. '... and (he) had him carried home to his lodge.' (251.10) (2) On J>es ilces geares let se kyning nimert his broöer Rotbert fram t>one biscop Roger of Saeresbyri .... O n this same year the king had his brother Robert taken from the bishop, Roger of Salisbury ....' (256.19-20) Example 1 was analyzed as a coordinated predication with a VO pattern and example 2 as a clause with a VSO pattern. There are two more things that need to be said about predicators. Consider the two following examples: (1) y he for to. Rome, y pxr wxs wxl underfangen fram pe pape Eugenie. 'And he went to Rome and was received there well by Pope Eugene.' (265.19-20) (2) y feole dwild wearen geseogen y geheord. 'And many illusions were seen and heard.' (250.11-12) In the first, the stretch of words he for to Rome was analyzed as an independent clause (SV pattern), and the rest of the T-unit was treated as a coordinated predication (V pattern). The second example was analyzed as an independent clause (SV pattern) with a coordinated elliptical verb. The elliptical verb geheord, however, did not enter into the frequency count of the structures of predication. Still another concern about the procedure of classification has to do with noncontiguous compound subjects. In Chapter 2 the variety of compound subjects was illustrated. In the following example 7 sone pe cosan abbot ferde 7 te muneces mid him to Oxenforde to t>e king. 'And soon the chosen abbot went (and with him the monks) to Oxford to the king.' (268.33-34) only the italicized portion was considered in the classification of the clause. The clause was classified as an SV pattern. A fourth point that must be kept in mind is that a dependent clause was recognized as a possible filler of subject and direct object slots. For instance, in these three examples (1) ... t>et hi mosten cesen of clerc hades man swa hwam swa swa hi wolden to ercebiscop. ..that they might elect from the secular clerks whomsoever they wanted as archbishop.' (251.23-24) (2) ... Jja bed se cyng heom pxt hi scolden cesen hem asrcebiscop to Cantwarabyrig.... '... then the king ordered them that they should elect themselves an archbishop to Canterbury ....' (251.18-19) (3) Fela soöfeste men saeidon pet pxr wxron manege mid micel unrihte gespilde.

72

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

'Many trustworthy men said that there were many mutilated with great injustice.' (254.34-35) there are these patterns in the order of their occurrence: SVO, VSIO, and SVO. The dependent clauses filling slots in independent clauses were, of course, later classified and entered into the frequency count according to the presence and position of the primary tagmemes within their boundaries. In the same examples, the patterns of the dependent clauses were classified as OSV, SVIO, and VS. The final point about the procedure in classification that needs to be emphasized is that in patterns with a series of objects such as is found in y J>ar he nam j>e biscop Roger of Sereberi y Alexander biscop of Lincol y te Canceler Roger hise neues .... 'And there he took Roger, bishop of Salisbury; Alexander, bishop of Lincoln; and the chancellor Roger, his nephew ....' (263.30-31) only the first object was counted. This particular pattern was classified SVO. Now that these special features in the procedure of classification have been taken care of, the rest of the chapter will deal with a report of the findings. There were 484 T-units in the corpus. This is equivalent to saying that there were 484 independent clauses. Dependent clauses numbered 279. Overt subjects appeared in 441 (91 per cent) of the independent clauses and 261 (93 per cent) of the dependent clauses. In the other clauses, subjects, though implied, were not expressed. In this discussion, clauses in which expressed subjects appear will be called SUBJECTFUL; those in which they are not expressed will be called SUBJECTLESS. Counting the 279 dependent clauses embedded in or attached to the 484 T-units, there was a total of 763 clauses, 702 (92 percent) of which were subjectful and sixty-one (8 percent) subjectless. The classification and frequency counts of the clause patterns are presented abstractly in the tables on page 73 through 81. The table headings are self-explanatory. The Arabic numbers in the first column stand for the groups, and the capital letters stand for the primary tagmemes: S (subject), V (predicator), Ο (direct object), I (indirect object), C (subject complement). The capital S enclosed by parentheses means 'subjectless'. The arrangement of the symbols into formulas represents the linear ordering of the primary tagmemes or the clause patterns. The percentages are to the nearest whole per cent, and the minus sign next to the numeral 1 means 'less than'. The numerical and percentage breakdown in Table IV is based on the total number of 763 clauses. Tables V and VI show the distribution and frequency of the two kinds of clauses: independent and dependent. Tables VII and VIII show the distribution of clauses in relation to the First Continuation and the Final Continuation. Tables IX, Χ, XI, and XII are based on the number of independent and dependent clauses in the First and Final Continuations respectively. Table IV indicates that of the 763 clauses, 385 (51 per cent) belonged to Group 1 (subject and predicator), 229 (30 per cent) to Group 2 (subject, predicator, and object),

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

73

TABLE IV

Total Frequencies of Clause Groups and Patterns Total Number Clauses : 763 Group

Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

SV vs

257 128

67 33

34 17

385

51

svo sov vso vos osv ovs

117 38 31 2 34 7

51 17 13 1 15 3

15 5 4 -1 4 1

229

30

3

SVC scv vsc CSV CVS

35 4 4 2 1

76 9 9 4 2

5 1 1 -1 -1

46

6

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

15 2 6 1 6 2 2 2 6

36 5 14 2 14 5 5 5 14

2 -1 1 -1 1 -1 -1 -1 1

42

6

5

(S)V

13

100

2

13

2

6

(S)VO (S)OV

21 4

84 16

3 -1

25

3

7

(S)CV (S)VC

15 2

88 12

2 -1

17

2

fS)VIO (S)VOI (S)IVO (S)OIV

1 2 2 1

17 33 33 17

-1 -1 -1 -1

6

1

Pattern 1

2

8

74

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE V

Frequencies of Independent Clause Groups and

Patterns

Total Number Clauses: 484 uroup

Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

126 122

51 49

26 25

248

51

SVO SOV VSO VOS osv ovs

67 24 28 2 15 7

47 17 20 1 10 5

14 5 6 -1 3 1

143

30

3

SVC scv vsc CSV CVS

14 0 3 2 1

70 0 15 10 5

3 0 1 -1 -1

20

4

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

8 2 6 1 6 2 2 1 2

27 7 20 3 20 7 7 3 7

2 -1 1 -1 1 -1 -1 -1 -1

30

6

5

(S)V

13

100

3

13

3

6

(S)VO (S)OV

21 3

88 12

4 1

24

5

7

(S)CV (S)VC

0 2

0 100

0 -1

2

-1

(S)VIO (S)VOI (S)IVO (S)OIV

1 2 1 0

25 50 25 0

-1 -1 -1 0

Pattern 1

2

8

SV VS

4

1

75

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE VI

Frequencies of Dependent Clause Groups and Patterns Total N u m b e r Clauses: 279 Group Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

131 6

96 4

47 2

137

49

50 14 3 0 19 0

58 16 4 0 22 0

18 5 1 0 7 0

86

31

81 16 4 0 0

7 1 -1 0 0

26

9

CSV CVS

21 4 1 0 0

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

7 0 0 0 0 0 0 1 4

58 0 0 0 0 0 0 8 33

3 0 0 0 0 0 0 -1 1

12

4

5

(S)V

0

0

0

0

0

(S)VO (S)OV

0 1

0 100

0 -1

1

-1

(S)CV (S)VC

15 0

100 0

5 0

15

5

0 0 1 1

0 0 50 50

0 0 -1 -1

2

1

Pattern

1

2

SV VS

svo sov vso vos osv ovs SVC

3

6 7

scv vsc

(S)VIO 8

(S)VOI (S)IVO (S)OIV

76

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE VII

Frequencies of Clause Groups and Patterns

in First

Continuation

Total Number Clauses: 471 Group

Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

SV vs

155 85

65 35

33 18

240

51

svo sov vso vos osv ovs

70 22 19 1 21 2

52 16 14 1 16 1

15 5 4 -1 4 -1

135

29

3

SVC scv vsc CSV CVS

25 0 2 0 1

89 0 7 0 4

5 0 -1 0 -1

28

6

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

10 1 6 1 6 2 2 1 2

32 3 19 3 19 7 7 3 7

2 -1 1 -1 1 -1 -1 -1 -1

31

7

5

(S)V

6

100

1

6

1

6

(S)VO (S)OV

9 4

69 31

2 1

13

3

7

(S)CV (S)VC

13 1

93 7

3 -1

14

3

(S)VIO (S)VOI (S)IVO (S)OIV

1 2 1 0

25 50 25 0

-1 -1 -1 0

4

1

Pattern 1

2

8

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

77

TABLE VIII Frequencies

of Clause Groups and Patterns

in Final

Continuation

Total Number Clauses: 292 oroup

Number Pattern

Per Cent Group

102 43

70 30

35 15

SVO SOV VSO VOS OSV

ovs

47 16 12 1 13 5

50 17 13 1 14 5

16 5 4 -1 4 2

3

SVC scv vsc CSV CVS

10 4 2 2 0

56 22 11 11 0

3 1 1 1 0

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSV1

5 1 0 0 0 0 0 1 4

45 10 0 0 0 0 0 9 36

2 -1 0 0 0 0 0 -1 1

5

(S)V

7

100

6

(S)VO (S)OV

12 0

7

(S)CV (S)VC (S)VIO (S)VOI (S)IVO (S)OIV

Pattern

1

2

8

SV VS

Per Cent Total

Number Group

Per Cent Total

145

50

94

32

18

6

11

4

2

7

2

100 0

4 0

12

4

2 1

67 33

1 —1

3

1

0 0 1 1

0 0 50 50

0 0 -1 -1

2

1

78

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE IX

Frequencies of Independent

Clause Groups and Patterns in First

Continuation

Total Number Clauses: 296 uroup

Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

SV VS

76 80

49 51

26 27

156

53

SVO SOV VSO VOS osv ovs

35 12 17 1 10 2

45 16 22 1 13 3

12 4 6 -1 3 1

77

26

3

SVC scv vsc CSV CVS

12 0 2 0 1

80 0 13 0 7

4 0 1 0 -1

15

5

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

5 1 6 1 6 2 2 1 1

20 4 24 4 24 8 8 4 4

2 -1 2 -1 2 1 1 -1 -1

25

8

5

(S)V

6

100

2

6

2

6

(S)VO (S)OV

9 3

75 25

3 1

12

4

7

(S)CV (S)VC

0 1

0 100

0 -1

1

-1

8

(S)VIO (S)VOI (S)IVO (S)OIV

1 2 1 0

25 50 25 0

-1 1 -1 0

4

1

Pattern 1

2

79

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE X

Frequencies of Independent Clause Groups and Patterns in Final Continuation Total Number Clauses: 188 Group Pattern

Number Pattern

Per Cent Group

Per Cent Total

vs

50 42

54 46

27 22

svo sov vso vos osv ovs

32 12 11 1 5 5

48 18 17 2 8 8

17 6 6 1 3 3

CSV CVS

2 0 1 2 0

40 0 20 40 0

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOl I SVO OSVI

3 1 0 0 0 0 0 0 1

5

(S)V

6

7

1

2

8

Per Cent Total

92

49

66

35

1 0 1 1 0

5

3

60 20 0 0 0 0 0 0 20

2 1 0 0 0 0 0 0 1

5

3

7

100

4

7

4

(S)VO (S)OV

12 0

100 0

6 0

12

6

(S)CV (S)VC

0 1

0 100

0 1

1

1

(S)VIO (S)VOI (S)IVO (S)OIV

0 0 0 0

0 0 0 0

0 0 0 0

0

0

SV

SVC 3

Number Group

scv vsc

80

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE XI

Frequencies of Dependent Clause Groups and Patterns in First Continuation Total Number Clauses: 175 uroup Pattern

Number Pattern

Per Cent Group

Per Cent Total

SV VS

79 5

94 6

45 3

svo sov vso vos osv ovs

35 10 2 0 11 0

60 17 3 0 19 0

20 6 1 0 6 0

SVC

CSV CVS

13 0 0 0 0

100 0 0 0 0

7 0 0 0 0

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

5 0 0 0 0 0 0 0 1

83 0 0 0 0 0 0 0 17

3 0 0 0 0 0 0 0 1

5

(S)V

0

0

6

(S)VO (S)OV

0

1

(S)CV (S)VC (S)VIO (S)VOI (S)IVO (S)OIV

1

2

3

4

8

scv vsc

Number Group

Per Cent Total

84

48

58

34

13

7

6

3

0

0

0

0 100

0 1

1

1

13 0

100 0

7 4

13

7

0 0 0 0

0 0 0 0

0 0 0 0

0

0

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

81

TABLE XII

Frequencies of Dependent Clause Groups and Patterns in Final Continuation Total Number Clauses: 104 Group Pattern

Number Pattern

Per Cent Group

Per Cent Total

Number Group

Per Cent Total

SV VS

52 1

98 2

50 1

SVO SOV VSO

15 4 1 0 8 0

54 14 4 0 29 0

14 4 1 0 8 0

SVC

CSV CVS

8 4 1 0 0

62 31 8 0 0

8 4 1 0 0

4

SVIO SVOI SIVO SIOV VSIO SOIV VSOI ISVO OSVI

2 0 0 0 0 0 0 1 3

33 0 0 0 0 0 0 17 50

2 0 0 0 0 0 0 1 3

6

6

5

(S)V

0

0

0

0

0

6

(S)VO (S)OV

0 0

0 0

0 0

0

0

7

(S)CV (S)VC

2 0

100 0

2 0

2

2

(S)VIO (S)VOI (S)IVO (S)OIV

0 0 1 1

0 0 50 50

0 0 1 1

2

2

1

2

3

8

vos osv ovs scv vsc

53

51

28

27

13

12

82

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

forty-six (6 per cent) to Group 3 (subject, predicator, and subject complement), and forty-two (6 per cent) to Group 4 (subject, predicator, indirect object, and direct object). Sixty-one (9 per cent) of the 763 clauses were subjectless and were distributed among Groups 5, 6,7, and 8. In the general discussion of the order and co-occurrence of primary tagmemes and their outstanding features, subjectless patterns will not be included because in these matters of word order only patterns with overt subjects are of real significance. They, nevertheless, will be presented separately and brought into the discussion of the order of individual primary tagmemes toward the end of the chapter. 3.1. GROUP 1 (SUBJECT AND PREDICATOR)

With regard to the relative linear order of subject and predicator, there are two kinds of word order possible: the subject-predicator order (SV), which is called the direct order, 7 and the predicator-subject order (VS), which is called the inverted order. Table IV on page 73 shows that 385 (51 per cent) of the total number of clauses belonged to Group 1. Two hundred fifty-seven (67 per cent) of these appeared with the subject before the predicator (SV) and 128 (33 per cent) with the predicator before the subject (VS). Before anything else is said about the order of these tagmemes, consider the following examples: SV (independent): (1) y on Pasches he weas on Norht hamtune. 'And on Easter he was at Northampton.' (250.2) (2) y feole dwild wearen geseogen y geheord. 'And many illusions were seen and heard.' (250.11-12) (3) y sex com to J)e iunge eorl Henri. 'And she came to the young count, Henry.' (268.5-6) (4) y hi swa diden. 'And they did thus.' (251.17) SV (dependent): (1) ... pe ]jaer binnen wseron .... '... which were inside ....' (250.7) (2) ... for se biscop of Szresbyrig wxs Strang .... '... because the bishop of Salisbury was strong ....' (251.31) (3) f>a he to Engleland com .... 'When he came to England ....' (268.25) VS (independent): (1) y t>asr comen pes eorles sander men of Angeow to him. 'And the messengers of the count of Anjou came there to him.' (251.2-3) 7 Included here is the S ... V order, in which an adverbial tagmeme came between the subject and predicator. This order occurred three times in independent clauses (after y, ac) and twenty-three times in dependent clauses.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

83

(2) Dis gaere for pe king Stephne ofer see to Normandi .... 'This year King Stephen went across the sea to Normandy ....' (263.24) VS (dependent): (1) ... t>et t>aer wxron manege mid micel unrihte gespilde. '... that there were many mutilated with great injustice.' (254.34-35) (2) ... J>et fraer mihte wel ben abuton twenti oder pritti hornblaweres. '... that there might well be about twenty or thirty hornblowers.' (258.25-26) Tables V and VI show that 248 (64 per cent) of the subject and predicator examples occurred in independent clauses and that 137 (36 per cent) occurred in dependent clauses. In the independent clauses, the distribution among SV and VS patterns was just about even. From Table V it is seen that the SV pattern occurred 126 times (51 per cent) and that the VS pattern occurred 122 times (49 per cent). For dependent clauses, Table VI shows that the SV pattern was the predominant form, appearing 131 times (96 per cent) out of 137 occurrences of the subject and predicator group. This, it seems, can be readily explained. In Chapter 2 it was stated that the most frequent filler of the subject slot of the 702 clauses was the pronoun. Fifty per cent of the pronoun subjects occurred in dependent clauses. The preference for the placing of the pronoun in the initial position in Classical Old English is generally recognized.8 To account for this patterning in independent clauses is more difficult. This matter will be discussed at length, just a little later, in the paragraphs on inversion. These same 385 clauses divided on the basis of their appearance in the First and Final Continuations are worth noting. See Tables VII and VIII. In the First Continuation, there were 240 subject and predicator clauses, with 155 (65 per cent) in the SV order and eighty-five (35 per cent) in the VS order. In the Final Continuation, 102 (70 per cent) of the 145 clauses occurred in the SV order and forty-three (30 per cent) in the VS order. A separate breakdown into independent clauses and dependent clauses sheds some light on the predominance of the SV pattern over the VS pattern. Tables IX through XII show why there was an almost two-to-one ratio in the First Continuation and a better than two-to-one ratio in the Final Continuation. First of all, Tables IX and X indicate that there was no great significance in the difference of the frequency of occurrence of SV and VS patterns in independent clauses. Next, Tables XI and XII, a breakdown of dependent clauses, reveal that the SV pattern was the characteristic order of dependent clauses. In the First Continuation, it occurred in seventy-nine (94 per cent) out of eighty-five examples; in the Final Continuation, in fifty-two (98 per cent) out of fifty-three examples. This is in accord with other recent studies (referred to several times before) which demonstrated the strong preference for the SV order in dependent clauses. 8

See, for example, C. A. Smith, "The Order of Words in Anglo-Saxon Prose", PMLA, (1893), 221.

VIII

84

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

3.1.1. Inversion of Subject and Predicator The most important feature of the study of direct (SV) and inverted (VS) orders is their distribution in independent clauses. At this point it may be helpful to restate their frequency of occurrence. In the First Continuation, the VS order was more frequent, occurring eighty times (51 per cent) out of the total of 156. In the Final Continuation, the SV order predominated in a ratio of 50 (54 per cent) to 42 (46 per cent). Added together these figures show that the SV order was just slighty more frequent than the VS order: 126 (51 per cent) to 122 (49 per cent). The point of interest centers on the strong persistence of the VS order. 9 A partial explanation is to be sought in the environment of these patterns, which apparently reveals, in the light of the arbitrariness of the inflectional system and the presence of modern analytic constructions in this corpus, an almost conscious attempt at preserving certain formulaic syntactic patterns from earlier Old English, which caused, with high predictability, inversion of subject and predicator.10 3.1.1.1. Inversion and fillers of the subject slot. - Since the pronoun as filler of the subject slot was frequently placed in the initial position in the relative order of the subject and predicator in dependent clauses, the question of its force in the determination of SV and VS orders in independent clauses arises. Table XIII just below shows that the noun filled the subject slot of Group 1 in 145 examples, occurring in the SV order forty-five times (31 per cent) and in the VS order 100 times (69 per cent). Also indicated is the fact that the subject slot was TABLE XIII

The Relationship Between the Noun and Pronoun as Fillers of the Subject Slot and the SV and VS Orders in Independent Clauses First Continuation Order

SV VS

Noun

Final Continuation

Pronoun

Noun

Total

Pronoun

Noun

Pronoun

No.

%

No.

%

No.

%

No.

%

No.

%

No.

%

23 68

25 75

53 12

82 18

22 32

41 59

28 10

74 26

45 100

31 69

81 22

79 21

* Cecily Clark, in her The Peterborough Chronicle 1070-1154 (Oxford, Oxford University Press, 1958), p. lxvi, was misleading in her reference to Rothstein's study, "Die Wortstellung in der Peterborough Chronik", Studien zur englischen Philologie, LXIV (1922), 108, as Rothstein was himself, when she wrote that the VS order was frequently abandoned after introductory adverbial phrases. 10 This observation was made independently, but other students of the corpus have been found to have made similar remarks. See, for example, Cecily Clark, op. cit., p. lvii.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

85

filled by a pronoun in 103 examples and that it occurred in the SV order eighty times (79 per. cent) and in the VS order twenty-two times (21 per cent). At first sight, these percentages seem significant enough to say that the subject slot filled by the pronoun was, as a rule, placed in the SV order and that inversion was more likely to take place when the subject was filled by a noun. Yet, it would be dangerous to make such a generalization, for the evidence is inconclusive in that this breakdown shows nothing about the presence or absence of adverbial tagmemes in the patterns. 3.1.1.2. Inversion and adverbial and connecting tagmemes. - It is well known that Classical Old English word order was strongly influenced by the presence of elements not associated with the subject in the initial position in independent clauses. This was especially true of adverbials. Shannon found it necessary to distinguish between initial and sequence clauses, perhaps because of the frequent use of such a formulaic adverbial as her, which almost invariably was followed by the inverted order (VS).11 This word and other formulaic patterns such as on pyssum geare were not as numerous in this corpus. There was only one use of her, and this occurred in a sequence clause (not in the initial clause of the annal entry). Here is the example: Her him trucode ealle his mycele craeftes. 'At this time all his great trickery failed him.' (262.4-5) The other formulaic phrase (or its variations) occurred thirty-six times. Sixteen of the examples occurred as initial adverbials in the clauses beginning the sixteen annal entries. A typical example is as follows: On pyssum geare waes se king Henri on Cristes tyde set Dunestaple. 'In this year King Henry was at Dunstable at Christmas.' (251.1-2) All the thirty-six phrases except one were followed by the inverted order (VS). That these formulaic patterns and certain other adverbials had inverting force on the subject and predicator is apparently confirmed by the figures in Table XIV on page 86. Since there was not a substantial difference in the percentages between the First and Final Continuations, only the totals will be read. An examination of Table XIV reveals the degree of the predominance of the VS order when an adverbial filled the initial position. (In the table, zero indicates that no other elements preceded the primary tagmemes.) The figures in Table XIV indicate that the VS order was found far more frequently than the SV order when adverbials occurred in the initial position of the clause. The subject and predicator group occurred 143 times with adverbials as initial elements. In these 143 examples, the VS order appeared 112 times (71 per cent) and the SV order 31 times (29 per cent). The relatively high frequency (compared with Classical Old English) of the SV order after initial adverbial elements may be explained if element 4 (adverb) is examined more 11

Shannon, op. cit., p. 47.

86

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES TABLE XIV

The Relationship Between the Presence of Adverbial and Connecting Tagmemes in Initial Position and the SV and VS Orders in Independent Clauses First Continuation SV

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Zero Conj. Neg. Adv. Mult. Adv. Noun Ph. Prep. Ph. Adv. CI. Part. Ph. Corr. CI. Totals

Final Continuation

VS

Totals

VS

SV

SV

VS

No.

%

No.

V /o

No.

%

No.

%

No.

%

No.

%

28 30 0 13 0 0 4 0 1 0

100 94 0 39 0 0 13 0 100 0

0 2 1 20 12 12 26 0 0 7

0 6 100 61 100 100 87 0 0 100

12 25 0 7 0 2 2 1 0 1

75 93 0 35 0 50 17 50 0 11

4 2 1 13 1 2 10 1 0 8

25 7 100 65 100 50 83 50 0 89

40 55 0 20 0 2 6 1 1 1

91 93 0 38 0 12 14 50 100 6

4 4 2 33 13 14 36 1 0 15

9 7 100 62 100 88 86 50 0 94

76

80

50

42

126

122

closely. Out of 53 occurrences of the adverb in the initial position in the subject and predicator group, the adverb was followed by the VS order thirty-three times (62 per cent) and by the SV order twenty times (38 per cent). Included in the frequency count of the twenty adverbs followed by the SV order were sixteen with adverbs such as pa, siddan, penon, swa, which were preceded by a coordinating conjunction (y, and) as in this example: y siddan Balduin acordede. 'And afterwards Baldwin agreed.' (263.19-20) S. O. Andrew, in his study of word order, observed that the clauses like the one just presented were not exceptions to the rule that inverted order was the usual order when adverbs were in the initial position.12 This is true, he explained, because the conjunction initiates the clause and the adverb is internal. The practice of not inverting the subject and predicator when such a pattern initiated the clause was in force in this corpus. For instance, in lines 23-30 on page 255,13 there were three close occurrences of the conjunction plus adverb (y siddon, 2; y peonon), all of which were followed by the SV order. On page 257 (lines 23-32) there was one example of y siddon (line 24) with the SV order and four examples with the adverb alone as initial element, all of which were followed by the VS order. If this formula can be considered 12

S. O. Andrew, "Some Principles of Old English Word Order", Medium yEvum, III (1934), 167-168. 13 The line and page references are to John Earle and Charles Plummer, Two of the Saxon Chronicles Parallel (London, Oxford University Press, 1892), I.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

87

operative (as it certainly seems to be), then the high percentage of the SV order with adverbs preceding the primary tagmemes can be accounted for. It is apparent then that in clauses where modifying elements not associated with the subject began the clause, the VS order was almost predictable.14 No practical purpose would be served by explaining all of the data in Table XIV, other than in a general way. It shows that adverbial tagmemes, as a rule, caused inversion of subject and predicator, especially so if they were multiple combinations and parts of correlative constructions.15 This statement is supported by the occurrence of the SV order 91 per cent of the time (40 out of 44) when there were no adverbial tagmemes in the initial position. Furthermore, the SV order occurred 93 per cent of the time (55 out of 59) when a coordinating conjunction was the initial element. Whether inversion should be characteristic of the English language in this period is neither here nor there. Whether it was in actual speech will probably never be known. Perhaps it was not. Nevertheless, inversion is characteristic of this corpus. The reason for its force probably lies, as was suggested earlier, in the scribes' feeling the force of the style and of certain formulaic patterns of manuscripts of earlier periods, even at this advanced stage of the English language.16 Now, it remains to go back to investigate the interrelationships of nouns and pronouns as fillers of the subject slots in the subject and predicator group with regard to the presence of adverbial and connecting tagmemes in the initial position. The analysis in Table XV seems to indicate, if anything, that inversion of subject and predicator was likely to occur if an adverbial tagmeme occurred initially, regardless of whether the subject slot was filled by a noun or pronoun. It also indicates that without adverbial tagmemes in the initial position the order was likely to be SV, no matter what kind of filler the subject had. All in all, then, the SV was the most frequent order. Inversion was strongly represented in the subject and predicator group and, apparently, took place, as a rule, when adverbials (words, phrases, correlatives) preceded the primary tagmemes. The

14 Sec footnote 8. A failure to recognize this apparently fixed rule may have been one of the reasons why Rothstein was led to state that the practice of inversion after adverbial phrases was almost abandoned. 15 C. A. Smith, op. cit., p. 223, states that the inverted order (VS) in correlatives such as:

t>a hwile f>e J^a muneces sungen J^aere messe ...pa com se fir on ufen weard Jxme stepel. 'While the monks were singing the mass ... then the fire came to the upper part of the tower.' (250.4-5) was not caused by the preceding adverbial clause, but by the second pa in the construction {pa ...pa). The adverbial clauses (element 8) in the table are not correlative, but just adverbial clauses preceding the main clauses. In one instance the order of the main clause was SV; and in the other, it was VS. " There are several discussions about when Old English ended and Middle English began in general, and about this corpus in particular. For example, see Kemp Malone, "When Did Middle English Begin?" Curme Volume of Linguistic Studies, Language Monographs, VII (1930), 110-17 and F. P. Magoun, "Colloquial Old and Middle English", Harvard Studies and Notes in Philology and Litera· ture, XIX (1937), 167-73.

88

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

figures rather consistently show that when adverbial modifiers came first, the predicator was drawn with them, and the subject was forced to follow.17

TABLE x v

The Relationship Between the Presence of Adverbial and Connecting Tagmemes in the Initial Position and the SV and VS Orders with Nouns (N) and Pronouns (P) in the Subject Slot in Independent Clauses First Continuation

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Zero Conj. Neg. Adv. Mult. Adv. Noun Ph. Prep. Ph. Adv. CI. Part. Ph. Corr. CI. Totals

Totals

Final Continuation

SV(N) VS(N) SV(P) VS(P) SV(N)|vS(N)jSV(P)

VS(P) SV(N) VS(N) SV(P) VS(P)

1 21 0 1 0 0 0 0 0 0

0 1 1 13 12 12 26 0 0 3

27 9 0 12 0 0 4 0 1 0

0 1 0 7 0 0 0 0 0 4

2 13 0 4 0 2 1 0 0 0

3 2 0 10 1 2 10 1 0 3

10 12 0 3 0 0 1 1 0 1

1 0 1 3 0 0 0 0 0 5

3 34 0 5 0 2 1 0 0 0

3 3 1 23 13 14 36 1 0 6

37 21 0 15 0 0 5 1 1 1

1 1 1 10 0 0 0 0 0 9

23

68

53

12

22

32

28

10

45

100

81

22

3.2. GROUP 2 (SUBJECT, PREDICATOR, AND DIRECT OBJECT)

Group 2 (subject, predicator, and direct object) constituted 30 per cent of the total number of clauses, which means that it occurred 229 times among 763 clauses. It is obvious from Table IV that the subject, predicator, and direct object occurred in every possible arrangement, but that there was a greater frequency of the SVO order, which appeared in 117 examples, 51 per cent of the 229 clauses with an overt subject. Since none of the other patterns had more than 17 per cent of the total, the SVO pattern apparently was normal. The next most frequent pattern was the SOV, of which there were thirty-eight examples (17 per cent). Close to this pattern in frequency were the OSV with thirty-four examples (15 per cent) and the VSO with thirty-one examples (13 per cent). The OVS pattern occurred seven times (3 per cent). The VOS pattern, the one that G. H. Vallins in his The Patterns of English described as one of the three standard patterns in Old English, occurred only twice, less than 1 per cent of the time.18 17 This is the general rule except for words such as pa, siddon, pair, and penon when preceded by a coordinator. " G. H. Vallins, The Patterns of English (London, Andre Deutsch, 1956), p. 56.

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

89

Table V on page 74 shows that 143 of these patterns (62 per cent) occurred in independent clauses, and Table VI on page 75 indicates that eighty-six (38 per cent) occurred in dependent clauses. Tables V and VI, furthermore, reveal that the SVO order was the most frequent in both kinds of clauses. A closer examination of these tables also shows that there was no appreciable difference in the occurrence of these patterns, excepting the VSO perhaps, in independent and dependent clauses, not enough at least to make any significant and solid conclusions about. The same can be said about the distribution of Group 2 in both the First and Final Continuations (see Tables VII and VIII). All these analyses significantly show that there was a general preference for the SVO order. A position analysis reveals that the subject stood before the predicator in 189 examples (83 per cent) and before the object in 186 (81 per cent). The subject stood in the first position 135 times (59 per cent), in the second sixty-five times (28 per cent), and in the last nine times (13 per cent). The predicator occupied the second position 124 times (54 per cent), the first thirty-three times (14 per cent), and the last seventytwo times (32 per cent). The object occurred last 148 times (65 per cent), first fortyone times (18 per cent), and second forty times (17 per cent). Despite the various positions in which the three primary tagmemes occurred, the subject was preferably placed in the first position, the predicator in the second, and the object in the last. This distribution of the patterns certainly shows a tendency to positions that are characteristic of Modern English. Since the remaining discussion of Group 2 will be concerned chiefly with the various patterns with nouns and pronouns as fillers of the nominal slots, it may be appropriate here to say a little about the frequency of occurrence of clauses and infinitives in this group. Eighteen of the examples of Group 2 occurred with clauses as fillers of the direct object slots. Thirteen of these occurred in the SVO pattern, with nine appearing in independent clauses and four in dependent clauses. The other five were VSO, all of which appeared in independent clauses. As would be expected, direct object slots filled by clauses appeared only in the last position. In the one instance that the infinitive phrase occurred as filler of the direct object slot, the pattern was

vos.

The following examples were selected to show the variety of fillers of the primary functional slots as well as the order in which they patterned: SVO (independent): (1) man ferode hine to Lincolne mid micel wuröscipe. 'They took him to Lincoln with great honor.' (251.11) (2) Ac pet ofer com Rome J>et of er cumed eall weoruld .... 'But that which overcomes all the world overcame Rome ....' (252.29-30) (3) hi sieden openlice pet Xpist slep .... 'They said openly that Christ slept ....' (265.9-10) (4) Se ilce Willelm hefde numen Fulkes eorles gingre dohter to wife of Angeow. 'This

90

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

same William had taken as wife the daughter of Fulk, the count of Anjou.' (254. 7-8) SVO (dependent): (1) f>e hwile J» pa munecas sungen psere messe .... 'While the monks were singing the mass ....' (250.4) (2) ... forjM Jiet he uuolde underpeden pet mynstre to Clunie .... '... because he wanted to subject the monastery to Clunny ....' (262.21-22) (3) ... forjri J>et hi uuendenpet he sculde ben alsuic alse the eom wes .... '... because they expected that he was going to be just as his uncle had been ....' (263.25-26) SOV (independent): (1) (2) (3) (4)

y eallepa biscopas him underfengen. 'And all the bishops received him.' (252.2-3) he hine ledde to Cantwarabyrig. 'He led him to Canterbury.' (255.21) oc Xpist it ne uuolde. 'But Christ did not want it.' (262.27) ac hi nan treuthe ne heolden. 'But they kept no pledge.' (263.35)

SOV (dependent): (1) ...paealldigelnessesedy wat.... '... who sees and knows all secrets ....' (254.35) (2) ... for his men him suyken .... '... because his men betrayed him ....' (266.24-25) (3) y se pe pet ne wolden done .... 'And whoever would not do that ....' (260.3-4) VSO (independent): (1) Da sone Jsaer aefter sende se kyng hise write ofer eall Englalande. 'Then soon after that the king sent his writs all over England.' (251.14-15) (2) On J)es ilces gaeres sende se papa of Rome to öise lande an cardinal Iohan of Creme was gehaten. 'In this same year the pope of Rome sent to this land a cardinal who was called John of Crema.' (255.17-18) (3) J)a t>e king Stephne to Englalande com £>a macod he his gadering at Oxenford. 'When King Stephen came to England, then he held his council at Oxford.' 263.29-30) (4) sxide se abbot of Clunni pet hi heafdon forloron Sancte IoHes mynstre {>urh him y Jjurh his mycele sotscipe. 'The abbot of Clunny said that they had lost St. John's monastery through him and through his great folly.' (262.9-10) VSO (dependent): (1) ... forjai ]set naefre ne luueden hi munece regol .... '... because they never liked the monastic rule ....' (251.27) (2) J>a namen hi pa men .... 'Then they took the men ....' (264.5-6)

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES

91

VOS (independent): y ne forstod noht ealle pa bodlaces. 'And all the orders profited nothing.' (260.8-9) VOS (dependent): no examples OSV (independent): (1) pis he dyde eall for jjes biscopes luuen. 'He did this all for the bishop's love.' (253.3-4) (2) Hugo of Mundford he sende to Engleland .... 'He sent Hugo of Mundford to England ....' (253.34) (3) Sume hi diden in crucethus Jjet is in an caeste t>at was scort y nareu. y undep .... 'They put some in a torture-house, that is, in a case that was short, narrow, and shallow ....' (264.15-16) OSV (dependent): (1) ... swa hwam swa swa hi woldon. '... whomsoever they wanted.' (251.19) (2) ... of ealle 5a £>ing pet se papa him on leide .... '... of all the things that the pope imposed upon him ....' (252.31-32) (3) ... pe hi wenden .... '... whom they expected ....' (264.6) OVS (independent): (1) y hine bebyrigde se biscop of Ceastre Rotbert Pecceö was gehaten. 'And the bishop of Chester, who was called Robert Pecceth, buried him.' (251.12-13) (2) Her him trucode ealle his mycele crxftes. 'At this time all his great trickery failed him.' (262.4-5) (3) Micel hadde Henri King gadered gold y sylver. 'King Henry had gathered much in gold and silver.' (263.27-28) OVS (dependent): no examples. 3.2.1. Nouns and Pronouns as Fillers of the Subject and Direct Object Slots The next phase of this section has to do with the question of the influence of the kinds of fillers in the subject and direct object slots on the order of the subject, predicator, and direct object in a clause. Is there any possible correlation between the combination of noun and pronoun and the patterning of the subject, predicator, and direct object in a clause? Table XVI presents the frequency of occurrence of the patterns of Group 2 according to the combination of noun and pronoun as fillers of the subject and direct object slots. The combinations of letters (NN, PP, NP, PN) represent linear order just as the patterns do. For example, the NP arrangement under independent clauses is noun followed by pronoun in linear order. The column of numbers under it indicates the number of times the noun + pronoun arrangement occurred in the six patterns of Group 2. The SOV pattern, for example, occurred in sixteen independent

Totals

THE ORDER OF PRIMARY CLAUSE-LEVEL TAGMEMES 5 oc \e Μ ^ in Ο m

Ο CS Os

CO

Ο

Ο Ο

Tf Ο Ο

Ο •«j·

»-Η

Ό 00

CO CO

1

Ο 00

cs ψ-* OS

Ο

-Η Ο Ο

Ο η

Ο Ο Ο

Ο Ο

© © * — < © —I

Ο Ο

PQ

« ο ο

-η

Ο Ο -Η Ο —

·«* Ο Ό

Ο Os

Ο Ο —

Ο >-Η CO SO CO


et is twegen sed laepas to six scillingas .... '... one sold seed wheat, that is, two seedlips, for six shillings an acre ....' (254.1415) (3) y se king hine underfeng mid micel wurdscipe. 'And the king received him with great honor.' (255.19) (4) diden him ealle hersumnesse swa swa hi scolden don here abbot. '(They) did complete obedience to him as they should do to their abbot.' (262.1)

4.1.5. The Purpose Tagmeme The purpose tagmeme, which occurred in fifteen examples, is an adverbial slot filled only by prepositional phrases. This adverbial expressed the purpose for or the beneficiary of the predication of the clause. All fifteen examples followed the predicator.

SECONDARY CLAUSE-LEVEL TAGMEMES

117

Examples: (1) J)is he dyde eall for pes biscopes luuen. 'This he did all for the bishop's love.' (253. 3-4) (2) Crist rsede for pa wrecce muneces of Burch y for J)et wrecce stede. 'Christ take counsel for the wretched monks of Peterborough and for that wretched place.' (262.17-18) 4.1.6. The Agent Tagmeme The agent tagmeme, like the purpose tagmeme, is an adverbial slot filled only by prepositional phrases. The agent tagmeme indicated either the instrument or means through which something was done or the personal agent of an action. Five (12 per cent) of the forty examples preceded the predicator, and thirty-five (88 per cent) followed the predicator. Examples: (1) ... for t>e land was al fordon midsuilce dxd.es. '... for the land was all destroyed through such deeds.' (265.8-9) (2) saelde se abbot of Clunni Jset hi heafdon forloron Sancte Iohes mynstre purh him y purh his mycele sotscipe. 'The abbot of Clunny said that they had lost St. John's monastery through him and his great folly.' (262.9-10) 4.1.7. The Relationship Tagmeme The relationship tagmeme is an adverbial slot filled by prepositional phrases denoting active opposition (him togeanes), accompaniment (mid him), separation (Jram him), or intermediacy (betwux heom), all of which expressed involvement of persons. Eighteen (19 per cent) of the ninety-seven examples stood before the predicator, and the other seventy-nine (81 per cent) stood after the predicator. Examples: (1) y se kyng held stranglice hem togeanes. 'And the king held strongly against them.' (253.12) (2) sume helden mid te king. 'Some held with the king.' (267.17) (3) y te cuen of France todaelde fra pe king. 'And the queen of France separated from the king.' (268.5) (4) He seide J>et fülle feoht was sett betwenen da Cristene y pa hedene. 'He said that a full battle was set between the Christians and heathens.' (259.20-21) 4.2. LINEAR ORDERING OF ADVERBIAL TAGMEMES

Although the purpose here is to describe the various positions of the adverbials relative to the predicator (and to a lesser degree, the other primary tagmemes), a

118

SECONDARY CLAUSE-LEVEL TAGMEMES

primary concern is to note how clause-level adverbials tended to be placed when one or more than one occurred in a structure of predication. Three hundred twenty of the 589 structures of predication with adverbials appeared with only one clause-level adverbial tagmeme. One hundred sixty-six occurred with two adverbials, seventyfour with three, and twenty-one with four. Only seven structures of predication occurred with five adverbials, and just one with six, which is the most to occur in a single pattern. 4.2.1. Structures of Predication With One Clause-Level Adverbial Tagmeme As was indicated above, the most frequent structure of predication with adverbials was that with one adverbial. The 320 structures of predication with one adverbial consisted of fifty-three coordinated predications, 101 dependent clauses, and 166 independent clauses. As a rule, the adverbials stood either before or after the primary tagmemes but at times stood between them. The introductory formula tagmeme, as would be expected, invariably occurred (excepting those, of course, which appeared in correlative constructions) before the predicator and the other tagmemes, and in only independent clauses. The purpose and agent slots, filled only by prepositional phrases, appeared always after the predicator. The time, location, manner, and relationship tagmemes had greater freedom of movement, but generally were placed after the predicator and the other primary tagmemes as well. Table XXVI shows the positions of the adverbial slots and their fillers with respect to the predicator in patterns with only one adverbial. It may also be noted that only the adverb in the location and manner slots had a relatively high degree of freedom of movement. This apparent freedom of movement was due, for the most part, to the position the adverbs took in dependent clauses, especially in the adjectival clauses, in which they stood between the subject and the predicator, leaving the latter as the final element in the clause. Here is an example: ... £>eper waeron .... '... who were there ....' (262.25) The occurrence of the relationship tagmeme before the predicator, noted earlier, was perhaps due to the scribe's use of it as a link between clauses. Here are two examples: (1) Se an waes gehaten Petrus ... mid him helden 5a of Rome ... Se oöer het Gregorius ...mid him held se kasere of Sexlande ....' One was called Peter ... with him held those of Rome ... The other was called Gregory ... With him held the emperor of Saxony ....' (260.18-24) (2) Des ilces geares com se abbot Heanri ... Alfter him com se abbot of Clunni .... This same year Abbot Henry came ... After him came the abbot of Cluny ....' (261.5-7) Other varied examples of patterns with one adverbial are as follows:

119

SECONDARY CLAUSE-LEVEL TAGMEMES

ο ζ

Ο Η

ο

ζ

ο

ο

ζ

υ •Μ C ο Τ3 β

ο

ζ

£Ό α η

ο

ζ

ο

ζ

υ Τ3 β £ υ 0

ο ζ

ο ζ

Ο Ο Ν h αο ν> ο ο-, νο

οο

I— Ο Λ ο

Ο Ο

Ο

r-

Ο Ο Ο Ο

—< Tt

-ψ Ο — — m Ο (Ν Γ-

OOTf-VO f^lr^^H

Γ-

Vi —i

Q TJ-

Ο Ο Ο Ο

CS ν-ι ® Ο Ο

Ο Ο Ο Ο Ο

VO 00

Ο

Ο

CS

Ο Ο Ο Ο

——Ο Ο Ο

Ο Ο Ο Ο Ο

Μ ——ο

Μ m Ο η >t ef-clause could possibly be considered an objective complement as well as an appositive. In example 3, because of the tendency of nominal tagmemes to move about in linear order, the swa hwam swa jwa-clause could be treated as the direct object with sercebiscop as objective complement. The Wua-clause in example 4 could be analyzed as forming a non-adjacent double object in the prepositional phrase or as a clause in apposition to Aim.5 With examples 5 and 6 there are even more possibilities. Carlton listed a unit like example 5 and called the ^»/-clause a subject complement.® Jespersen considered such a construction as having its clause placed in 'extraposition' at the end of the unit and represented by it at the beginning.7 One may also consider the />ef-clause here as filler of the subject slot and the hit as an expletive. It seems that example 1 can only be considered an appositive, yet it is not in apposition to a single word, but to the whole idea expressed in the independent clause of the T-unit. 6.2.2. The Adjectival Clause Another type of dependent clause is the adjectival, which occurred seventy-eight times and had a modifying relationship to a nominal tagmeme. As a modifier, it modified nouns, pronouns, and pronominal adjectives which manifested nominal functional slots in clauses. The adjectival clause was different from other dependent clauses in this corpus in that its subordinator typically filled subject, object, or indirect object slots in the clause it introduced. It was the first item in the clause it introduced and, as a rule, stood next to its head or antecedent. There were eight adjectival clauses, however, which were not adjacent to their heads. Excluding the fifteen subjectless (set clauses) clauses, there was only one which did not have a subordinator; this one occurred in an interdependent clause (see page 193). The adjectival clause, the subordinators of which are called relative pronouns, was introduced by se twice, pe twenty times, pa eighteen times, d& p& once, and pet {pat) twenty-one times. The fifteen subjectless clauses had no relatives and always appeared with a subject complement. In one example the relative was omitted, but the clauses were held together by interdependence. Se, in the two examples, took singular verbs and showed agreement with its antecedent. The unusual form da pa occurred with a plural verb and also showed 6

See Robert A. Palmatier, A Descriptive Syntax of the "Ormulum" (The Hague, Mouton, 1970). • Charles R. Carlton, Descriptive Syntax of the Old English Charters (The Hague, Mouton, 1970), pp. 13-22. 7 Otto Jespersen, Essentials of English Grammar (New York, Henry Holt, 1933), reprinted, University of Alabama Press, 1964, p. 349. See also Randolph Quirk and C. L. Wrenn, loc. cit.

THE RELATIONSHIP AND JOINING OF CLAUSES

191

agreement with its antecedent. The other forms, pe, pa, and pet occurred with both singular and plural verbs and antecedents. The verb in the clause it introduced, however, generally agreed with the number of the antecedent. All the subjectless clauses were formulaic and always occurred with the singular form of the verb; the subject complement always showed agreement with the antecedent. Although there were some exceptions, pe usually was used with animate antecedents and pet with inanimate.8 Following is a list of the syntactic patterns of the seventy-eight adjectival clauses along with the number of times each occurred: SV, 22; SVO, 10; SOV, 6; VSO, 1; OSV, 12; SVC, 7; SCV, 16; ISVO, 1; and OSVI, 3. 6.2.2.1. Non-contiguous clauses. - It was mentioned earlier that the adjectival clause usually was adjacent to the word it modified and also that there were eight examples for which this was not true. The five examples below illustrate these non-contiguous clauses: (1) Ac pet ofer com Rome pet of er cumeö eall weoruld t>et is gold y seolure. 'But that which overcomes all the world overcame Rome, that is, gold and silver.' (252. 29-30) (2) y t>a muneces herdon da horn blawen pet hi blewen on nihtes. 'And the monks heard the horns blowing that they blew at night.' (258.23-24) (3) for xuricman sone raude ojser pe mihte. 'For everyone who could soon robbed the other.' (263.6-7) (4) y te king makede Teodbald aercebiscop pe was abbot in the Bee. 'And the king made Theobald archbishop, who was abbot of Lebec.' (266.14-15) (5) sume ieden on aelmes pe waren sum wile rice men. 'Some who had been at one time powerful men went on alms.' (264.33-34) 6.2.2.2. Subordinators as fillers of primary functional slots. - The words se, pe, pa, da pa, and pet, may manifest the subject, direct object, or indirect object slots within the boundaries of the clauses they introduce. These connectives filled the subject slot in forty-six instances, the direct object slot in fifteen instances, and the indirect object slot in one instance. Sixteen clauses occurred without subordinators. Examples of subordinators as subjects: (1) Des ilee gaeres he gaef J)one abbotrice of Burch an of Peitowe. se haefde his abbotrice Sancte Johns same year he gave the abbacy of Peterborough to Henry of Poitou who had possessed his abbacy of 13-15) 8

abbot Heanri waes gehaten of Angeli on hande. 'This an abbot (who) was called St. Jean of d'Angely.' (257.

With the exception of "{se landes pe lien to £>e circewican" (265.21), this rule was invariable in the Final Continuation. For a discussion of this matter in early Middle English as a whole, see Angus Mcintosh, "The Relative Pronoun pe and pat in Early Middle English", English and Germanic Studies, I (1947-8),73-90.

192

THE RELATIONSHIP AND JOINING OF CLAUSES

(2) for agenes him risen sone {3a ricemen pe waeron swikes. 'For against him rose the powerful men who were traitors.' (263.17-18) (3) y se man pa heafde twa hundred oöJ>e öre hundred swin ne beleaf him noht an. 'And the man who had two or three hundred pigs, for him not one was left.' (261.22-23) (4) J>is bebaed se sercebiscop Willelm of Cantwarabyrig y ealle J>a leod biscopes dapa waeron on Englalande. 'This ordered Archbishop William of Canterbury and all the diocesan bishops who were in England.' (260.5-7) (5) Eall J>is geare weas se kyng Heanri on Normandi for J)one unfriö pet wss betwenen him y his nefe öone eorl of Flanders. 'All this year King Henry was at Normandy on account of the strife that was between him and his nephew, the count of Flanders.' (258.29-31) Examples of subordinators as objects: (1) t>et wes for se miccle unfriö pet he heafde wiö se king Loöewis of France, y wiö se eorl of Angeow. y wiö his agene men aire mest. 'That was because of the great strife that he had with King Louis of France and with the count of Anjou, and with his own men most of all.' (253.21-23) (2) Da siööon ferde se king y wan ealle ties eorles castelas Waleram £>a waeron on Normandi. y ealle fra oöre pa his wiörewines healden him togeanes. 'Then, later, the king went and won all Count Waleran's castles that were in Normandy and all the others that his enemies held against him.' (254.3-5) (3) mid him com se cwen y his dohter pet he ieror hafde giuen t»one kasere Henri of Loherenge to wife. 'With him came the queen and his daughter whom he had given as wife to the emperor, Henry of Lorraine.' (256.9-10) Example of subordinator as indirect object: him com togaenes Willelm eorl of Albamar pe Jse king adde beteht Euorwic y to other aeuez men mid faeu men. 'And Count William of Aumale, to whom the king had entrusted York, and the other trustworthy men with a few men came against him.' (266.4-6) 6.2.2.3. The head of adjectival clauses. - As modifiers of fillers of the primary functional slots, adjectival clauses operated basically on a phrase level and were tied to the head of the phrase which they helped to expand. The head of adjectival clauses in this corpus may be a noun, pronoun, or pronominal adjective. Of the seventy-eight adjectival clauses, fifty-eight appeared with nouns as heads, seven with pronouns, and thirteen with pronominal adjectives. Some typical examples of each are as follows: Noun as head: (1) of Willelm Malduit pe heold Rogingham fix castel he wan Cotingham y Estun. 'From William Maudit who held Rockingham Castle, he won Cottingham and Easton.' (265.24-25)

THE RELATIONSHIP AND JOINING OF CLAUSES

193

(2) for agenes him risen sona {3a riceman pe wzeron swikes. 'For against him soon rose the powerful men who were traitors.' (263.17-18) (3) for ouer sithon ne forbaren hi nouther circe ne cyrceiaerd oc namen al J>e god pet par inne was. y brenden sythen fce cyrce y al te giedere. 'For against custom, they respected neither the church nor church yard but took all the property that was inside and later burned the church and everything together.' (264.36-37) Pronoun as head: (1) Ac pet ofer com Rome pet ofer cumed eall weoruld }set is gold y seolure. 'But that which overcomes all the world overcame Rome, that is, gold and silver.' (252.29-30) (2) Ne hi ne forbaren biscopes land ne abbots nor preostes. ac raeueden munekes y clerekes. y xuric matt other pe ouer mythe. 'Neither did they respect the bishop's land nor the abbot's nor the priest's, but robbed the monks and clerics, and everyone who had the greater power, the other.' (265.1-3) (3) fet wss for se man de hafde an pund he ne mihte cysten aenne peni at anne market. 'That was because the one who had a pound was not able to get the value of a penny at the market.' (255.9-10)9 Pronominal adjective as head: (1) Eall pet pa beon dragen toward swa frett t>a drane y dragaö fraward. 'Everything that the bees dragged in, thus the drones ate and carried out.' (258.12-13) (2) sume ieden on aelmes pe waren sum wile rice men. 'Some who were at one time powerful men went on alms.' (264.33-34) 6.2.2.4. Adjectival clauses without subordinators. - There were two kinds of constructions which did not have overt subordinators signaling inclusion. One, of which there was only one example, was what may be considered an interdependent clause. The clause, which follows, occurred in an adverbial clause: ... oc aefre t>e mare he iaf heom J>e waerse hi waeron him. '... but always the more (that) he gave them, the worse they were to him.' (266.18-19) The other construction was a formulaic subjectless clause usually occurring with wies gehaten, with a proper noun before it as subject complement.10 A typical example is as follows: y hine bebyrigde se biscop of Ceastre Rotbert Pecced wxs gehaten. 'And the bishop of Chester, (who) was Called Robert Pecceth, buried him.' (251.12-13) 9 Since this example is used in an indefinite sense, there is some question about man of the unit, se man de, that is, whether it means 'man' or 'one'. It could be a variation of se pe and is considered as such here. 10 This clause will be discussed at greater length in the section labeled "Parataxis".

194

THE RELATIONSHIP AND JOINING OF CLAUSES

A variation of this occurred with a proper name before the predicator (wxs) and the subject complement after it. Here is the example: Da sone in J>e lenten ferde se aercebiscop to Rome aefter his pallium ... y Gifard wazs pes kinges hird clerc. 'Then early in the spring the archbishop went to Rome after his pallium ... and (also) Giffard (who) was the king's court clerk.' (252.17-21) 6.2.2.5. Adverbial clauses with adjectival function. - There were two adverbial clauses connected by par to a noun in a nominal slot, performing therefore an adjectival function. Here are the examples: (1) Des ilees gaeres on Jjone lenten tide w»s se eorl Karle of Flanders ofslagen on ane circe pzr he Ixi y bsed hine to Gode tofor pone weofede among pane messe fram his agene manne. 'This same year, in the spring, Count Karl of Flanders was killed by his own men in a church where he lay and praying to God before the altar during mass.' (257.3-5) (2) Hi dyden heom in quarterne par nadres y snakes y pades wxron inne y drapen heom swa. 'They put them in prisons where there were adders and snakes and toads and killed them this way.' (264.13-15) 6.2.3. The Adverbial Clause The third type of dependent clause is the adverbial, which occurred 125 times and operated at three levels. It might have a modifying relationship to certain functional slots of the independent clause of the T-unit; it might fill such clause-level slots as time, location, and manner; and it might fill the dependent clause slot of the unit which has a modifying relationship to the independent clause as a whole. The adverbial clause was introduced by subordinators which normally did not have primary functions within the clauses they introduced. The adverbial clause occurred in various syntactic patterns and stood before or after the independent clause of the T-unit. In adding information to the T-unit, the adverbial clause might express such notional relationships as time, location, comparison, cause, result, condition, and concession. 6.2.3.1. Subordinators. - In this corpus, adverbial clauses were brought into structural relationship most frequently by combinations of swa with swa, swa with pet, forpi pet and pa with pa. Here is a complete list of the subordinators along with the number of times each occurred; swa + adj+swa, 5; swa -f- noun + swa, 3; swa swa, 4; swa, 1; alswa + noun + swa, 1; suile als, 1; alsuic alse, 1; alse, 2; pan (ne), 5; swa pet, 8; swa + adj + pet, 1; swa + noun + pet, 1; pet, 9; to pzt, 1; swile, 2; gif, 5; pxh, 3; pop, 1; pxr, 4; swa + adv + swa, 4; war, 2; forpi(an) pet, 13; for, 9; purh pxt, 2; hwilepe ... pa, 3; pa ... pa, 17; xr, 6; pa hwilepet, 1; sone swa ... swa, 1; fram pet, 1; siööan pet, 1; pa, 2; til, 2; and wile, 2. One clause did not have a subordinator, but expressed the concessive relation by appearing with its verb, wxr, initially in the clause.

THE RELATIONSHIP AND JOINING OF CLAUSES

195

As a rule, the adverbial subordinates merely served as a connector, but there were two clauses which, it seemed, had the subordinators filling the subject slot. They are presented below: (1) y fca for mid h i m y after him swa micel folc swa nsefre ser ne dide .... 'And there then went with him and after him so large a number of people as never had done....' (259.18-19) (2) y hit ward sone suythe god pais, sua Jset neure was here. 'And it was soon a very good peace such that never had been here.' (268.17-18) 6.2.3.2. Order of clauses. - Of the 125 adverbial clauses, thirty-three occurred before the independent clause of the T-unit to which it was related, and ninety-two occurred after it. The most frequently occurring adverbial clause in the dependent-independent clause order was the time clause, which appeared in twenty-four of the thirty-five examples. Significant in this count is the fact that of the twenty-four time clauses, twenty-one supported correlation, a relationship which always had the dependentindependent clause order in this corpus. Of the seven location clauses, two occurred in this order. Only two of the twenty-five comparison clauses had the dependentindependent clause order; these, like most of the time clauses, expressed correlation. Of the remaining seven examples in the dependent-independent order, the dependent clause was causal once and conditional four times. The two types of adverbial clauses not represented in the dependent-independent order were the concessive and result clauses. In the independent-dependent order, the time clause occurred 16 times, the location five times, the comparison twenty-three times, the causal twenty-four times, the conditional three times, the concessive five times, and the result eighteen times. Wherever possible, an example of each type of adverbial clause before and after the independent clause is presented below: Time (before): Da hwile pet se xrcebiscop wses ut of lande, geaf se kyng öone biscoprice of Baöe Jjes ewenes canceler Godefreiö waes gehaten. 'While the archbishop was out of the country, the king gave the bishopric of Bath to the queen's chancellor who was called Godfrey.' (252.34-35) Time (after): God hit bete pa his wille bed. 'May God amend it when it is his will.' (261.25) Time (correlative): pa hi ne leng ne muhten polen J)a stali hi ut y flugen. 'When they could endure (it) no longer, they stole away and fled.' (267.1-2) Location (before): War sse me tilede. J>e erthe ne bar nan corn, for }>e land was al fordon. mid suilce

196

THE RELATIONSHIP AND JOINING OF CLAUSES

dsedes. 'Wherever one tilled the soil, the earth bore no corn because the land was all destroyed with such deeds.' (265.8-9) Location (after): Hi dyden heom in quarterne par nadres y snakes y pades wseron inne. y drapen heom swa. 'They put them in quarters where there were adders, snakes and toads inside and killed them this way.' (264.13-14) Comparison (before): Suilc y mare panne we cunnen ssein. we holenden *ixx· wintre for ure sinnes. 'This and more than we can tell, we endured for nineteen years on account of our sins.' (265.10-11) Comparison (after): ealle heoldon here wifes be J)es kynges leue swa swa hi ear didon. 'All of them kept their wives by the king's permission as they had done before.' (260.9-10) Causal (before): {Dar aefter purh pet he wxs des kynges msei of Engleland y pes eorles of Peitowe t>a geaf se eorl him Jsone abbotrice of Sancte Iohns minstre of Angeli. 'After that, because he was a kinsman of the king and the count of Poitou, the count gave him the abbacy of St. Jean d'Angely.' (257.25-27) Causal (after): ac he ne myhte for he wart it war. 'But he could not because he was aware of it.' (266.9-10) Conditional (before): Gif twa men oper -iii· coman ridend to an tun. al £>e tunscipe flugaen for heom. 'If two or three men came riding to a village, all the villagers fled from them.' (265.3-4) Conditional (after): nu him behofed J>et he crape in his mycele codde in aelc hyrne gif pxr wxre hure an unwreste wrenc J>et he mihte get beswicen anes Crist y eall Cristene folc. 'Now he was forced to creep into each corner of his great bag (to see) if there were yet one poor trick with which he might yet betray Christ and all Christian people.' (262.5-7) Concessive (only after): y hi togaedere comen y wuröe sshte. pop it litel forstode. 'They came together and reached agreement although it amounted to little.' (263.22-23) Result (only after): y J>aer man him held pet he ne mihte na east na west. 'And there one held him so that he could not go east or west.' (262.8-9)

THE RELATIONSHIP AND JOINING OF CLAUSES

197

6.2.3.3. Syntactic groups and patterns. - Eighty-one of the 125 clauses belonged to Group 1 (subject and predicator) with the SV order occurring seventy-seven times and the VS four times. Group 2 (subject, predicator, object) was represented by thirty-two examples with the following distribution: SVO, 24; SOV, 5; VSO, 2; and OSV, 1. Group 3 (subject, predicator, subject complement) had nine examples with eight occurring in the SVC order and one in the SCV. Group 4 (subject, predicator, indirect object, direct object) had only three examples with one each in the SYIO, SIVO, and OSVI orders. 6.2.3.4. Expression of relationships. - The adverbial clauses, it was learned, were joined by a variety of subordinating constructions and expressed various relationships. According to the meaning they expressed, they were roughly grouped into adverbial clauses of time, location, comparison, cause, condition, concession, and result. Often the meanings were indicated by the subordinators, but since some of the subordinators expressed various relationships, the meanings usually had to be gathered from the T-units as a whole. Here the purpose will be to illustrate how these clauses were structured to show these relationships. Since there were nuances within the groups themselves, this section will be presented in such a way as to bring out the differences within the groups (wherever examples are available) as well as between the groups. Furthermore, the examples will be listed according to the type of subordinator. The TIME clause is the first adverbial clause that will be considered here. The time clause occurred either before or after the independent clause of the T-unit. Of the forty time clauses, twenty-four occurred before the independent clause and sixteen after it. Twenty-one of those occurring before the independent clause were members of a correlative construction which, as a rule, had the dependent-independent clause order. In this corpus, time clauses were joined by pa hwile pe ...pa (3),11 swa lange swa (1), swa lange pet (1), swa radlice swa (1), pa ...pa (17), xr (6), pa hwile pet (1), sone swa ... swa (\),fram pet (1), siddan pet (1), pa (2), to pet (1), til (2), and wile (2). The time clauses indicated the following six temporal relations: time when, immediate sequence, duration, time in reference to a preceding action, and time in reference to a subsequent action.12 The first type of time clause to be considered here is the clause that expressed TIME WHEN. There were nineteen clauses which expressed this temporal relation, seventeen of which were joined by pa ...pa and two by pa. An example of each follows: (1) Pa hi dider coman da nam man an y an y benam aelc öone riht hand y t>a stanes beneöan. 'When they came to that place, then they took them one by one and took the right hand of each and their testicles below.' (255.13-14) 11

The number in parentheses indicates the number of times each occurred. This classification owes much to Arthur Adams, The Syntax of the Temporal Clause in Old English Prose (New York, Henry Holt and Company, 1907), p. 9. 14

198

THE RELATIONSHIP AND JOINING OF CLAUSES

(2) Jjer efter in £>e lengten Festrede J^e sunne y te daei. abuton non tid daeies. pa men eten. öet me lihtede candles to aeten bi. 'After that in the spring the sun became dark and about noon time when men were eating, one had to light candles to eat by.' (266.11-12) A different type consisted of a noun phrase modified by an adjectival clause: pat ilce dsei pat Martin Abbot of Burch sculde pider far en. pa saeclede he y ward ded iiii N. Ian. 'The same day that Abbot Martin of Peterborough was to go there, he then became sick and died on 2 January.' (268.29-30) The second type of time clause is that which expressed IMMEDIATE SEQUENCE, There were only two examples in the corpus. One was introduced by sone swa ... swa and the other, by swa radlice, both of which are below: (1) £>a sone swa he hxfde wind swa ferde he ofer in Normandie .... 'Then as soon as he had wind, he went into Normandy ....' (253.6-7) (2) ... for hit wass ful cuö ofer eall land t>et swa radlice swa hepxr com ... }ja son J)2er asfter ]m sasgon y herdon fela men feole huntes hunten. '... because it was well known everywhere that as soon as he came there many men soon after that saw and heard many hunters hunting.' (258.15-19) The third type of time clause is the clause that denoted DURATION, of which there were eight examples. Two each were joined by pa hwile pe ...pa and wile, and one each by pa hwile pet ... pa, pa hwile pet, swa lange swa, and swa lange pet. An example of each follows: (1) pa hwile pe pa munecas sungen pxre messe ...pa com se fir on ufen weard {Done stepel. 'While the monks were singing the mass ... the fire came to the upper part of the tower.' (250.4-6) (2) y {>et lastede Jje -xix- wintre wile Stephne was king. 'And that lasted the nineteen winters while Stephen was king.' (264.25-26) (3) pa hwile pet he pxr wxs pa geaf he Jione biscoprice of Lincolne an clerc Alexander waes gehaten. 'While he was there, he then gave the bishopric of Lincoln to a clerk who was called Alexander.' (253.2-3) (4) £>a hwile pet se sercebiscop wxs ut of lande, geaf se kyng öone biscoprice of Baöe {>es canceler Godefreiö waes gehaten. 'While the archbishop was out of the country, the king gave the bishopric of Bath to the queen's chancellor who was called Godfrey.' (252.34-35) (5) oc he wolde hauen bade on hand, y swa hafde swa lange swa Godes wille waes. 'But he wanted to have both in hand and so had as long as it was God's will.' (257.21-22) (6) and laeste swa lange pet hit wxs liht ofer eal. 'And it lasted so long that it was light everywhere.' (250.23) The fourth type of time clause is the clause which expressed

TIME IN REFERENCE TO

THE RELATIONSHIP AND JOINING OF CLAUSES

199

The two examples in the corpus (both of which are given below) were joined by fram pet and siddan pet. A PRECEDING ACTION.

(1) Isis waes saegon y herd fram pet he pider com eall pet lented tid on an to Eastren. 'This was seen and heard from the time he came there all of Lent up to Easter.' (258.26-27) (2) y t>a for mid him y after him swa micel folc swa naefre aer ne dide siddon pet se firste fare was on Vrbanes dxipape f>eah hit litel behelde. 'And then went with him and after him so large a number of people as never had done since the first journey in the days of Pope Urban, although it amounted to little.' (259.18-20) The fifth type of time clause is the clause which expressed TIME IN REFERENCE TO A SUBSEQUENT ACTION. There were six examples, and all six were introduced by ser. A typical example is given below: Da com se aercebiscop of Cantwarabyrig y waes öaere fülle seoueniht ser hi mihte cumen to pes papes sprzce. "Then the archbishop of Canterbury came and was there a full week before he could speak to the pope.' (252.25-26) The last type of time clause is the clause which indicated the TIME OF THE TERMINATION OF THE ACTION OF THE INDEPENDENT CLAUSE. There were only three examples, two of which were introduced by til and one by to pet.13 An example of each is given below: (1) for he besäet heom til hi aiauen up here castles. 'For he besieged them until they gave up their castles.' (267.27) (2) Me dide cnotted strenges abuton here haeued. y uurythen to pet it gxde to pe hxrnes. 'One put knotted ropes around their heads and twisted them until (to the point that) they penetrated to the brains.' (264.12-13) The next type of adverbial clause is the LOCATION clause, which expressed location or place. There were seven examples in the corpus, two of which occurred before the independent clause and five after it. Three of these clauses were introduced by pxr. War occurred in two examples and swa hwar swa, in one. An example of each is given below: (1) On t>is gar waerd J>e king Stephne ded y bebyried per his wif y his sune war on bebyried set Fauresfeld. 'In this year King Stephen died and was buried there where his wife and son were buried at Faversham.' (268.21-22) (2) for ware se he com he dide mare yuel Joanne god. 'For wherever he came, he did more evil than good.' (267.32) (3) ^Efter him com se abbot of Clunni Petrus gehaten to Englelande bi J)es kynges leue y waes underfangen ouer eall swa hwar swa he com mid mycel wuröscipe. 18

Ν. R. Ker reads this as it pet. "Some Notes on the Peterborough Chronicle", Medium ALvum III (1934), 138. If this is accurate, then the clause would be result instead of time. Plunimer's reading of the text (the manuscript) was a more accurate reading considering the arrangement of the letters in the manuscript. In his glossary, he translates to pet as 'to the point that'. Perhaps Ker was led to discount Plummer's reading because if to pet is the actual reading, then it would be the only use recorded in the manuscripts which have survived the Middle Ages. See Adams, op. cit., p. 141.

200

THE RELATIONSHIP AND JOINING OF CLAUSES

'After him came the abbot of Cluny named Peter to England by permission of the king and was received with great honor wherever he came.' (261.7-9) Another type of adverbial clause is the COMPARISON clause, which occurred twenty-five times. Included in this group were clauses which appeared to express manner and degree simply because comparison seemed to be expressed as well. Only two of the twenty-five comparison clauses occurred in the dependent-independent order. Following is a list of the eleven comparison subordinators along with the number of times each occurred: swa + adj. + swa, 5; swa + noun-f- swa, 3; swa swa, 3; swa, 1; alswa + noun + swa, 1; suilc als, 1; suilc, 2; alsuic alse, 1; swa ... alse, 2; pan (ne), 5; and sua pet, 1. An example of each follows: (1) ... swa t»et ealle öe hit saegon waeron swa offered swa hi nxfre xr ne wxron. '... so that all who saw it were so afraid as they had never been before.' (261.17-18) (2) y ahengen t>aer swa fela pefas swa nxfre xr ne wxron. 'And (they) hanged there so many more thieves than ever had been before.' (254.31-32) (3) diden him ealle hersumnesse swa swa hi scolden don here abbot. '(All) did him homage as they should have done their abbot.' (262.1) (4) y t>asr he wunede eall riht swa drone dod on hiue. 'And there he lived just as drones do in a hive.' (258.11) (5) y wuröon pa alswa gode freond swa hi wxron xror feond. 'And they became just as good friends as they had been enemies.' (259.29-30) (6) y uuard J)e sunne suilc als it uuare thre niht aid mone, an sterres abuten him at mid daei. 'And the sun became as if it were a three-night old moon with stars about at noon.' (263.1-2) (7) y se wolcne undide on fower healfe and faht f>aer togeanes. swilc hit scolde acwencen. 'And the sky opened on four sides and fought against it as if it were to quench it.' (250.20-21) (8) ... forJ)i }>et hi uuenden t>et he sculde ben alsuic alse the eom wes .... '... because they expected that he would be just as his uncle was ....' (263.25-26) (9) for ne uuaeren naeure nan martyrs swa pined alse hi wxron. 'For never were martyrs so tortured as they were.' (264.8-9) (10) ne naeure hethen men werse ne diden pan hi diden. 'No heathen men did worse than they did.' (264.35-36) (11) hit ward sone suythe god pais, sua pet neure was here. 'It was a very good peace such that never was here.' (268.17-18) A fourth type of adverbial clause in this corpus is the CAUSAL. Only one of the twentyfive examples occurred in the dependent-independent order. The most frequent subordinator used was forpi pet, which appeared in twelve examples. Variations of this connective were forpan pet with one example and for with nine examples. Other connectives expressing the causal relationship were swa swa with one example and purh pxt with two examples. Examples are as follows:

THE RELATIONSHIP AND JOINING OF CLAUSES

201

(1) J)a forlaes he Jset mid rihte forpi pet he hit hxfde seror beieten mid unrihte. 'Then he lost it justly because he got it before unjustly.' (257.29-30) (2) J>et waes forpan pet hit waes don done pape to understanden {>et he hasfde underfangen öone aercebiscoprice togeanes J)a muneces of t>e mynstre y togeanes rihte. 'That was because it was given the pope to understand that he had received the archbishopric against the monks of the monastery and uncanonically.' (252.26-29) (3) al folc him luuede for he dide god iustise y makede pais. 'All the people loved him because he had good justice and made peace.' (268.19-20) (4) Oc se ilce Heanri dide Jjone king to understandene {set he haefde laeten his abbotrice for J>et miccele unsibble ... y purh pset he was legat of done Rome scott. 'But this same Henry had the king to understand that he had left his abbacy on account of the great unrest... and because he was legate of the Romescot.' (257. 17-21) The fifth type of adverbial clause is the CONDITIONAL, which appeared in only five examples. Four of the five clauses occurred in the dependent-independent order and one in the independent-dependent order. This clause, which was introduced by gif, generally had a modifying relationship with the independent clause as a whole instead of with one of its parts. A typical example follows: i>a bejiohte he him |>et gif he mihte ben rotfest on Engleland J)et he mihte habben eal his wille. 'Then he thought to himself that if he could be firmly rooted in England, he might be able to have all his desires.' (257.33-34) The next type of adverbial clause is the CONCESSIVE clause, of which there were five examples. This clause always occurred after the independent clause and had a modifying relationship with the independent clause just as the conditional clause did. It was introduced by pseh three times and pop once. The other clause showed the concessive relationship by the use of the subjunctive form of wxran in the initial position in the clause. An example of each is given below: (1) y siööon he heold his concilie on Lundene fülle £>reo dagas ... y bead Jsaer pa ilce lagas J>a Anselm aercebiscop haefde sror beboden y feala ma peah hit litel forstode. 'And later he held his council in London for three full days ... and ordered there the same laws which Archbishop Anselm had ordered before (and many more) although it amounted to little.' (255.26-30) (2) y hi togaedere comen y wuröe saehte. pop it litelforstode. 'And they came together and reached agreement although it amounted to little.' (263.22-23) (3) he hit dide foröi J)et he wolde frurh his micele wiles öaer beon wxr it tweolf mond odde mare. 1 siööon ongeon cumen. 'He did it because he wanted through his great trickery to be there, even though it was twelve months or more, and then to come back.' (259.7-9)

202

THE RELATIONSHIP AND JOINING OF CLAUSES

The last type of adverbial clause is the RESULT clause. There were eighteen examples, and all eighteen occurred after the independent clause. The result clause may have a modifying relationship with a part of the independent clause or with the independent clause as a whole. It was usually introduced by ίννα pet or pet, of which there were seven and nine examples respectively. One clause was joined by the swa + adj + pet combination and another by the swa + noun + pet combination. Examples of each follow: (1) Des ilces geares waeron fasla untime on Englelande. on corne y on ealle westme. swa pet betweonen Cristes messe y Candel messe man sxlde pet acer ssed hwxte ...to six scillingas .... 'This same year there was so much bad weather in England on corn and all crops that between Christmas and Candlemas one sold seed wheat ... for six shillings an acre ....' (254.12-15) (2) y frasr behet se abbot Heanri him fret he scolde beieton him frone mynstre of Burch pet hit scolde beon underded into Clunni. 'And there Abbot Henry ordered him that he should get for him the monastery of Peterborough so that it could be subjected to Cluny.' (261.10-11) (3) I>et was forfri fret corn waes litel. and se penig waes swa ifel pet se man pa hxfde at an market an pund he ne mihte cysten pser of for nan ping twelfe penegas. 'That was because the corn was scarce and the penny was so bad that the man who had a pound at the market could not by any means get the value of twelve pennies for it.' (254.16-19) (4) On öes ilces geares wearö swa micel fold on Sancte Laurenties messe dxig pet feola tunes y men weordan adrencte .... 'In this same year there was so great a flood on St. Lawrence's Day that many villages were flooded and people drowned ....' (255.37)

6.3. THE PARENTHETICAL CLAUSE

A parenthetical clause is an injected statement that interrupts another clause with which it is only slightly related. In this corpus, it did not have a distinct modifying relationship to the clause it interrupted; it merely added specific qualifying or restrictive details. Generally the interruption was very noticeable because it was abruptly signaled by pet is or pet wxs. There were two examples, however, which were not easily recognized as parenthetical clauses because they were unobtrusive and introduced by markers which could lead one to consider them otherwise. The one introduced by pe could be considered an adjectival clause structurally and the one introduced by pes pe could be treated perhaps as an adverbial clause of restriction. Be that as it may, they were treated here as parenthetical clauses. These plus the six with pet is and the four with pet waes as markers amounted to a total of twelve parenthetical clauses in the corpus.

THE RELATIONSHIP AND JOINING OF CLAUSES

203

The pet is and pet wxs parenthetical clauses were followed by words or phrases which more or less functioned as appositives to words or phrases in the clause which they interrupted. Three of these clauses marked by pet is or pet wxs were followed by prepositional phrases which made more specific the prepositional phrases functioning as adverbiale of time or place in the preceding clause. Here are two examples: (1) I>a tidde hit on an Wodnesdei. pet wies on -iiii· idus Ianrii J>et se king rad in his derfald .... 'It happened on Wednesday, that was on the tenth of January, that the king rode in his deer park ....' (251.4-5) (2) Sume hi diden in crucethus pet is in an cxstepat was scort y narrow y undep .... 'Some they put in a torture chamber, that is, in a cast that was short, narrow, and shallow ....' (264.15-16) Four clauses supplied specific detail to direct objects. A typical example is as follows: On fris gaer sende se king Henri toforen Cristes messe of Normandi to Engla lande y bebead fret man scolde beniman ealle fra minitere fre waeron on Engla lande heora liman pet wxs here elces riht hand y heora stanen benedan. 'In this year King Henry sent from Normandy before Christmas to England and ordered that one should take from all the minters who were in England their limbs, that was, from each their right hand and their testicles below.' (255.6-9) One clause made specific a relative pronoun that filled the subject slot of a preceding clause: Ac fret ofer com Rome fret ofer cumeö eall weoruld pet is gold y seolure. 'But that which overcomes all the world overcame Rome, that is, gold and silver.' (252.29-30) Another pet wxs clause made specific an idea contained in an adverbial clause of time: ... for hit waes ful cuö ofer eall land fret swa radlice swa he fraer com pet wxs pes Sunendxiespet man singaö Exurge Quare 0. D '... for it was well known everywhere that as soon as he came there, that was, the Sunday that one sings the Exurge Quare O. D ' (258.16-1) Still another pet is clause added information that was in apposition to the past participle form of the predicator: frat was sua maced pet is fxstned to an beom. 'That was thus made, that is, fastened to a beam" (264.19-20) Unlike the preceding examples and not as distinctively parenthetical were those marked pe and pes pe, both of which are presented below: (1) saeidon pes pe heom Puhte fret fraer mihte wel ben abuton twenti oöer frritti horn blaweres. '(They) said, as it seemed to them, that there might well be about twenty or thirty hornblowers.' (258.24-26)

204

THE RELATIONSHIP AND JOINING OF CLAUSES

(2) E>a namen hi J>a men fie hi wenden J>et ani god hefden .... 'Then they took the men, whom they thought, that had any property ....' (264.5-6)

6.4. COORDINATION OF T-UNITS

In most English grammars it has been customary to speak of simple, compound, and complex sentences, all of which are defined on the basis of the type, number, and combination of the clauses they contain. Although the T-unit provides the investigator with a provisional unit by which he can detect and describe all the most significant syntactic relations he needs to consider, a descriptive syntax seems almost incomplete if it does not consider the possibility of structures even larger than those already considered in this chapter. Whenever one first approaches some early English texts, he is usually struck by what appears to be an awkward structure of a string of sentences. Whether or not there is a designed structure consisting of two or more syntactically coordinated or equivalent predications in these texts is difficult to tell. In this corpus, to say with absolute certainty that the coordinator y (and) was used to form larger structures to indicate coordinated and equivalent patterns was almost impossible. The coordinator could simply be a means of accumulating and stringing structures together in sequential order without regard to any other relationships and meaning. Since the corpus, however, showed that the combination of independent and dependent clauses were relatively close to that of Modern English, although not always expressed as precisely, the investigator considered the possibility of coordination as a structural device. He, however, did not find evidence to say with certainty that coordination14 (of T-units) was an actual structural device for showing notional relationships.15 Even if the evidence was not conclusive enough to say that coordination was a structural device in expressing notional relationships, strings of T-units should be considered, if only briefly. T-units were most frequently connected by y and and, but were also linked by ac, for, and ne.16 The first, y, and (and), is usually called a COPULATIVE COORDINATOR, which connects two or more units and their meaning; the second, ac (but), an ADVERSATIVE COORDINATOR, which connects the units but contrasts their meaning; the third, for (for), a CAUSAL COORDINATOR, which adds a unit to explain the preceding unit; and the last, ne (neither, nor), a DISJUNCTIVE COORDINATOR, which connects the units but disconnects their meanings. Of the 484 14

Coordination, as used here, would mean that the T-unit following the conjunction is operating at the same level as the preceding structure forming an overall closely related structure. 15 Let it be known, however, that perhaps there is just as much evidence here as there is in a selected corpus of Modern English if we consider the failure of linguists to find a suitable definition of the sentence and if H. A. Gleason, Linguistics and English Grammar (New York, Holt, Rinehart and Winston, Inc., 1965), pp. 333-336, is close to being correct when he states that absolute coordination implies interchangeability and that it is hard to distinguish subordination and coordination. See also John P. Hughes, The Science of Language (New York, Random House, 1962), p. 185. " These, of course, generally expressed rather definite relationships.

THE RELATIONSHIP AND JOINING OF CLAUSES

205

T-units, 191 began with y or and, forming strings which varied from two to ten Tunits. Thirty-nine examples occurred with a string of two, 1 7 fifteen a string of three, sixteen a string of four, five a string of five, one a string of six, two a string of seven, and one a string of ten. Examples of compound units with strings of two, three, and ten are given below: (1) aerost man hem beraefoö her eahte and siööon man hem ofslaeö. 'First one deprived them of their property, and afterwards one killed them.' (255.2-3) (2) On J)is geare waes se king Heanri on Cristes maessan on Norht wie. y on Pasches he weas on Norththamtune. y on })one lenten tyde Jsaer toforen forbearn se burch on Gleaweceastre. 'In this year King Henry was at Norwich at Christmas, and at Easter he was at Northampton, and the spring before that the village in Gloucester burned.' (250.1-4) (3) beteahte hine siööon J>one aercebiscop Willeime of Cantwarabyrig. y he hine ledde to Cantwarabyrig. y he waes £>aer under fangen mid micel wuröseipe y mid micel processionem. y he sang öone heh messe on Eastren daei aet Cristes wefod. y siööon he ferde ofer eall Engle lande to ealle t>a biscoprices y abbot rices {5a waeron on J>is lande, y ofer eall he waes underfangen mid wuröseipe y ealle hine iaefen micele gife y maere. y siööon he heold his concilie on Lundene fülle ]?reo dagas on natiuitas Sancte Marie on Septemb mid aercebiscopes y mid leod biscop y abbotes y laered y lawed. y bead fciaer \>a ilee lagas ]m Anselm aercebiscop haefde aeror beboden y feala ma J>eah hit litel for stode. y t>eonon he for ofer sse sone aefter Sancte Michaeies messe y swa to Rome, y se aercebiscop Willelm of Cantwarabyrig. y se aercebiscop Turstein of Eferwic. y se biscop Alexander of Lincolne. y se biscop of LoJ>ene Iohan. y se abbot of Sancte Alban Gosfreiö y waeran J)aer underfangen of bone pape Honorius mid micel wuröseipe. y waeron jisre eall t>one wintre. '(He) entrusted him later to Archbishop William of Canterbury, and he was received there with great honor and with great ceremony, and he sang the high mass on Easter at the altar of Christ, and afterwards he went all over England, to all the bishoprics and abbacies which were in the land, and everywhere he was received with honor, and all the people gave great gifts and more, and later he held his council at London for three full days on the Nativity of St. Mary in September with archbishops and diocesan bishops and abbots, clerics, and laity and ordered there the same laws which Archbishop Anselm had ordered before (and many more although it amounted to little), and thence he went over the sea soon after St. Michaelmas and then to Rome (and Archbishop William of Canterbury) ... and (they) were received there by Pope Honorius with great honor, and (they) were there all winter.' (255.20-36) At times a relational adverb immediately followed the connective y and tended to make 17

A string of two means a T-unit plus a T-unit connected by y ; a string of three, a T-unit plus two T-units connected by y, etc.

206

THE RELATIONSHIP AND JOINING OF CLAUSES

the coordinative relationship more than a mere sequential one (y, y siddari). There were four examples, three of which began with y forpi and one with y popwaethere. An example of each follows: (1) Siööen £>a nam he ]ses kynges wifes swuster of France to wife, y forpi iaef se kyng him pone eorldom of Flanders. 'Afterwards then he took as wife the sister of the king of France, and, therefore, the king gave him the earldom of Flanders.' (257.

10-12) (2) y te king ferde agenes him mid micel mare ferd. y popwaethere fuhtten hi noht. 'And the king went against him with a much bigger army, and nevertheless, they did not fight.' (268.7-8) Closely related to this construction was one example without y before the relational adverb: y Dauid king of Scotland toe to uuerrien him. pa pohuuethere pat here sondes feorden betwyx heom. 'David, King of Scotland, began to make war on him. Then, despite that, their messengers went between them.' (263.21-22) Twenty-two T-units began with ac. There were eight examples with a string of two; the longest string of T-units with ac was three. An example each of ac forming a string of two and of ac combining with a T-unit introduced by y forming a larger string follows: (1) t>es ilee Willelm haefde aeror numen öes eorles dohter of Angeow to wife oc hi wxran sidden totweamde for sibreden. 'This same William had taken as wife the daughter of the count of Anjou, but they were later separated on account of consanguinity.' (257.8-9) (2) Des ilee gaeres he geaf Jsone abbotrice of Burch an abbot Heanri waes gehaten of Peitowe se haefde his abbotrice Sancte Johns of Angeli on hande ... oc se ilee Heanri dide pone king to understandene pet he haefde Ixten his abbotrice for pet micele unsibbe ...oc hit ne waes nadema eall swa. oc he whole hauen bade on hand.... 'This same year he gave the abbacy of Peterborough to an abbot who was called Henry of Poitou who already had his abbacy of St. Jean d'Angely in hand ... but this same Henry had the king understand that he had left his abbacy on account of the great strife ... but it was none the more so for all that, but he wanted to have both in hand ....' (257.13-22) Eleven T-units began with /or. 18 There were five examples with a string of two Tunits, and six other examples which had for joining the other connectives forming larger strings. An example of each set is presented below: 18

The investigator is aware that for is a shortened form of the subordinator forpi (and in some cases has been treated as such), but he feels that there were some uses which fit the description of causal coordinator. All of these occurred in the Final Continuation.

THE RELATIONSHIP AND JOINING OF CLAUSES

207

(1) I>a westre19 sona Jsas landes. for seuricman sone rzuede oper pe mihte. 'Then the lands soon became dark, for everyone who could robbed the other.' (263.6-7) (2) wende to begaston Normandi |)aer {wrh. oc he spedde litel. y be gode rihte for he was an yuel man. for ware se he com he dide mare yuel panne god. '(He) expected to get Normandy through that, but he profited little; and by good right for he was an evil man, for wherever he came, he did more evil than good.' (267. 30-32) Three T-units began with ne. One was part of a string of four T-units, and the other two were in the same string of four T-units. The examples are as follows: (1) I ne can ne i ne mai teilen alle pe wunder ne alle pe pines {jet hi diden wrecce men on J>is land, y fet lastede f>a -xix- wintre wile Stephne was king y aeure it was uuerse y uuerse. Ί cannot nor am I able to relate all the horrors nor all the pain that they did to the wretched men in this land; and it lasted the nineteen years while Stephen was king, and it always was worse and worse.' (264.24-27) (2) Wes naeure gast mare wreccehed on land, ne nzure hethen men werse ne diden pan hi diden for ouer sithon ne forbaren hi nouther circe ne cyrceixrd. oc namen al l>e god J)et J>ar inne was ... Ne hi ne forbaren biscopes land ne abbotes ne preostes. ac rsueden munekes y clerekes .... 'Never was there as yet more wretchedness in the land, nor have heathen men done worse than they did; for against custom, they neither respected the church nor church yard but took all the goods that were inside ... nor did they respect bishop's land nor abbot's nor priest's, but robbed monks and clerks ....' (264.35-265.3) There were 257 T-units which did not begin with a connective. One hundred fortytwo of these did not have a connective before or after the T-units. Of the 257 T-units, one hundred twenty-five began with adverbials such as prepositional phrases (On pa ilea tyme), adverb phrases (pus earmlice), modified-noun phrases {Des ylce geares), negatives (na), or adverbs (siddan). Five began with subordinators (pa hwile pet, gif etc.). Fifteen T-units began with the pa of the correlative construction (pa ... pa). Nineteen T-units began with verbs, eight of which were inverted constructions and eleven of which were verbs of subjectless constructions. The subject was the first element in the T-unit in seventy-two examples. Here is a list of the fillers of the subject slot with the number of times each occurred: dem., 20; pers. pron., 21; pronom. adj., 5; dem. + adj. + noun, 4; pronom. adj. + noun, 1; dem. + noun, 4; pronom. adj. + dem., 1; impers. pron. 4; proper noun, 6; dem. + pronom. adj., 2; rel. pron., 2; adj. + noun, 1; and neg., 1. Sixteen of the T-units began with objects which were manifested by the following fillers: dem., 4; pers. pron., 2; noun, 1; pronom. adj., 4; dem.-fnoun, 2; pronom. adj. + dem., 2; and pronom. adj. + num., 1. 19

This is Plummer's reading, but the analyst has accepted for his translation the emendation pestreden, suggested by N. R. Ker, loc. cit.

208

THE RELATIONSHIP AND JOINING OF CLAUSES

Three T-units began with subject complements; two of the slots were filled by an adj. + noun and one by an adv. + adj. + noun. It seems then that even though there was not enough evidence to state conclusively that coordination was a structural device, the corpus was not merely a loosely connected string of word groups as other early portions of the Anglo-Saxon Chronicle have been characterized. 6.5. PARATAXIS

In the forming of syntactic structures larger than the clauses, some combinations of clauses are loosely connected, and others are closely joined in their structure. When a structure consists of an independent clause and a dependent clause not only notionally connected but formally joined by a connective, it is said to be HYPOTACTIC. The T-units with nominal, adjectival and adverbial dependent clauses joined to the independent clause by an overt subordinator are examples of hypotactic structures. When an independent clause and dependent clause are placed next to each other showing a logical relationship, but not formally joined by a subordinator, the structure is said to be PARATACTIC. The distinctive feature then of parataxis, which is generally characterized as often loose and clumsy and considered to be the oldest sentenceforming technique, is the absence of an overt subordinator in syntactic combination. When considering Old English syntax one may use parataxis, according to S. O. Andrew in Syntax and Style in Old English, in two senses. The first he describes as "simply a lack of grammatical subordination such as we find in the language of children and some primitive peoples".80 To illustrate what he meant, he presented the following examples: (1) Then I asked him; then he replied .... (2) They came to a place on the road; there stood a temple. (3) There lived in the convent a certain monk; he was called Martin: he said .... These sentences, Andrew called immature and almost childish, and characteristic of certain Old English prose, especially the earlier annals of the Anglo-Saxon Chronicle. The earlier discussion on coordination in this study noted the relatively extensive use of such time indicators as pa and siödan in beginning T-units. The following five T-units, for example, occurred in sequential order: (1) Siödon Jjurh his micele wrences öa beiaet he Jione aercebiscoprice of Besencum .... 'Afterwards by his great trickery he got the archbishopric of Besancon ....' (257. 27-28) (2) Pa forlaes he fet mid rihte .... 'Then he lost it justly ....' (257.29) (3) siddon J>a beiet he ]x>ne biscoprice of Seintes Jjet was fif mile fram his abbot rice. 10

S. O. Andrew, Syntax and Style in Old English (Cambridge, Cambridge University Press, 1940), p. 87.

THE RELATIONSHIP AND JOINING OF CLAUSES

209

'Then he got the bishopric of Seintes which was five miles from his abbacy.' (257.30-31) (4) Pet he hsfde ful neah seoueniht on hande. 'That he possessed for almost a week.' (257.31-32) (5) peonon brohte se abbot him of Clunni swa swa he aeror dide of Besencum. 'Then the abbot of Cluny brought him away as he had done earlier from Besancon.' (257.32-33) This string was even preceded and followed by similar T-units. This type of patterning is not uncommon in this corpus, but it is not as extensive as it is in earlier parts of the Anglo-Saxon Chronicle. Andrew commented on this kind of parataxis this way: ... we have seen that many of the supposed forms of principal sentence can really only be subordinate, while on the other hand some possible forms of principal sentence areambiguous, i.e., may equally well be dependent clauses.21 Andrew's description, of course, may be true of certain texts of Old English prose and poetry, and for that matter, of the five T-units above. But because of the extensive use of a variety of subordinators22 and hypotactic constructions (not even considering the correlative patterns), there was generally less ambiguity in distinguishing independent and dependent clauses in this corpus and, therefore, some justification for not considering the sequences given above as distinctively paratactic. These sequences and others in the corpus, however, demonstrated the first kind of parataxis Andrew described, that is, in the sense that this series of T-units lacked subordination. Andrew described the second kind of parataxis as a rhetorical device "idiomatically used to indicate subordination". 23 The first example he gave of this included sentences beginning with wolde, wende, and cwsed, which he thinks were equivalent to an adverbial clause of purpose or reason. In this corpus there were three units beginning with wende or wendan and one with wolde. All of these were segmented as subjectless T-units, but the notional relationship with the preceding clause is close enough to consider the clauses with wende and wolde as adverbial clauses of cause and purpose. Examples are as follows: (1) Gif twa men oJ>er ·ΐϋ· coman ridend to an tun. al £>e tunscipe flugsn for heom. wendan pet hi wseron rseueres. 'If two or three men came riding to a village, the people of the village fled from them (because) (they) expected that they were robbers.' (265.3-5) (2) On {jis gaer com Dauid king of Scotlande mid ormete faerd to Jdis land, wolde winnan pis land. 'In this year David, King of Scotland, came to this country with a very large army (because) (he) wanted to win this land.' (266.3-4) " " pa "

Ibid. Andrew also commented (p. 91) on the use of only a few subordinators and the heavy reliance on in earlier Old English prose and poetry. Ibid.

210

THE RELATIONSHIP AND JOINING OF CLAUSES

In this paratactic structure (subjectless), as is evident, the predicator stood first and contiguous to the last word of the preceding clause. Another type of paratactic structure recognized by Andrew and found in this corpus was what has been treated in this study as a subjectless adjectival clause. There were fifteen examples of this structure. In all of these the relative pronoun was omitted, and the subject complement (a proper noun) stood first in all the examples except one. In the one that it did not, the predicator stood first and the subject complement last. Eleven of the examples occurred with wzs gehaten filling the predicator slots; three appeared with hatte (het) and one with wses. An example of each variety is as follows: (1) On J?es ilces gaeres sende se papa of Rome to öise lande an cardinal Iohan of Creme wxs gehaten. 'In this same year the pope of Rome sent to this land a cardinal (who) was called John of Crema.' (255.17-18) (2) he was canonie of an mynstre Cicc hatte. 'He was canon of a monastery (which) was called Cicc.' (252.1) (3) y mid him ferde se biscop Bernard of Wales ... y Gifard wses pes kinges hird clerc. 'And with him went Bishop Bernard of Wales ... and Giffard (who) was the king's court clerk.' (252.18-21) Andrew recognized other paratactic structures, but only the types that have been illustrated were found in this corpus. If the first kind of parataxis (lack of subordination) can be looked upon as the state of the art of composition of the period, then it is safe to say that parataxis as a rhetorical device was used very little in this corpus.

6.6. CORRELATION

If clause combination in the English language can be described as having moved from parataxis to hypotaxis, perhaps it then can be said that correlation was a stage existing somewhere between the two in the development of grammatical subordination. Whereas there is not an overt unit connecting two closely related clauses in paratactic structures, there is an overt connective called the subordinator in hypotactic structures. In correlative structures there are normally two units present, and they are identical most of the time. Correlation can be defined then as a means of expressing a notional relationship that exists between two clauses usually by means of two units (corresponding words) which, as a rule, introduce the two clauses. In this corpus there were several kinds of correlative constructions, most of which fit the definition as given above. There were a few, however, which did not. Some of these were appositional in nature, and some had only a single word between the two clauses. The most frequent correlative construction was the kind which expressed the demonstrative and relative notions between two clauses by having each introduced by

THE RELATIONSHIP AND JOINING OF CLAUSES

211

a connecting unit. Of the twenty-eight constructions fitting this description, seventeen were joined by pa ... Pa, three by pa hwile pet... pa, two by swa ... swa, and one by each of the following: sone swa ... swa, swa radlice swa ... pa, xr... pa, purhpet... pa, gif... alse, and noht forpipet ... oc. An example of each is as follows: (1) pa hi pider comon da ne was hit noht buton laesunge. 'When they came to that place, then it was nothing but lies.' (259.21-22) (2) pa hwile pet he pxr wxs pa geaf he Jsone biscoprice of Lincolne an clerc Alexander waes gehaten. 'While he was there, then he gave the bishopric of Lincoln to a clerk (who) was called Alexander.' (253.1-2) (3) y swa mycel hearm pxr wxs gedon swa nan man hit cuöe o£>er secgen. 'And so much injury was done there that no one was able to tell about it.' (253.18-19) (4) t>a sone swa he hxfde wind swa ferde he ofer in to Normandie .... 'Then as soon as he had wind, then he went across the sea to Normandy ....' (253.6-7) (5) ... swa radlice swa he pxr com ...pa son t>aer after J>a sasgon y herdon fela men feole huntes hunten. ' . . . a s soon as he came there ... then soon after that, many men saw and heard many hunters hunting.' (258.16-19) (6) Mr he were wel ded. pa waere f>aer coren twa papes. 'Before he was dead very long, then there were two popes elected.' (260.17-18) (7) par xfter purh pet he wxs des kynges mxi of Engle land ...pa geaf se eorl him Jjone abbo trice .... 'After that because he was a kinsman of the king of England, then the count gave him the abbacy ....' (257.25-26) (8) y gif he leng moste liuen. alse he mint to don of J» horderwycan. 'And if he could live long enough, he meant to do the same with the treasury.' (265.22-23) (9) £>er efter waex suythe micel uuerre betuyx J>e king y Randolph eorl of Caestre noht forpi pet he ne iaf him al pet he cuthe axen him. alse he dide alle othre. oc ffifre t»e24 mare he iaf heom J>e waerse hi waeron him. 'After that there grew a great war between the king and Rannulf, count of Chester, not because he did not give him all that he asked him for just as he did all the others, but the more he gave them, the worse they were to him.' (266.16-19) Sometimes the dependent clause may have other adverbial elements preceding it as in: y pet oper dei pa he lai an slep in scip. pa Jwstrede }>e daei ouer al landes. 'The other day when he lay asleep in the ship, the day became dark over all the land.' (262.34) Another special construction had a noun phrase plus an adjectival clause bearing the relative meaning temporal and serving as part of the correlative. The example is as follows: Pat ilce dxi pat Martin abbot of Burch sculde pider faren. pa saeclede he .... 'The same day that Abbot Martin went to that place, then he became sick....' (268.29-30) 21

The interdependent clause from here on could possibly be considered a correlative construction: fie mare ...fie wxrse.

212

THE RELATIONSHIP AND JOINING OF CLAUSES

Correlation may also occur in constructions where adjectival or nominal dependent clauses were in apposition to one of the nominal tagmemes of the independent clause. Elsewhere these have been treated as appositives and double subjects or double objects. The four examples are given below: (1) se man pe seni god heafde him me hit beraefode mid strange geoldes y mid strange motes. 'The man who had any property, one deprived him of it by severe taxes and severe courts.' (255.3-4) (2) de hafde an pund he ne mihte cysten aenne peni at anne market. 'He who had a pound, he could not get the value of a penny at the market.' (255.10) (3) y se man pa hefde twa hundred oÖpe öre hundred swin ne beleaf him noht an. 'And the man who had two or three hundred pigs, not one was left for him.' (261.22-23) (4) Wua sua bare his byrthen gold y sylure. durste nan man sei to him naht but god. 'Whosoever bore his burden in gold and silver, no one dared to say to him nothing but good.' (263.10-12) A variation of this is illustrated by the following example: ... £>a weorö hit eall of earcedsecnes wifes y of preostes wifespet hi scolden hiforlxten be Sancte Andreas messe. '... then it was all about archdeacon's and priest's wives, that they should give them up by Saint Andrew's Mass.' (260.2-3) Some ol the correlative constructions had a single word between the clauses serving as both the demonstrative and relative. One of the three examples is given below: On f>is gaer waerd £>e king Stephne ded y bebyried per his wif y his sune wxron bebyried at Fauresfeld. 'In this year King Stephen died and was buried there where his wife and son were buried at Faversham.' (268.21-22) Closely related to these correlative structures were two constructions which consisted of a pronominal adjective plus adjectival clause and swa plus an independent clause. These constructions, of course, were correlative only if swa is accepted as a pronoun. An example follows: Eall J>et Jse beon dragen toward swa frett f>a drane .... 'All that the bees dragged out, that the drones ate ....' (258.12)

6.7. G E N E R A L CONCLUSIONS

The purpose of this chapter has been to point out the relationships and functions of clauses and to focus on some of the connecting devices by which they are joined. As a result, the following conclusions can be drawn: (1) The corpus contained single clause and multiple clause T-units.

THE RELATIONSHIP AND JOINING OF CLAUSES

213

(2) At first sight, clause combination in this corpus seemed to be awkward and clumsy, but a closer examination showed that different relationships were generally expressed effectively, though not always with the precision and smoothness of Modern English prose. (3) Some clauses were loosely joined in their structure (paratactic), while others, the great majority, were closely joined in their structure (hypotactic). (4) On the basis of function and the notional relationships they express, dependent clauses can be classified as nominal, adjectival, and adverbial. (5) Dependent clauses had a variety of functions; they might manifest clause-level nominal and adverbial slots, phrase-level nominal, adjectival, and adverbial slots or have a modifying relationship with the T-unit as a whole. (6) Although the dependent clauses were more various, expressed diverse relationships, and were introduced by many different connecting tagmemes, the internal structure (patterns) did not vary greatly from that of independent clauses, excepting that the subject was more likely to stand before the predicator in dependent clauses. (7) Nominal clauses typically operated at the clause and phrase levels; adjectival clauses operated at the phrase level; and adverbials operated at the clause, phrase, and T-unit levels. (8) The nominal clause as filler of the subject slot might occur both before and after the predicator and as filler of the direct object slot only after the predicator. (9) The adjectival clause as filler of a modifying slot of a modified-noun phrase, as a rule, stood after and adjacent to the head of the phrase, but occasionally was separated from its head. (10) The adjectival clauses and some nominal clauses were joined by subordinators which also filled primary slots of the clause they introduced. (11) Adjectival clauses might occur with or without overt subordinators. (12) The adverbial clause, which might express a variety of notional relationships, might occur both before and after the independent clause of the T-unit. (13) The adverbial clause was joined by a variety of subordinators, which, for the most part, did not fill slots within the boundaries of the clauses they introduced. (14) Coordination, though present in the corpus, apparently was not used as a deliberate structural device for showing notional and structural relationships. (15) Parataxis except for certain formulaic clauses (adjectival clauses) was not extensively used. (16) Frequently the joining of clauses was supported by correlation.

7 SUMMARY AND CONCLUSIONS

The portion of the Peterborough Chronicle investigated in this study has generally been considered by scholars such as Plummer, Whitelock, Ker, and Clark 1 to be an uncorrupted original document written and composed at Peterborough during, for the most part, the earlier half of the twelfth century and is the earliest extensive example of the East Midland Dialect, the dialect thought to be the ancestor of Modern English. In these pages an effort has been made to describe the syntactic patterning of the language of the First and Final Continuations of the Peterborough Chronicle from the years 1122 to 1154. Although the chief concern of this study has been the syntactic features of the corpus, the most significant and relevant relations between morphology and syntax, especially case inflections and the expression of the subject, indirect object, and direct object relations, have been briefly noted. Using methods of analysis, description, and reporting derived and adapted from both the scholarly traditional grammars and the techniques of modern descriptive linguistics, more particularly the work of Benjamin Elson and Velma Pickett, An Introduction to Morphology and Syntax,2 the investigator centered his attention on the order of primary and secondary clause-level functional slots and their fillers, the various modified-head phrases, and the function and relationship of clauses and the modes of joining them. There were some procedures used in this study believed to be improvements over those of some of the other recent and similar studies. They are briefly listed here: (1) separate frequency counts and descriptions for independent and dependent clauses, (2) the use of the tagmeme as a correlative of functional slot and class filler, (3) the use of levels of analysis, (4) separate frequency counts and descriptions for SVO and SVIO patterns, (5) a separate frequency count and description for the coordinated predication, (6) a treatment of adverbials on the basis of their separability from 1

John Earle and Charles Plummer, Two of the Saxon Chronicles Parallel (London, Oxford University Press, 1892-99), II, p. xxxv; Dorothy Whitelock, The Peterborough Chronicle (Copenhagen, Rosenkilde and Bagger, 1954), pp. 13-35; N. R. Ker, Catalogue of Manuscripts Containing AngloSaxon (Oxford, The Clarendon Press, 1957), pp. 424-426; and Cecily Clark, The Peterborough Chronicle 1070-1154 (Oxford, Oxford University Press, 1958), pp. xxx-lxvi. ' Benjamin Elson and Velma Pickett, An Introduction to Morphology and Syntax (Santa Anna, California, Summer Institute of Linguistics, 1964).

SUMMARY AND CONCLUSIONS

215

the predicator, and finally, (7) the use of the minimal terminable syntactic unit, or T-unit, for segmentation of the corpus. Because of the scribe's tendency to string structural units together by the symbol y, the investigator posited an arbitrary provisional unit called the minimal terminable syntactic unit, or T-unit, with which he could begin the analysis. The T-unit proved to be a workable and effective unit by means of which the investigator could not only determine and analyze the primary and secondary clause-level tagmemes, but also describe the function and combination of clauses filling the predication slots of the clauses and T-units. A survey of the formal case inflections confirmed the observations of Behm, Meyer, and Clark 3 that the inflectional system was in a state of dissolution and decomposition and so irregular that it was, as a rule, incapable of consistently and distinctively marking the subject, indirect object, and direct object in most patterns. The disintegration of the inflectional system of this corpus should not be too surprising, for Saitz, it was pointed out, demonstrated that even the Classical Old English inflectional system (case endings) could not have distinguished the subject-object relationship in 59 per cent of the patterns he studied. 4 Later, Carlton concluded that, even though inflections were generally distinctive in his corpus, word order was regular enough to be a supplementary syntactic signal.5 Shannon, furthermore, showed that nouns, pronouns, and modified-noun phrases as fillers of the subject and direct object slots were ambiguous in case in over two-thirds of the patterns in her corpus.® In comparison with the declensional system of much earlier stages of the English language, which has generally been characterized as full and complete (and which Saitz, Carlton, and Shannon have shown not to be), the inflectional system of this corpus was exceedingly disintegrated and simplified. Fillers of the primary nominal slots were not always distinctively marked. In the First Continuation, though inflectional distinctiveness was greater than that of the Final Continuation, the indirect object and direct object were frequently indistinguishable by case endings such as shown in these two examples: Pais he makede men and daer 'He made peace for men and animals' and y iaf hine pone eorldom 'and (he) gave him the earldom'. The situation was different with the subject, but the distinctiveness of the marked subject was due to the masculine forms he and se, the latter of which was used with other functions and for all nouns despite their original gender. In the Final Continuation the accusative gave way to the dative (i namen him 'and captured him'). Furthermore, * O. P. Behm, The Language of the Latter Part of the Peterborough Chronicle (Uppsala [n.n.], 1884); Heinrich Meyer, Zur Sprache der jüngeren Teile der Chronik von Peterborough (Leipzig, Buchhandlung Gustav Fock, 1889); and Cecily Clark, loc. cit. 4 Robert L. Saitz, "Functional Word Order in Old English Subject-Object Patterns", unpublished Ph.D. dissertation, University of Wisconsin, Madison, 1955, pp. 4-41. 4 Charles R. Carlton, Descriptive Syntax of the Old English Charters (The Hague, Mouton, 1970), p. 193. • Ann Shannon, A Descriptive Syntax of the Parker Manuscript of the Anglo-Saxon Chronicle from 734 to 891 (The Hague, Mouton, 1964), pp. 55-56.

216

SUMMARY AND CONCLUSIONS

the subject was usually only distinctively marked when the nominative forms of the personal pronoun filled the subject slot. In the 234 patterns with subject and direct object slots manifested by nouns, noun phrases or pronouns, twenty-nine, or 12 per cent, of the patterns occurred with a subject with a distinctive nominative ending and with an object with either a distinctive accusative or dative ending.7 Ninety, or 38 per cent, of the patterns occurred with neither the subject nor direct object distinctively marked. The remaining 115, or 50 per cent of the patterns occurred with only one function distinctively marked. The arbitrariness of the formal case inflections becomes even more pronounced, considering that, when the subject was distinctively marked, it was marked by se, which, as was just pointed out, was used with most nouns, no matter what the original gender had been, and by the pronoun he, the nominative form of present-day English. The ambiguity here, however, can be resolved by inflections for number, context, lexical association, and word order. For most patterns, as a general rule, lexical compatibility could determine the functions. As implied in the foregoing paragraphs, word order, if this corpus can be considered indicative of the language of the period, must have played, if not the primary role, at least a very important one in distinguishing the subject-object relationship and in showing the direction of modification of modified-noun phrases. Although the primary tagmemes take up various positions relative to each other, the statistically prominent clause patterns were SV, SVO, SVC, and SVIO. In predications with only the predicator and object, the order by a wide margin was VO. The percentages of patterns in the various groups were generally closer in distribution in independent clauses than in dependent clauses. For example, in Group 1 (subject and predicator), the distribution was 51 per cent for the SV order and 49 per cent for VS in independent clauses, whereas it was 96 per cent for the SV order and 4 per cent for the VS order in dependent clauses. Inversion of subject and predicator was functional to a large degree when adverbials occupied the initial position in the clause. Because of the decided tendencies to relatively consistent and limited patterns, it must be concluded that word order was an important syntactic signal in this corpus. Yet the modernity of the word order can be exaggerated. A comparison with the word order of the Parker Manuscript of the Anglo-Saxon Chronicle from the years 734 to 891 is interesting. Subjoined below are figures (percentages) from Shannon's study for comparison. 8

SV VS SVO 7 8

Parker MS. (734-891) 64 36

Peterborough MS. (1122-1154) 67 33

30

51

Six of these twenty-nine objects were dative. Shannon, op. tit., p. 13.

217

SUMMARY AND CONCLUSIONS

sov vso vos osv ovs

35

SVC

86

scv vsc

9

CSV CVS vo ov

18

17 13

3

1

11

15 3

3

6 0 0 18 82

76 9 9 4

2 91 9

The differences in approaches in these studies prevent broad generalizations, but the comparison may be instructive. First of all, it shows that the syntactic patterning of the primary tagmemes of this corpus, excepting the VO and OV patterns, is not necessarily as modern as scholars have said. Second, it seems unusual for two corpora, one early and having a relatively full inflectional system and the other comparatively late with an arbitrary and confused inflectional system, to have such a close variability in clause patterns. And third, this comparison demonstrates the continuity in stages of a language and shows that primary tagmemes were placed in favorite patterns very early in the history of the English language, perhaps giving some support to the view that word order preceded inflectional loss. And, furthermore, since there were favorite patterns, the comparison shows that the primary tagmemes were not just placed anywhere as is implied by this account from a recent book: The syntax of Old English is inseparable from its morphology, because the syntax is immediately dependent upon the morphology. Synthetic rather than analytic, Old English syntax is the result of a full inflectional system. When inflections indicate grammatical relationships, syntactical positioning becomes secondary. In a synthetic language like Old English, therefore, WORD ORDER HAS LITTLE OR NO INFLUENCE UPON MEANING: THE INFLECTIONAL MORPHEMES DO THE WORK OF COMMUNICATION. [Emphasis mine.]®

It is also generally stated that the direction of modification in the modified-noun phrase was indicated by case inflections and that, as a result, the modifiers could stand almost anywhere. This may be possible, but the fact of the matter is that the modifiers had rather well-established linear positions with respect to the head and to each other. Both Carlton and Shannon acknowledged the distinctiveness of inflections in the modified-noun phrase, but they (particularly Carlton) also demonstrated that word order was operative. In fact, Carlton showed that there was a rather fixed linear ordering of the modifiers in the modified-noun phrase. Except for some very • John Nist, A Structural History of English (New York, St. Martin's Press, 1966), p. 124. Statements like this reveal either an ignorance of, or a lack of interest in, the recent studies accumulating information on early English syntax.

218

SUMMARY AND CONCLUSIONS

special features, the results of this study, of Carlton's, and Palmatier's10 indicate explicitly that the configuration of the modified-noun phrase of earlier periods of the English language is essentially that of Modern English described by Hill and Gleason.11 In this corpus, if all the modifying slots co-occurred, they would have to occur in this order: (1) pre-determiner (2) determiner (3) quantifier (4) particularizer (5) qualifier (6) adjunctor (7) head (8) locater (9) classifier (10) identifier. At times there were occasional departures from this linear ordering, but they were so few that they could be considered anomalous. 12 Since gender and case inflections in the modifiednoun phrase were used haphazardly and since there was a rather well-established linear ordering of the modifiers relative to the head and each other, word order was apparently the primary syntactic signal in indicating the direction of modification. The adverbiale, traditionally treated as 'verb-modifiers', can most effectively be handled on the basis of their separability from the predicator. On this basis, a distinction can be made between those adverbials which tended to form a rather cohesive phrasal unit with the verb of the predicator slot and those that tended to take up various positions in the clause. In this study, the first were called adverbial modifiers of the modified-verb phrase and the second clause-level adverbials. The clause-level adverbials were discussed first and divided into seven types on the basis of their function and the meaning they expressed: (1) the introductory formula tagmeme, (2) the time tagmeme, (3) the location tagmeme, (4) the manner tagmeme, (5) the purpose tagmeme, (6) the agent tagmeme, and (7) the relationship tagmeme. It was difficult to establish any consistent and fixed rules about the relative positions of these adverbials, but there were three significant facts worth noting here. First, despite the latitude in the placing of the clause-level adverbials, all except the introductory formula were more frequently placed after the predicator and whatever other primary tagmemes that might be present. Second, the slots filled by light fillers (adverbs) generally were more variable and stood before those with heavy fillers (prepositional phrases or clauses). Third, restriction of position apparently was due to the type of the filler rather than to the function and meaning of the slot. The modified-verb phrase, as treated here, was a relatively close-knit phrasal unit consisting of a verb (simple or complex) and the negator, the aspect indicator, the degree indicator, or direction indicator. These adverbials, although they sometimes stood before, between the finite and non-finite forms of, and after the verb of the predicator slot, as a rule stood adjacent to it, serving to express such notional rela10

Robert A. Palmatier, A Descriptive Syntax of the "Ormulum" (The Hague, Mouton, 1970). Archibald Hill, An Introduction to Linguistic Structures (New York, Harcourt, Brace and Company, 1958), pp. 230-255; and H. A. Gleason, Jr., Linguistics and English Grammar (New York, Holt, Rinehart, and Winston, Inc., 1965), pp. 409-412. " For an excellent comparison of the modified-noun phrase in Old English, Early Middle English, and Modern English, see Palmatier, op. cit.. For a comparison of the modified-noun phrase in Old English, Modern English and Modern German, see Gleason, loc. cit. 11

SUMMARY AND CONCLUSIONS

219

tionships as negation, aspect, degree, and direction. Multiple negation was the rule rather than the exception, occurring in 69 per cent of the examples.18 Important also is the fact that the auxiliary verb never occurred after the main verb. Although the clauses in the corpus appeared at times to be awkwardly joined and lacked the precision and smoothness of Modern English, they were generally combined effectively and by a variety of conjunctions. At times, adjectival clauses were separated from the head word. Some clauses were loosely joined in their structure (paratactic), while the majority were closely joined in their structure (hypotactic). Frequently the joining of clauses was supported by correlation. Except for the great preference of SV order over VS order in dependent clauses, clause patterns of independent and dependent clauses did not vary considerably. There is general agreement that the latter part of the Peterborough Chronicle is 'precious' among medieval texts for the history of the English language. The question of its place in the growth and development of the English language is controversial and apparently still an open one. Is it Old English or Middle English? Or is it in a transitional stage? Cecily Clark, who feels that it should be considered Middle English, acknowledged that "the weight of opinion has inclined to classing it as Old English".14 Rothstein, quite a few years earlier, implied that the word order was modern.1® Bruce Mitchell questioned the modernity of the language. In his article, the aim of which was to show what is modern and what is archaic about the language, Mitchell concluded that as far as inflectional reduction and the presence of some analytic constructions are concerned, Clark may be right; but as far as word order and several other syntactic features are concerned, she is wrong and asserted that the only claim to modernity of this portion of the Peterborough Chronicle is the 'complete absence' of the S. noun OV pattern.16 This study, the main purpose of which has been descriptive and not comparative, corroborated the observations of Mitchell that the word order is closer to Old English than to Modern English, despite the inflectional loss and presence of analytic constructions. The study, however, questioned Mitchell's statement that the S. noun OV pattern did not occur at all. Earlier in the present chapter a comparison with Shannon's analysis of a much earlier portion of the Anglo-Saxon Chronicle showed that the word order of this corpus was close to Classical Old English. A further comparison may even be more instructive. Below are appended the analyses of the clause patterns of the Parker Manuscript (734-891),17

18 It is interesting to note that Charles C. Fries remarked in American English Grammar (New York, Appleton-Century-Crofts, Inc., 1940), p. 35, that in Old English the multiple negative occurred in less than 35 per cent of the total number of negative statements. 14 Clark, op. cif., p. xlix. u Ewald Rothstein, "Die Wortstellung in der Peterborough Chronik", Studien zur englischen Philologie, LXIV (1922), 108. " Bruce Mitchell, "Syntax and Word Order in the Peterborough Chronicle 1122-1154", Neuphilologische Mitteilungen, LXV (1964), 113-144. 17 Shannon, loc. cit.

220

SUMMARY AND CONCLUSIONS

the Pastoral Care (900),18 the Ormulum (1200),19 and the Peterborough Chronicle for comparison: Parker MS. (734-891) 64 36

Pastoral Care (900) 82 18

svo SOV vso vos osv ovs

30 35 18 3 11 3

26 44 6 1 21 2

51 17 13 1 15 3

62 14 11 1 8 4

SVC scv vsc CSV CVS vcs

86 9 6 0 0 0

70 13 2 9 5 0

76 9 9 4 2 0

77 8 1 3 2 9

vo OV

18 82

51 49

91 9

75 25

sv vs

Peter. MS. (1122-1154) 67 33

Ormulum (1200) 84 16

Again the differences in analyses assumed in these studies must be kept in mind. Mitchell based his refutation of Clark and Rothstein on the similarity of word order patterns between an earlier corpus and the one investigated in this study and the presence of historically anomalous forms. In trying to determine the place of a given corpus in the growth and development of a language, the solution may not necessarily lie, as Mitchell would have us believe, in the carry-over of remnants in a later stage from an earlier stage of a language; continuity is just as much a part of the historical stages of a language as change is. The figures appended above reveal, except for the decrease in SOV and OV patterns, that the change in syntactic positioning was not greatly dramatic and that there was a decided preference for certain similar patterns in all these corpora. It is realized that statistics can be misleading in trying to settle the matter of historical position of a corpus, but so can too much emphasis on historical anomalous forms. Like Rothstein and Clark, Mitchell produced a 'misleading impression' when he implied that iElfric would not have felt that the word order of the Peterborough Chronicle differed from his own.20 The above figures show that people from any of these periods 18

William H. Brown, Jr., A Descriptive Syntax of King Alfred's "Pastoral Care" (The Hague, Mouton, 1970). " Palmatier, op. cit.. " Mitchell, op. cit., p. 122.

SUMMARY AND CONCLUSIONS

221

would have had little difficulty with any of the other word orders. And for that matter, people of the fourteenth and fifteenth centuries might also have had little difficulty with them. Quibbling about whether the syntax is Old English, Middle English, or Modern English seems to be trifling and is like asking to be engulfed in quicksand. The important thing to note is that continuity is evident in all stages of language and that word order, the primary syntactic signal in present-day English, played an important role in signaling syntactic relationships in earlier periods of the English language, as this study makes clear.

BIBLIOGRAPHY

WORKS RELATING TO EARLY ENGLISH GRAMMAR AND SYNTAX Adams, Arthur, The Syntax of the Temporal Clause in Old English Prose (New York, Henry Holt and Company, 1907). Andrew, S. O., "Some Principles of Old English Word-Order", Medium JEvum, III (1934), 167-188. , Syntax and Style in Old English (Cambridge, Cambridge University Press, 1940). Brown, William H., Jr., A Descriptive Syntax of King Alfred's "Pastoral Care" (Janua Linguarum, Series Practica, 101) (The Hague, Mouton, 1970). Brunner, Karl, Outline of Middle English Grammar (Cambridge, Harvard University Press, 1963). Carlton, Charles R., Syntax of the Old English Charters (Janua Linguarum, Series Practica, 111) (The Hague, Mouton, 1970). , "Word Order of Noun Modifiers in Old English Prose", Journal of English and Germanic Philology, LXII (1963), 778-783. Cassidy, Frederic G., "The Backgrounds in Old English of the Modern English Substitutes for the Dative Object in the Group Verb + Dative Object + Accusative Object". Unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1938. Fries, Charles C., American English Grammar (New York, Appleton-Century-Crofts, Inc., 1940). , "On the Development of the Structural Use of Word-Order in Modern English", Language, XVI (1940), 199-208. Harris, David Payne, "Word-Order in Twelfth Century English", Studies in Language and Linguistics in Honor of Charles C. Fries, edited by Albert H. Marckwardt (Ann Arbor, The English Language Institute, University of Michigan, 1964), 187-198. Kruisinga, Etsko, "How to Study Old English Syntax", English Studies, VIII (1926), 44-49. Kurath, Hans and Sherman Kuhn (eds.), Middle English Dictionary (Ann Arbor, University of Michigan Press, 1952-). Magers, Mildred K., "The Development of the Grammatical Use of Word-Order for Relationships Expressed by the Accusative with Special Reference to the Development in Subordinate Clauses". Unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1943. Magoun, F. P., "Colloquial Old and Middle English", Harvard Studies and Notes in Philology and Literature, XIX (1937), 167-173. Malone, Kemp, "When Did Middle English Begin?" Curme Volume of Linguistic Studies·. Language Monographs, VII (1930), 110-117. Moore, Samuel and Albert H. Marckwardt, Historical Outlines of English Sounds and Inflections (Ann Arbor, George Wahr Publishing Company, 1960). Mosse, Fernand, A Handbook of Middle English (Baltimore, The Johns Hopkins University Press, 1952). Nist, John, A Structural History of English (New York, St. Martin's Press, 1966). Palmatier, Robert Α., A Descriptive Syntax of the "Ormulum" (Janua Linguarum, Series Practica, 74) (The Hague, Mouton, 1970).

BIBLIOGRAPHY

223

Quirk, Randolph and C. L. Wrenn, An Old English Grammar (New York, Holt, Rinehart and Winston, Inc., 1957). Saitz, Robert L., "Functional Word-Order in Old English Subject-Object Patterns". Unpublished Ph.D. dissertation, University of Wisconsin, Madison, 1955. Shannon, Ann, The Descriptive Syntax of the Parker Manuscript of the Anglo-Saxon Chronicle from 734 to 891 (Janua Linguarum, Series Practica, 14) (The Hague, Mouton and Company, 1964). Small, George W„ "On the Study of Old English Syntax", PMLA, LI (1936), 1-7. Smith, C. Alphonso, "The Order of Words in Anglo-Saxon Prose", PMLA, VIII (1893), 210-243. Swieczkowski, Walerian, Word Order Patterning in Middle English (The Hague, Mouton and Company, 1962). Thomas, Russell, "Syntactical Processes Involved in the Development of the Adnominal Periphrastic Genitive in the English Language". Unpublished Ph.D. dissertation, University of Michigan, Ann Arbor, 1931. Vallins, G. H., The Patterns of English (London, Andre Deutsch, 1956).

WORKS RELATING TO THE PETERBOROUGH CHRONICLE Behm, O. P., The Language of the Latter Part of the Peterborough Chronicle (Uppsala, [n.n.], 1884). Clark, Cecily, "Gender in the Peterborough Chronicle, 1070-1154", English Studies, XXXVIII (1957), 109-115. , "Studies in the Vocabulary of the Peterborough Chronicle, 1070-1154", English and Germanic Studies, V (1952-3), 63-89. , The Peterborough Chronicle 1070-1154 (Oxford, Oxford University Press, 1958). Earle, John and Charles Plummer, Two of the Saxon Chronicles Parallel, 2 vols. (London, Oxford University Press, 1892-98). Garmonsway, G. N., The Anglo-Saxon Chronicle (London, Everyman, 1953). Ker, N. R., Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, The Clarendon Press, 1957). , "Some Notes on the Peterborough Chronicle", Medium /Evum, III (1934), 136-138. Mcintosh, Angus, "The Relative Pronouns pe and pat in Early Middle English", English and Germanic Studies, I (1947-8), 73-90. Meyer, Heinrich, Zur Sprache der jüngeren Teile der Chronik von Peterborough (Leipzig, Buchhandlung Gustav Fock, 1889). Mitchell, Bruce, "Syntax and Word-Order in The Peterborough Chronicle 1122-1154", Neuphilologische Mitteilungen, LXV (1964), 113-144. Roth, Wilhelm, Die Wortstellung im Aussage-Hauptsatz Angelsächsischen Originalprosa, Annalen 800-900, 1066-1154 (Berlin, Mayor and Miller, 1914). Rothstein, Ewald, "Die Wortstellung in der Peterborough Chronik", Studien zur englischen Philologie, LXIV (1922), 1-108. Tucker, S. I., "The Anglo-Saxon Chronicle (1042-1154)" in English Historical Documents 1042-1198. Edited by David C. Douglas and George W. Greenway (London, Eyre and Spottiswoode, 1953). Whitelock, Dorothy, The Peterborough Chronicle (Copenhagen, Rosenkilde and Bagger, 1954). WORKS RELATING TO ENGLISH GRAMMAR, LANGUAGE, AND LINGUISTIC METHODS Bryant, Margaret M., A Functional English Grammar (Boston, D. C. Heath and Company, 1959). Elson, Benjamin and Velma Pickett, An Introduction to Morphology and Syntax (Santa Anna, California, Summer Institute of Linguistics, 1964). Francis, W. Nelson, The English Language: An Introduction (New York, W. W. Norton and Company, 1965). , The Structure of American English (New York, The Ronald Press Company, 1958). Fries, Charles C., The Structure of English (New York, Harcourt, Brace and Company, 1952).

224

BIBLIOGRAPHY

Gleason, Henry Α., Jr., An Introduction to Descriptive Linguistics (New York, Henry Holt and Company,1955). , Linguistics and English Grammar (New York, Holt, Rinehart and Winston, 1965). Harris, Zellig S., Methods in Structural Linguistics (Chicago, University of Chicago Press, 1951). , String Analysis of Sentence Structure (The Hague, Mouton and Company, 1962). Hart, H., "Hierarchical Structuring of Amuzgo Grammar", International Journal of American Linguistics, XXIII (1957), 141-164. Hill, Archibald, An Introduction to Linguistic Structures (New York, Harcourt, Brace and Company, 1958). Hughes, John P., The Science of Language (New York, Random House, 1962). Hunt, Kellog W., Differences in Grammatical Structures Written at Three Grade Levels, the Structures to be Analyzed by Transformational Methods, Report to the U.S. Office of Education, Cooperative Research Project No. 1998 (Tallahassee, Florida, 1964). Jespersen, Otto, Essentials of English Grammar (New York, Henry Holt and Company, 1933). (Reprinted University of Alabama Press, 1964). , Progress in Language (New York, Macmillan and Company, 1894). Longacre, Robert E., "String Constituent Analysis", Language, XXXVI (1960), 63-88. Pike, Kenneth L., "Grammemic Theory", General Linguistics, II (1957), 35-41. , Language in Relation to a Unified Theory of the Structure of Human Behavior, 3 vols. (Glendale, Summer Institute of Linguistics, 1954-1960). Shell, Olive, "Cashilio II", International Journal of American Linguistics, XXIII (1957), 179-218. Zandvoort, R. W., A Handbook of English Grammar, 9th ed. (Groningen, J. B. Wolters, 1964).

INDEX

Adams, Arthur, 197, 199 Andrew, S. O., 86, 208, 209, 210 Behm, O. P., 22, 24, 66, 152, 215 Brown, William H. Jr., 18, 21, 220 Brunner, Karl, 24 Bryant, Margaret M., 25 Carlton, Charles R., 18,19. 20, 35,154, 190,215, 217 Cassidy, Frederic G., 19, 22 Clark, Cecily, 16, 17,22, 6 4 , 6 6 , 8 4 , 1 0 9 , 1 5 1 , 1 5 2 , 163, 214, 219, 220 Earle, John, 17, 23 Elson, Benjamin, 23, 24, 27, 110, 214 Francis, W. Nelson, 23, 154 Fries, Charles C., 18, 20, 23, 68, 219 Garmonsway, G. N., 17 Gleason, Η. Α., 23, 204, 218 Harris, David Payne, 18 Harris, Zellig, 23 Hart, H., 27 Hill, Archibald, 23, 218 Hughes, John, 24, 204 Hunt, Kellog, 25 Jespersen, Otto, 15, 25, 35, 190 Ker, N. R., 16, 199, 207, 214 Kruisinga, Etsko, 15, 24

Magoun, F. P., 87 Malone, Kemp, 87 Marckwardt, Albert, 24 Mcintosh, Angus, 22, 191 Meyer, Heinrich, 22, 24, 66, 152, 215 Mitchell, Bruce, 22, 23, 94, 219, 220 Moore, Samuel, 24 Mossi, Fernand, 24 Nist, John, 217 Palmatier, Robert Α., 18, 21, 24, 109, 190, 220 Pickett, Velma, 23, 24, 27, 110, 214 Pike, Kenneth, 27, 28 Plummer, Charles, 17, 23, 86, 199, 207, 214 Quirk, Randolph, 176, 187 Roth, Wilhelm, 22 Rothstein, Ewald, 22, 84, 87, 219, 220 Saitz, Robert L., 19, 20, 35, 215 Shannon, Ann, 18, 20, 21, 35, 85, 215, 216, 217 Shell, Olive, 27 Small, George W., 15 Smith, C. Α., 15, 83, 87 Swieczkowski, Walerian, 24 Thomas, Russell, 19, 22, 169 Tucker, S. I., 17 Vallins, G. H „ 88

Longacre, Robert E., 24, 27, 28

Whitelock, Dorothy, 16, 17, 214 Wrenn, C. L., 176, 187

Magere, Mildred K., 20

Zandvoort, R. W., 25

JANUA LINGUARUM STUDIA MEMORIAE NICOLAI VAN WIJK DEDICATA Edited by C. H. van

Schooneveld

SERIES PRACTICA I. MARILYN CONWELL and ALPHONSE JUILLAND: Louisiana French Grammar, I: Phonology,

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 26. 27.

Morphology, and Syntax. 1963. 207 pp., 2 maps. Cloth. Gld. 50.— The Jewish Neo-Aramaic Dialects of Persian Azerbaijan: Linguistic Analysis and Folkloristic Texts. 1965. 342 pp., map. Cloth. Gld. 90.— MORRIS F. GOODMAN: A Comparative Study of Creole French Dialects. 1964. 143 pp., map. Gld. 29.— ROLAND HARWEG: Kompositum und Katalysationstext, vornehmlich im späten Sanskrit. 1964. 164 pp. Gld. 32.— GUSTAV HERDAN: The Structuralistic Approach to Chinese Grammar and Vocabulary: Two Essays. 1964. 56 pp., 4 figs. Gld. 23 — ALPHONSE JUILLAND: Dictionnaire Inverse de la Langue Frangaise. 1965. 564 pp., 9 figs. Cloth. Gld. 99 — A. HOOD ROBERTS: A Statistical Linguistic Analysis of American English. 1965. 437 pp., 11 figs., 6 tables. Cloth. Gld. 67.— VALDIS LEJNIEKS: Morphosyntax of the Homeric Greek Verb. 1964. 92 pp. Gld. 22.— ROBERT Ε. DIAMOND: The Diction of the Anglo-Saxon Metrical Psalms. 1963. 59 pp. Gld. 18.— JOSEPH E. GRIMES: Huichol Syntax. 1964. 105 pp. Gld. 25.— CLARA N. BUSH: Phonetic Variation and Acoustic Distinctive Features: A Study of Four General American Fricatives. 1964. 161 pp., 64 figs., 84 tables. Gld. 39.— WILLIAM E. CASTLE: The Effect of Selective Narrow-Band Filtering on the Perception of Certain English Vowels. 1964. 209 pp., 53 figs., 84 tables. Gld. 48.— ANN SHANNON: A Descriptive Syntax of the Parker Manuscript of the Anglo-Saxon Chronicle from 734-891. 1964. 68 pp. Gld. 18.— EicHi KOBAYASHI: The Verb Forms of the South English Legendary. 1964. 87 pp. Gld. 22.— HOMER L. FIRESTONE: Description and Classification of Sirionö, a Tupi Guarani Language. 1965. 70 pp., 7 figs. Gld. 21.— WOLF LESLAU: Ethiopian Argots. 1964. 65 pp. Gld. 21.— viDYA NIWAS MISRA: The Descriptive Technique of Pänini. 1966. 175 pp., some tables. Gld. 48.— EUGENE A. NIDA: A Synopsis of English Syntax. Second, revised edition. 1966. 174 pp. Gld. 25.— ROBERT τ. OLIPHANT : The Harley Latin-Old English Glossary, edited from British Museum, MS Harley, 3376. 1966. 223 pp. Gld. 48.— ERICA REINER: A Linguistic Analysis of Akkadian. 1966. 155 pp., graph. Gld. 36.— M. J. HARDMAN: Jaqaru: Outline of Phonological and Morphological Structure. 1966. 131 pp., 2 figs., map, 20 tables Gld. 30.— MARVIN κ. MAYERS (ed.): Languages of Guatemala. 1966. 318 pp. Gld. 48.— ROBERT LIVINGSTON ALLEN: The Verb System of Present-Day American English. 1966. 303 pp., 7 tables, 24 figs. Gld. 46.— ANDREW MCLEISH: The Middle English Subject Verb Cluster. 1969. 276 pp. Gld. 70.— EMMA GREGORES and JORGE A. SUAREZ: A Description of Colloquial Guarani. 1967.248 pp. Gld. 54.— IRENE GARBELL:

29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 50. 52. 53. 55. 57. 58. 59. 60. 61. 62. 63. 64. 65.

w. LAW: The Obligatory Constructions of Isthmus Nahuat Grammar. 1966. 73 pp., 21 tables Gld. 21.— MARVIN H. FOLSOM: The Syntax of Substantive and Nonfinite Satellites to the Finite Verb in German. 1966. 96 pp. Gld. 21.— DEBI PRASSANA PATTANAYAK: A Controlled Historical Reconstruction of Oriya, Assamese, Bengali and Hindi. 1966. 91 pp. Gld. 27 — ABDELGHANY A. KHALAFALLAH: A Descriptive Grammar of Sas:di Egyptian Colloquial Arabic. 1969. 124 pp. Gld. 36.— Η. Μ. ABOUL-FETOUH: A Morphological Study of Egyptian Colloquial Arabic. 1969.150 pp. Gld. 42.— SAUD Μ. GAMAL-ELDIN: A Syntactic Study of Egyptian Colloquial Arabic. 1967. 117 pp. Gld. 36.— H. MARCOS HANNA: The Phrase Structure of Egyptian Colloquial Arabic. 1967. 58 pp. Gld. 21.— IRMENGARD RAUCH: The Old High German Diphthongization: A Description of a Phonemic Change. 1967. 130 pp. Gld. 30.— JOSEPH HAROLD FRIEND: The Development of American Lexicography, 1798-1864. 1967. 129 pp., 4 facs. Gld. 32.— WILLIAM J. SAMARIN: A Grammar of Sango. 1967. 280 pp. Gld. 72.— DEAN Η. OBRECHT: Effects of the Second Formant on the Perception of Velarization Consonants in Arabic. 1968. 104 pp., 57 figs. Gld. 29.— YOLANDA LASTRA: Cochabamba Quechua Syntax. 1968. 104 pp. Gld. 29.— KAZUKO INOUE: A Study of Japanese Syntax. 1969. 160 pp. Gld. 36.— R. s. p. BEEKES: The Development of the Proto-Indo-European Laryngeals in Greek. 1969. xxiv + 324 pp. Gld. 82.— HARWOOD H. HESS: The Syntactic Structure of Mezquital Otomi. 1968. 159 pp. Gld. 40.— PAUL w. PILLSBURY: Descriptive Analysis of Discourse in Late West Saxon Texts. 1967. 91 pp. Gld. 20.— MADELINE ELIZABETH EHRMAN: The Meaning of the Modals in Present-Day American English. 1966. 106 pp. Gld. 21.— VIKTOR KRUPA: Morpheme and Word in Maori. 1966. 83 pp., 26 tables, 1 fig. Gld. 24.— JOHN c. FISHER: Linguistics in Remedial English. 1966. 71 pp., 4 tables Gld. 15.— Μ. Α. κ. HALLIDAY: Introduction and Grammar in British English. 1967.61 pp., 2 folding tables Gld. 18 — MARY RITCHIE KEY: Comparative Tacanan Phonology: with Cavinena Phonology and Notes on Pano-Tacanan Relationship. 1968. 107 pp. Gld. 32.— RUTH MARGARET BREND: A Tagmemic Analysis of Mexican Spanish Clauses. 1968. 128 pp. Gld. 28.— HAROLD H. KEY: Morphology of Cayuvava. 1967. 73 pp. Gld. 20.— L. ROMEO: The Economy of Diphthongization in Early Romance. 1968.127 pp. Gld. 24.— ALAN CAMPBELL WARES: A Comparative Study of Yuman Consonantism. 1968. 100 pp. Gld. 30.— JEAN PRANINSKAS: Trade Name Creation: Processes and Patterns. 1968. 115 pp. Gld. 24.— GEORGE GIACUMAKIS, JR.: The Akkadian of Alalab. 1970. 119 pp. Gld. 32.— JOAN RUBIN: National Bilingualism in Paraguay. 1968. 135 pp. Gld. 36.— SALMAN Η. AL-ANI: Arabic Phonology: An Acoustical and Physiological Investigation. 1970. 104 pp., 18 figs., 13 ill., 3 diag. Gld. 36.— CURTIS p. HEROLD : The Morphology of King Alfred's Translation of the Orosius. 1968. 80 pp. Gld. 20.— JAN SVARTVIK: On Voice in the English Verb. 1966. xiv + 200 pp. figs., and tables, Gld. 29.— MARVIN R. WILSON: Coptic Future Tenses: Syntactical Studies in Sahidic. 1970. 143 pp. Gld. 40.— RUSSELL N. CAMPBELL: Noun Substitutes in Modern Thai: A Study in Pronominality. 1969.70 pp. Gld. 21.— HOWARD

66.

A Descriptive Analysis of Cypriot Maronite Arabic: 1969. 69 pp. Gld. 20.— 70. BRENT BERLIN: Tzeltal Numeral Classifiers: A Study in Ethnographic Semantics. 1968. 243 pp., 118 plates Gld. 72.— 71. ROBERT D. STEVICK: Suprasegmentale, Meter, and the Manuscript of "Beowulf". 1968. 88 pp. Gld. 27.— 73. AERT H. KUIPERS: The Squamish Language: Grammar, Texts, Dictionary. 1967. 470 pp. map. Gld. 96.— AERT H. KUIPERS: The Squamish Language: Grammar, Texts, Dictionary. Vol. II. 1969. 98 pp., 1 photograph. Gld. 18.— 74. ROBERT ALLEN PALMATIER : A Descriptive Syntax of the "Ormulum". 1969·. 137 pp. Gld. 38.— 75. HELMUT R. PLANT: Syntaktische Studien zu den Monseer Fragmenten: Ein Beitrag zur Beschreibung der inneren Form des Althochdeutschen. 1969. 96 pp. Gld. 18.— 77. j. R. RAYFIELD: The Languages of a Bilingual Community. 1970. 118 pp. Gld. 28.— 79. HENRY G. SCHOGT: Le systdme verbal du fran^ais contemporain. 1968. 74 pp. Gld. 18.— 80. KAMIL ZVELEBIL: Comparative Dravidian Phonology. 1970. 202 pp. Gld. 54.— 81. DAVID COHEN: Iitudes de linguistique simitique et arabe. 1970. 178 pp. Gld. 54.— 82. OARY J. PARKER: Ayacucho Quechua Grammar and Dictionary. 1969. 221 pp. Gld. 64.— 83. JAMES w. GAIR: Colloquial Sinhalese Clause Structures. 1970. 164 pp. Gld. 40.— 85. ELMAR SEEBOLD: Vergleichendes und Etymologisches Wörterbuch der Germanischen Starken Verben. 1970. 571 pp. Gld. 132.— 86. GENEviivE N'DIAYE: Structure du dialecte basque de Maya. 1970. 249 pp., Folded table. Gld. 66.— 87. ANNA FUCHS: Morphologie des Verbs im Cahuilla. 1970. 76 pp. Gld. 35.— 93. MARIO SALTARELLI: A Phonology of Italian in a Generative Grammar. 1970. 96 pp. Gld. 24.— 94. VLADIMIR MILTNER: Theory of Hindi Syntax: Descriptive, Generative, Transformational. 1970. 72 pp. Gld. 28.— 96. J0ZEF TOMPA: Ungarische Grammatik. 1968. 426 pp. Gld. 72.— 97. RODOLFO JACOBSON: The London Dialect of the Late Fourteenth Century: A Transformational Analysis in Historical Linguistics. 1970. 193 pp. Gld. 48.— 100. Q. ι. M. MOK: Contribution ä l'itude des catigories morphologiques du genre et du nombre dans le francais parte actuel. 1968. 155 pp. Gld. 30.— 101. WILLIAM H. BROWN, JR.: A Syntax of King Alfred's Pastoral Care. 1970. 91 pp. Gld. 18.— 102. GARY DEAN PRIDEAUX: The Syntax of Japanese Honorifics. 1970. 107 pp. Gld. 28.— 105. RUDOLF P. BOTHA: The Methodological Status of Grammatical Argumentation. 1970. 70 pp. Gld. 11.— 106. ANDRE-MARCEL D'ANS: Le Cr6ole fran^ais d'Ha'iti: Etude des unites d'articulation, d'expansion et de communication. 1968. 181 pp. Gld. 50.— 111. CHARLES CARLTON: Descriptive Syntax of the Old English Charters. 1970. 200 pp. Gld. 48.— 112. MICHAEL GRADY: Syntax and Semantics of the English Verb Phrase. 1970. 84 pp. Gld. 14.— 113. ALEXANDER ISACENKO and HANS-JOACHIM SCHÄDLICH: A Model of Standard German Intonation. 1970. 66 pp., folded plate, record. Gld. 21.— 116. SAMUELN, ROSENBERG : Modern French Ce: The Neuter Pronoun in Adjectival Predication. 1970. 222 pp. Gld. 36.— MARIA TSIAPERA:

MOUTON · PUBLISHERS

THE

HAGUE