The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe 9789463721929, 9789048550166

Citation preview

KNOWLEDGE COMMUNITIES

Mark Amsler

The Medieval Life of Language Grammar and Pragmatics from Bacon to Kempe

The Medieval Life of Language

Knowledge Communities The Knowledge Communities series focuses on innovative scholarship in the areas of intellectual history and the history of ideas, particularly as they relate to the communication of knowledge within and among diverse scholarly, literary, religious and social communities across Western Europe. Interdisciplinary in nature, the series especially encourages new methodological outlooks that draw on the disciplines of philosophy, theology, musicology, anthropology, paleography and codicology. Knowledge Communities addresses the myriad ways in which knowledge was expressed and inculcated, not only focusing upon scholarly texts from the period, but also emphasizing the importance of emotions, ritual, performance, images and gestures as modalities that communicate and acculturate ideas. Knowledge Communities publishes cutting-edge work that explores the nexus between ideas, communities and individuals in medieval and early modern Europe. Series Editors Clare Monagle, Macquarie University Mette Bruun, University of Copenhagen Babette Hellemans, University of Groningen Severin Kitanov, Salem State University Alex Novikoff, Fordham University Willemien Otten, University of Chicago Divinity School

The Medieval Life of Language Grammar and Pragmatics from Bacon to Kempe

Mark Amsler

Amsterdam University Press

Cover image: From BL, Stowe MS 49, fol. 220r (c. 1300) © The British Library Board Cover design: Coördesign, Leiden Typesetting: Crius Group, Hulshout isbn 978 94 6372 192 9 e-isbn 978 90 4855 016 6 (pdf) doi 10.5117/9789463721929 nur 613 ǀ 616 © Mark Amsler / Amsterdam University Press B.V., Amsterdam 2021 All rights reserved. Without limiting the rights under copyright reserved above, no part of this book may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the written permission of both the copyright owner and the author of the book. Every effort has been made to obtain permission to use all copyrighted illustrations reproduced in this book. Nonetheless, whosoever believes to have rights to this material is advised to contact the publisher.



Table of Contents

Acknowledgements

7

Abbreviations

9

Introduction: Where is Medieval Pragmatics?

11

1 Medieval Pragmatics: Philosophical and Grammatical Contexts Three Terms and a Theory Roger Bacon’s Semiotics and Pragmatics Peter (of) John Olivi: Pragmatics and the Will to Speak

31 32 43 72

2 Interjections: Does Affect have Grammar?

85

3 Allas Context Allas: A Case for Context

103 109

4 Alisoun’s Giggle, or the Miller Does Pragmatics Does a Giggle Mean? Impoliteness, Hedging, and Textual Pragmatics Polysemy, Bullseyes, Misfires, or How Narrative Escapes Intention Centrifugal Narrative Contracts

139 143 147 156 161

5 How Heretics Talk, According to Bernard Gui and William Thorpe Pragmatic Talk, Pragmatic Action Bernard Gui’s Conversation Analysis and Institutional Discourse William Thorpe’s Relationship Pragmatics

165 167 170 183

6 Margery Kempe’s Strategic Vague Language Cooperate or Else Vaguing Pragmatics Kempe Comes to the Archbishop Kempe Tells a Tale

205 206 211 212 234

One More Thing

241

Bibliography

251

Index

263

List of Images and Tables Images Image 6.1 Oak pulpit Image from St. James Church, Castleton Acre, Norfolk. Available at: http://www.norfolkchurches.co.uk/ castleacre/images/dscf0260.jpg. Accessed 1 May 2020 Image 6.2 Luther, Passional Christi und Antichristi (Wittenberg, 1521), pp. 18-19 Woodcuts by Lucas Cranach the Elder. Internet Archive. Available at: https://archive.org/details/ passionalchristi00luth/page/n17/mode/2up. Accessed 1 May 2020 Tables Table 3.1 Table 3.2

233

233

Allas in Chaucer’s Knight’s Tale117 Allas in Chaucer’s Miller’s Tale130

Acknowledgements For help making this book, I especially thank: Anneli Luhtala (Helsinki), for suggestions about Boethius’ Latin and for her generous intellect over the years with our mutual interest in the history of linguistics; colleagues connected with the Australia and New Zealand Association for Medieval and Early Modern Studies (ANZAMEMS), whose intellectual hospitality makes medieval studies smart, fun, and edgy, especially Jenna Mead, Constant Mews, Juanita Feros Ruys, Andrew Brown, Amanda McVitty, and Stephanie Trigg; Elizabeth Robertson (Glasgow), for her friendship and encouragement, her intellectual challenges and comradeship, and her new ways to explore medieval ideas about the senses and pragmatics; colleagues at the University of Leuven (Belgium): Pierre Swiggers, for arranging a visiting fellowship at the Center for the Historiography of Linguistics at the University of Leuven, which enabled me to complete the manuscript; Pierre, Toon Van Hal, and Alfons Wouters, for many years of fresh ideas and stimulating talk about language and for their generous high school exam supervision when asked; Stephen Turner (Auckland and Amsterdam), for talking text and sending Nicholas of Paris’ Syncategoreumata at a crucial moment; Bernadette Luciano (Auckland), for explaining Italian interjections; librarians at Erasmushuis Artes Library (University of Leuven) and the Rare Books Collection at McGill University Library (Montreal), for their generous help accessing their collections; Faculty of Arts (University of Auckland), for research leaves and material and intellectual support; University Auckland Libraries staff for keeping the collection vital in sometimes difficult circumstances; Shannon Cunningham, Amsterdam University Press, for her encouragement and editorial advice on this project. Portions of Chapter 2 originally appeared in Before Emotion: The Language of Feeling, 400-1800, ed. Juanita Feros Ruys, Michael W. Champion, and Kirk Essary (New York and London: Routledge, 2019), 26-37. Republished by permission of Taylor and Francis Group LLC.

Abbreviations CCCM Corpus Christianorum, continuatio medievalis CIMAGL Cahiers de l’Institut du Moyen Age Grec et Latin CMEPV Corpus of Middle English Prose and Verse CT Chaucer, Canterbury Tales CTMPT The Cambridge Translations of Medieval Philosophical Texts: Volume 1: Logic and the Philosophy of Language. Ed. Norman Kretzmann and Eleonore Stump Early English Text Society EETS Heresies Heresies of the High Middle Ages. Trans. Walter W. Wakefield and Austin P. Evans Grammatici latini. Ed. Heinrich Keil KGL LLT-A/B Library of Latin Texts: Series A/B Middle English ME Middle English Dictionary MED Old English OE Oxford English Dictionary OED Patrologia cursus completus. Series Latina. Ed. JacquesPL Paul Migne Stanford Encyclopedia of Philosophy. Online: https:// SEP plato.stanford.edu/archive TC Chaucer, Troilus and Criseyde WB/F-M The Holy Bible, Containing the Old and New Testaments, with the Apocryphal Books, in the Earliest English Versions Made from the Latin by John Wycliffe and His Followers. Ed. Josiah Forshall and Frederic Madden. 4 vols.



Introduction: Where is Medieval Pragmatics? Abstract This book recovers pragmatics within the history of medieval linguistics. The introduction outlines the study of pragmatics from a critical history of linguistics perspective, situating language study in a complex social field and comparing medieval pragmatic ideas and metapragmatics with assumptions in contemporary pragmatic theory. Pragmatics embraces communication, expression, and understanding; it prioritizes meaning, context, affect, and speaking position over formal grammar. Relevant texts for late medieval pragmatics include grammatical and logical texts, especially those by Roger Bacon, Robert Kilwardby, and anonymous grammarians, and Peter (of) John Olivi. Other sources for medieval pragmatics include life narrative (Margery Kempe), poetry (Chaucer), and heresy records. Theoretical and everyday texts reveal provocative intersections of Latin and vernacular intellectual and religious cultures and different assumptions and ideologies concerning meaning, speech, and speakers. Across these heterogenous, sometimes antagonistic discursive fields, medieval intellectual history crosses paths with social history. Keywords: critical history of linguistics, medieval pragmatics, pragmatics and semiotics, discourse theory

The history of linguistics and language study is a branch of intellectual history and an adjunct to the history of ideas. In modern times, histories of linguistics have usually adopted one of three approaches: 1 history of the development of “Linguistics” the discipline, as a sequence of Great Texts or Key Ideas from pre-Linguistics to present-day Linguistics (internal History A), or history of one or more developments within Linguistics the discipline, for instance, individual theories and terminology or a subfield or disciplinary paradigm (phonology, syntax,

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_intro

12 

The Medieval Life of L anguage

comparative-historical linguistics, transformational grammar, pragmatics, sociolinguistics, L2 acquisition, etc.) (internal History B); 2 history of the relations between ideas about language and language study, not restricted to Linguistics the discipline, and other intellectual, political, and social ideas, sometimes incorporating a critical perspective on language ideologies, epistemes, paradigms, or dominant and minority discourses which determine what is “linguistic” (the stuff) or “linguistics” (the subject) (intellectual History); 3 documentary history presenting the development of Linguistics the discipline with anthologies of ‘classic’ texts, interviews, memoires, or summaries (docu-History). Most histories of linguistics take account of grammar, philosophy of language, historical materials, and the like. Few, however, expand the textual archive to include literature or primary documents in social history. Pragmatics has emerged as an important and productive subfield in Linguistics the discipline, but among historians of linguistics, pragmatics has been underrecognized. In this book I discuss pragmatics as an important aspect of premodern understanding of language and meaning. I focus on pragmatic ideas and metapragmatic thinking in theory and practice in late medieval Europe, especially in England and France. My general approach is aligned with critical intellectual History. I explore relations and influences between late medieval linguistic ideas of grammar and pragmatics and other social and institutional contexts and practices. My argument is that far from being marginal or underdeveloped, pragmatic ideas and metapragmatic awareness were important aspects of medieval thinking about language and communication in grammatical, philosophical, and religious discourses. This critical historical analysis of medieval linguistics is informed by Foucauldian knowledge/power critique, Critical Discourse Analysis (CDA), Conversation Analysis (CA), and close textual reading to explore pragmatic implications of intellectual discourse and textual representations. (For overviews of CA and CDA research and theory, see Van Dijk 1993; Wooffitt 2005; Wetherell 2012; Fairclough 2014. Cf. Foucault 1980, 1997; Bourdieu 1990.) Analyzing ideational and textual discourse from CA and CDA perspectives gives us sharper understandings of how “structuring structures” of knowledge, institutions, and power, the ‘giveness’ or habitus of thought behavior, supported medieval pragmatics and metapragmatic ideas. We will also see how explicit and implicit pragmatic ideas nourished medieval people’s strategies for dissent and counter or alternate hegemonic discourse, manifested in texts representing both spoken and written discourse.

Introduc tion: Where is Medieval Pr agmatics?

13

Between 1100 and 1450, pragmatic ideas and metapragmatic thinking were discussed and debated not only in Latin by scholars and other intellectuals, but also in some vernaculars, in particular English and French. Vernacular contexts for discussing pragmatic ideas and metapragmatic thinking included not only scholars but also lay people engaged in social and intellectual interactions outside education, usually in literature or in relation to religious authority (cf. Gramsci 1971: 5-23; Copeland 2001). Language use always implies pragmatic practice, but what I am referring to here are medieval pragmatic ideas and metapragmatic awareness as discussed, exploited, and revised by clergy and lay people, poets, devout religious, dissenters, and heresy hunters. Medieval pragmatics was part of intellectual and socio-cultural work on several levels, within different social groups, and often in contested contexts. Medieval pragmatic thinking was often at the center of strategies for securing discursive control or power or performing agency. From a critical intellectual History perspective, we discover late medieval pragmatics theory and practice to be constituted as a multilingual and multilevel field of individual, social, and institutional discourses. Some actors embody established intellectual subjects (ars grammatica, dialectica, rhetorica), while others participate in elite or official intellectual culture as it was appropriated for everyday or more local knowledge and experience. Some deploy medieval pragmatic discourse by referring explicitly to intellectual and theoretical concepts (conceptus, affectus, intentio, equivocatio, sinceritas, suppositio), while others work more intuitively or experientially with less disciplinary vocabulary, but with no less pragmatic sophistication in order to control, explain, or critique meaning-making. They interact pragmatically and strategically, express themselves as discursive subjects, or dissent from dominant institutional authority. The objects in a discursive field or field of study are constructed by the metalanguage, strategies, and assumptions we use to organize and evaluate our talk about those objects. We construct the objects, give them names, then talk about the names and descriptions of those objects. Our languages about the physical world or visual art are complex; they are constitutive of what we know about the world or the art. For other subjects – I’m thinking of the history of linguistics, linguistics, literary studies, rhetoric, philosophy – the objects of study are primarily verbal. In those cases, the relations between metalanguage and object language are complex but in a different way. Computational and analphabetic systems of representation have been devised to represent language, but for the most part our ideas and theories about language have been articulated with language-based discourses which draw from the very objects being investigated and remake those words to serve metalinguistic,

14 

The Medieval Life of L anguage

conceptual, or critical purposes. This imbricated verbal condition is important for both the subject of this book – pragmatics – and how we talk about it. Not all languages, however, produce or render up a metalanguage to describe them. The accepted language for any grammatical or linguistic description and evaluation occupies a privileged position within a cultural or intellectual context. Metalanguage has power, power to claim to know and power to promote, contain, or exclude what we claim to know. Like other forms of power, the power of metalanguage is asymmetrical. In modern and contemporary linguistics, European languages, especially English, French, and German, have provided the majority of metalanguage and discursive modes for language analysis regardless of the linguistic theory or perspective. In the European Middle Ages, Latin was the dominant language for intellectual discourse, the liberal arts, and grammatica. After 900 CE some grammatical writings appeared in a vernacular (Old Icelandic, Old Irish, Old French, Old and Middle English, Anglo-Norman [aka French of England], Welsh), but much of medieval linguistic work was composed in Latin and about Latin as a manifestation of Langage. In the early Middle Ages European scholars, almost all affiliated with the Church, adapted the Roman grammarians’ models and discourse to discuss language. After about 1050 CE grammarians and philosophers began to construct a different sort of grammatical discourse, again mostly in Latin and about Latin as Langage. Some grammarians rethought medieval logic’s vocabulary and concepts and redirected that discourse on propositions to more semiotic and ultimately pragmatic understanding. Their rereading of propositional logic in a semiotic framework informs a good deal of late medieval pragmatic thinking. The terms linguistics and pragmatics were not used in the Middle Ages, but the ideas as well as inherent linguistic practices were. Intellectually, pragmatics, pragmatic ideas, and metapragmatic awareness were part of medieval sign theory and philosophy of language. Interpersonally and linguistically, they were active and critical parts of people’s strategic use of language in everyday and elite spoken and written situations. Both spoken and written contexts are relevant for a critical history of linguistics. After 1100 CE pragmatic topics sparked the interest and challenged received linguistic ideas of grammarians and philosophers. Not only were pragmatic ideas and metapragmatic thinking foregrounded in some medieval grammatical and semiotic theory. They were also deployed strategically and reflected on in everyday, official, and poetic written discourse. Pragmatic ideas and metapragmatic awareness were often foregrounded in conflicts over religious orthodoxy, conformity, and dissent. As we will see, medieval pragmatics manifests the relations between medieval ideas about language and medieval social and religious identities.

Introduc tion: Where is Medieval Pr agmatics?

15

By now, some readers are wondering what exactly do I mean by pragmatic ideas and metapragmatic awareness? In this book I use the term pragmatics to designate: 1) the study of language from the perspective of users and as it is used in practice and 2) the specific purposes and ways in which language users make meaning, understand expressions, create discursive topics, interact, and position themselves politically and socially. Functionally, the pragmatic field is comprised of text (what is said/written), position/stance (speaker/ writer’s intent, attitude to content), and interaction (who is addressed). Pragmatics embraces communication, expression, and understanding; pragmatics prioritizes meaning, context, and speaking position (not necessarily in that order) over grammar. The theory of pragmatics and metapragmatic awareness foregrounds our implicit communicative repertoires and tools for understanding the relations between what is intended, what is said, and what is understood. Consequently, our conventional pragmatic repertoires moderate the interactions and relations between speakers and listeners. In different texts and contexts, the speaker’s discursive position and utterance distinguishes what are pragmatic theory, pragmatic practice, and metapragmatic thinking. Pragmatics as a functional perspective aligns with semiotics and sociolinguistics and thus differs from more structural or propositional perspectives on language which align with grammatical and linguistic formalism. Pragmatics focuses on how speakers use or express linguistic, paralinguistic (gestures), and nonlinguistic (grunts, sighs) signs to make meaning, communicate, and interact in different social circumstances. In addition, metapragmatic awareness and critical reflection call into question naive notions of ‘intention’ with respect to what is said and what is meant. The inferred ‘intended meaning’ and the “illocutionary purpose” for which an utterance is articulated are constructed or contested in social interaction, but not by the individual speaker alone. The history of linguistics depends largely on textual materials. Evidence for how verbal subjects and behaviors in the past were understood and represented is almost entirely textual. The available evidence and materials for medieval pragmatic theory, pragmatic strategies, and metapragmatic awareness are no different (although some visual art evidence is relevant) (cf. Jucker 2000). Relevant texts for the history of late medieval pragmatics include grammatical and logical treatises and commentaries, especially those by Roger Bacon, Robert Kilwardby, the grammarians known as PseudoKilwardby and Magister Johannes, and Peter John Olivi. But pragmatics as part of medieval culture exceeds a definition of ‘intellectual’ restricted to elite or dominant culture. Gramsci makes the point directly: “although one

16 

The Medieval Life of L anguage

can speak of intellectuals, one cannot speak of non-intellectuals because non-intellectuals do not exist” (1971: 9). The critical archive for the history of medieval pragmatics also includes chronicles of institutional power and surveillance, life-writing by men, women, and religious dissenters, letters, literature, and translation theory and practice. Grammars and treatises on the philosophy of language are familiar materials for the history of linguistics. But other kinds of texts are just as relevant to histories of linguistics when we are searching records of people’s social and linguistic understanding and participation in social life. Medieval grammars and logic texts get special attention from historians because they ‘look’ like the subject Linguistics as determined by how the field has been defined in modern times. For the history of medieval pragmatics, however, we need to look beyond the explicitly grammatical and philosophical archive and attend to texts and contexts belonging to other discourses and cultural domains. A social or critical intellectual history of linguistics is much like the history of a language. Relevant evidence and materials include not only grammars, textbooks, and official discourse but also evidence for everyday use, language attitudes, metacommentary, social or linguistic controversies, and representations of linguistic ideas or use. Some poetic, institutional, and everyday texts relevant to medieval pragmatics reveal provocative intersections of Latin and vernacular intellectual and religious cultures and sometimes show speakers’ different assumptions and ideologies about meaning and speech and the people who are speaking. In this study of medieval pragmatics, I read across these discursive fields, where medieval intellectual history crosses paths with social history. We encounter some striking and sophisticated linguistic and pragmatic thinking about meaning-making and the value of ambiguity, the trouble with vagueness, and struggles for agency, even survival, in contested circumstances. As a perspective on the history of medieval linguistics, my approach to the study of post-1050 pragmatics is expansive and critical, in line with what is sometimes referred to as the Continental or European approach to pragmatics and historical pragmatics (e.g., Traugott 2006). The broadly cultural Continental approach is often opposed, theoretically and historically, to a narrower Anglo-American perspective founded on speech act theory (cf. Levinson 1983: 2, 4-5; Verscheuren 1999: 1; Mey 2001: 3-35; Taavitsainen and Jucker 2010). The Continental approach emphasizes utterance and discourse and draws its methodology from propositional analysis, sociolinguistics, discourse analysis, and semiotics. The so-called Anglo-American approach typically restricts itself to the propositional analysis of what is said and what is meant from the speaker’s point of view, the anchors of speech act theory

Introduc tion: Where is Medieval Pr agmatics?

17

and analysis. For the Continental approach and in this study, speech act theory and analysis are one part but not the only or most important part of pragmatic understanding, both historically and theoretically. Uptake or the consequences of speech and the ideological constructions of speech and performance contexts are also crucial for pragmatic understanding. A socio-semiotic approach to pragmatics theory and pragmatic practices, including text analysis, enables us to better explore medieval pragmatic practices and theories both for their alterity and as a reusable past. Pragmatics matters for the history of medieval linguistics and language thought because pragmatics focuses our attention on the particularity and constitutive power of context. Pragmatic theoretical analysis interrogates the conditions for communication and the ways meaning can be communicated or understood even when meaning does not map directly onto the utterance’s surface form. Pragmatics also investigates how speech or interactions can be and are understood differently by speaker and listener depending on each participant’s presuppositions, goals, and contextual or institutional constructions for situating speech and action. Shared understandings make communication possible, but not all interactions maintain shared understanding. Pragmatics is first and foremost language in use, but pragmatics is also a perspective on language, a way of thinking about language as social activity and as subject formation and positioning. Pragmatics is intimately involved with broader issues of agency, cooperation, and struggles for discursive power. Pragmatics and metapragmatic awareness perform and foreground how linguistic forms and structure are used and contextualized in order to make, control, or change meaning for different speakers’ purposes. Contemporary pragmatics investigates meaning and communication in several ways, especially with respect to implicitness, speech acts, deixis, face (politeness, impoliteness), constitutive contexts, social consequences, and nonverbal interaction. Pragmatics involves social constructions of discursive space. As we shall see, medieval pragmatics theory and understanding considered such questions, but often deployed a different metalanguage and addressed these topics with different, sometimes conflicting assumptions, not all of them ‘modern.’ Pragmatic theory and practice have been largely omitted from broad surveys of the history of linguistics. Among major histories (internal History A), R. H. Robins’ standard survey (1967/1997) mentions pragmatics just once, in the final chapter in a paragraph devoted to the post-1950 interest in semantics. Vivien Law’s The History of Linguistics in Europe: From Plato to 1600 (2003) stops at Early Modern linguistics and doesn’t mention pragmatics at all. Nor does Holger Pedersen’s much earlier and influential Linguistic Science

18 

The Medieval Life of L anguage

in the Nineteenth Century (1924 in Danish, 1931 in English); the Neogrammarians’ theory did not account for extraverbal contexts. Hans Aarsleff (1982) mentions syntax, etymology, semantics, and comparative philology as topics in the history of linguistics but not pragmatics. Julia Kristeva in Le langage, cet inconnu (1969/1981) adopts a structural-semiotic approach to the history of linguistic ideas. She devotes a chapter to “La pratique du langage” (1981: 275-291), but mostly discusses rhetorical speech (oratoire) and literary writing. Unlike Robins and Kristeva, Margaret Thomas’ Fifty Key Thinkers on Language and Linguistics (2011), a succinct version of internal History A, reflects the modern orientation to pragmatics. Thomas includes entries for philosophers and linguists who have made important contributions to Anglo-American pragmatics (Grice, Searle, McCawley), but pragmatics and medieval linguistics do not intersect in her episodic survey. Anthologies of key linguistics texts and oral self-histories (docu-History) from North American linguists collected in the First-person Singular volumes (1980, 1991, 1998) likewise maintain a strong disciplinary focus. If we believe those linguists’ self-accounts, twentieth-century linguistic intellectual work had little or nothing to do with pragmatics. That’s definitely not the case for European linguists. For most broad surveys of Linguistics the discipline, pragmatics belongs to whatever is called ‘modern’ linguistics. In the history of linguistics, intellectual History surveys blur the line between Linguistics the discipline and general intellectual or cultural history. The historiographic texts by Chomsky (3rd ed. 2009, first published 1966), Kristeva (1981), and Andresen (1990, 2013) are ‘critical’ to the extent they connect linguistic theories, ideas, and practice with philosophical traditions and nationalist, ideological, and institutional formations, but they offer varying degrees of persuasion and evidence. Many intellectual historians of linguistics are interested in precursors as well as ruptures in knowledge frameworks (for example, Padley 1976, Kristeva 1969/1981, Aarsleff 1982, and Swiggers 1997). But again, these surveys say little or nothing about theories of pragmatics. They take the subject to be a modern component of the discipline but not a genuine part of linguistics’ past. Some historians of linguistics, however, have recognized that pragmatics has a past. Their work and the significance they attach to pragmatics for the development of linguistic ideas challenge us to rethink the history of linguistics. I’m referring especially to Brigitte Nerlich (2006), Andreas Jucker (2012), Jacob Mey (2013), and in more depth, Nerlich and D. D. Clark (1996) and Irène Rosier (1994, 2004, 2010, 2016). Although most of these studies remain within a disciplinary framework for linguistics and pragmatics, they don’t always restrict themselves or the subject to a presentist focus. In

Introduc tion: Where is Medieval Pr agmatics?

19

particular, Rosier’s work and the contributors to the special issue of Vivarium (2011) devoted to medieval pragmatics make important contributions to our historical and intellectual understanding of medieval pragmatics and medieval linguistics. But there is more to pragmatics history when we look beyond Linguistics the discipline and philosophy to other social discourses and contexts. Mey and Jucker construct a post-1800 history of pragmatics in order to show how pragmatics has become an important subfield in contemporary linguistics. Jucker has been one of the most prolific theorizers and promoters of history of pragmatics and historical pragmatics research. But he maintains a pretty hard distinction between the history of linguistics and historical pragmatics research. His historical pragmatics work has focused on late medieval and early modern pragmatic uses and strategies for politeness, pronouns of address, and so forth in literary and social texts. But when Jucker turns to the history of linguistics, he adopts a more linear disciplinary perspective. In his survey “Pragmatics in the history of linguistic thought” (2012) he states: “I shall take a broad view on both these issues by including not only pragmatics avant la lettre but also a brief and necessarily selective account of the development of the discipline itself by adopting a broad, basically Continental [European] view of pragmatics …” (2012: 496). But Jucker’s chapter focuses almost entirely on disciplinary linguistics from 1830 to the present, despite his advocacy of the Continental approach to pragmatics. When he says that pragmatics “is still a relatively young branch of linguistics” (2012: 495), his presentist disciplinary assumption means that he effectively rules out medieval pragmatics. In his sketch survey, avant la lettre becomes a very short period of time indeed. Jucker briefly notes the relation of pragmatics the field to semiotic theory, in particular, the work of C. S. Peirce (1839-1914) and Charles Morris (1901-1979), and he notes pragmatic thinking before the late nineteenth century with a few sentences about classical rhetoric. However, despite his interest in the “broad … Continental view” of pragmatics, Jucker restricts his brief history to pragmatics’ role in the construction of the modern discipline of Linguistics and how pragmatics is part of a paradigm shift from formal to functional linguistics (2012: 503-505). Jacob Mey (2013) similarly elides medieval pragmatic theory and ideas and metapragmatic awareness in his sketch of the development of pragmatics. Like Jucker, Mey has been a strong advocate of the Continental approach to pragmatics. At first, his distinctions among “Early Origins of Pragmatic Thinking,” “Recent Pragmatic Ancestry,” and “Modern Approaches” hint at a more complex historical view of what the field of pragmatics includes and

20 

The Medieval Life of L anguage

has included. Mey helpfully discusses what he calls pragmatic thinking, not just the discipline of pragmatics. Nonetheless, his brief narrative is short on historical and critical analysis. Using terms like “origins” and “approaches,” Mey emphasizes the longevity of the “pragmatic turn,” but he keeps to the disciplinary frame with respect to the history of linguistics. He restricts his historical perspective on linguistics to intellectual discourse as purveyed by contemporary linguistics and philosophical theory (Austin, Grice, Searle), foundational for much contemporary pragmatics and semantics. While Mey suggestively links ancient philosophy, rhetoric, and linguistics with a pragmatics thread, he discusses the Sophists only briefly and medieval philosophy barely at all. He acknowledges the relation of pragmatics to semiotics but dismisses Peirce’s “rudimentary pragmatic insights” outright (2013: 589), apparently because he believes Peirce’s semiotic pragmatics is fundamentally different from linguistic pragmatics. (It’s not, or at least needn’t be.) Mey recognizes a few pre-1900 developments and an historical perspective which looks beyond the contemporary discipline of linguistics, but he ends up partly reiterating the Anglo-American approach to pragmatics by calling contemporary speech act theory the core of ‘proper’ pragmatics. This move is curious given Mey’s critique elsewhere of the limitations of speech act theory for pragmatics theory (e.g., 2001: 104-105). Mey implies that with every pre-Frege, pre-Wittgenstein, pre-Austin, pre-founding father of pragmatic thinking, we inch closer to the sunlit present of understanding, glimpsing but not yet fully seeing the “modern pragmatic truth about language,” a truth revealed only in the contemporary discipline of pragmatics, that there is no language prior to utterance and no meaning prior to use (2013: 589). Nerlich and Clark’s book-length study (1996) is the most generous of the historical accounts of post-1700 pragmatics. Their expansive historical view of language study, linguistics, and pragmatics is a kind of antidote to Pedersen’s triumphalist account of the Neo-grammarians. Nerlich and Clark reconnect modern pragmatics with Kant (1724-1804) and early twentieth-century philosophies of subjectivity, meaning, language, and truth. They stress the theoretical and disciplinary significance of Kant’s notion of pragmatisch and his distinction between practical reason (the “pure” kind) and pragmatic reason (the “empirical” kind). Their historical arc positions Kant as an important progenitor of modern pragmatics and rebuts Searle’s off-the-cuff dismissal of Kant as having any place in a genealogy of pragmatics: “You can’t go and find Kant’s view on apologizing or congratulating, as far as I know” (1984: 25). Still, Nerlich and Clark, like Mey and Jucker, manifest a modernist and disciplinary approach to pragmatics theory and history of

Introduc tion: Where is Medieval Pr agmatics?

21

linguistics (internal History B). When they take note of pre-1800 pragmatic ideas or theory (e.g., Aristotle, the Stoics, or Kant), they represent those insights as intimations of future adequacy rather than as fully accredited linguistic pragmatics or metapragmatic understanding within a specific sociocultural and intellectual context. Most historians of medieval linguistics and philosophy adopt internal History A/B or intellectual History approaches. Julie Andresen (1985) has identified the comparison of theories approach as one of the principal strategies for presenting historicized accounts of linguistics. In histories of medieval linguistics (internal or intellectual History) the comparison of theories approach is embodied in narratives of the shifts from early medieval and Carolingian word-based grammar to post-1100 Priscian- and dialectic-influenced or syntax-based grammar to Humanist rhetoric and the recovery of classical models of Latin expression. The comparison of theories approach also shapes the continued interest in the influence of the grammatica speculativa on other theories of grammar in the late Middle Ages and on the description of pre-speculative and post-speculative periods (e.g., Kneepkens 1983, 1990, 2013; Studies in Medieval Linguistic Thought, ed. Koerner, Niederehe, and Robins [1980]). Despite these limitations, some scholars who adopt intellectual History approaches argue the importance of the role of pragmatics in medieval intellectual history and ideas about language. The history of medieval pragmatics challenges those received versions of medieval linguistics which emphasize grammatical theory, pedagogy, and philosophical commentary. Some historians of medieval philosophy and intellectual culture have explored the relation of linguistics to philosophy and pragmatics to semantics (notably, Gabriel Nuchelmans [1988] and the contributions to the Vivarium 2011 medieval pragmatics issue). Nonetheless, some of these scholars, though not Nuchelmans, describe pragmatic signif ication or meaning not as constitutive, but as an “intrusion” into the domain of proper propositional or logical form. That is, they regard pragmatics as a separate domain from logic or proper propositional form (e.g., Mora-Márquez 2011; cf. Passnau and Toivanen 2018). In the history of medieval linguistics, sign theory holds a special place for the scholastics, not only those associated with the grammatica speculativa but many who oppose them as well. Aristotle and Augustine’s theories of signs were formative for several not entirely compatible strands of linguistic and semiotic thinking in the schools. Exploring the significance of sign theory in medieval theories of language and mind, the collections of essays in Umberto Eco and Costantino Marmo (1989) and Eugene Vance and Lucie

22 

The Medieval Life of L anguage

Brind-Amour (1983) and Marcia Colish’s influential Mirror of Language (1968/1983) brought together semiotics and cultural criticism, although not necessarily with a pragmatics focus. This research set new directions for later scholars and historians of linguistics. Some scholars directly connect medieval sign theory with pragmatics, in particular, Sten Ebbesen (e.g., 1977, 1980) and Rosier (1994, 2004, 2010). They have argued for the importance of pragmatics for more fully understanding medieval linguistics, semiotics, and semantics. Focusing on scholastic methodology and the apparatus of speech analysis, Nuchelmans (1988) and Louis Kelly (2002) have worked out some of the complex relations between scholastic grammar, logic, pragmatics, and theories of actus. Some historians of medieval philosophy of language situate pragmatics within the scope of medieval ideas about true and false propositions, syllogisms, and criteria for formal analysis. However, Rosier takes a different approach based on the thirteenth-century shift toward Aristotelian theories of language and intention. She specifically argues that medieval pragmatic ideas belong to the history of medieval linguistics. Rosier’s influential work sets post-1050 grammar broadly in relation to philosophical and theological theorizing. Her studies of syntax and pragmatics rely on important archival research and bring new manuscript evidence to the study of medieval linguistics. Rosier’s La parole comme acte: Sur la grammaire et la sémantique au xiiiesiècle (1994) sets pragmatics within grammatical theory. In that work and elsewhere (Rosier 1983, 1996, 2004, 2010, 2018) she draws on philosophical and theological discourses to argue that medieval ‘intentionalist’ grammarians and their pragmatic and semiotic orientation to language and usage have profound implications for ideas and attitudes about language in the thirteenth and early fourteenth centuries. Rosier’s groundbreaking research does much to fill in the theoretical gap between the Stoics and Kant that we find in other histories of pragmatics. However, Rosier relies primarily on grammatical and philosophical texts, which shows some of the limitations of the comparison of theories approach to the history of linguistics (internal History A/B, intellectual History). An intellectual history of linguistics, however broadly conceived, often remains at the ideational level of grammars, treatises, and similar high theory texts. This is not unlike the “history of ideas” approach to intellectual history associated with Lovejoy (1948), which rests on the assumption that “unit ideas … which have long life-histories of their own” (1948: 9) can be recognized, described, and analyzed in different historical and cultural contexts (science, philosophy, literature, education, etc.) as part of intellectual history. The comparison of theories approach to the history

Introduc tion: Where is Medieval Pr agmatics?

23

of linguistics is significant in that it can clarify how a particular theory of language constructs objects identified as “linguistic” or is embedded within an ideology or episteme with its own assumptions and ideologies. But the comparison of theories approach only really works if we do not accept uncritically the present model for what is “linguistic” as the criterion for comparison. A second problem is that focusing on theory and theorizing as a distinct mode of discourse gives primary attention to philosophical and grammatical theory texts as the basis for a history of linguistics as part of intellectual history. Sometimes this theory preference is not implicit at all. Anat Biletzki, for instance, claims that the “traditional disciplines of grammar and rhetoric are first candidates for such research [on the history of pragmatics], but it is in philosophy – and specifically, philosophy of language of the past – that true pragmatics is unearthed” (1996: 455). Grice and Searle would no doubt agree (but without Kant). I’m not sure how much Rosier would. My approach to the history of medieval pragmatics in this book is different. Grammatical theory is fundamental for understanding medieval pragmatics, but a history of linguistics is more than theory. There’s theory and practice and metaunderstanding in between. In this book I read out a critical history of linguistics. I explore linguistic and grammatical ideas and practice in various medieval discourses, not only linguistic, semiotic, and philosophical theory but also discourses situated in other intellectual, institutional, and social contexts, including life-writing, poetry, and accounts of dissent recorded by Church authorities and dissenters themselves. Not exactly “history from the ground up,” as social historians often say to characterize their work, but not altogether different either. I focus on languages of knowledge/power, implicit meaning, strategic language, and discursive conflict, recognizing and examining where a cultural hegemony supporting agreed-upon signification, meaning, and communication modes begins to fracture, multiply, or disseminate beyond itself in alternate hegemonies of dissent. In these alternate spaces, expression, agency, and understanding are called into question and resubjectified by secular and religious authorities and also by counterhegemonic and alienated voices. A critical history of linguistics or medieval linguistics should make manifest structures of knowledge and the ideological and power relations in dominant theory and ideas as well as the implications of praxis for both hegemonic and counter or alternate hegemonic practice. Ideas about language and pragmatics do not belong only or even primarily to grammarians, linguists, and philosophers. They are part of everyone’s implicit working knowledge and metaunderstanding of language as they use and perceive it (cf. Gramsci 1971: 5-9). The recent

24 

The Medieval Life of L anguage

emergence of the history of missionary linguistics in the historiography of linguistics has helped clarify the complex relations between linguistic description, linguistic theory, social ideology, and material history and foregrounded how people’s constructions of linguistic ‘knowledge’ take different forms in different areas of society. Lovejoy’s “unit ideas” often belong to the elite Western world of philosophical and theological discourse, but linguistic knowledge was and is distributed and constructed across social, historical, and geographical spectrums, however unevenly. Our objects of historical inquiry are constructed by the language and assumptions of the inquiry itself and also by the twin challenges – resistance and insistence – of materials of the past, if we recognize them as historical materials. They challenge and test our assumptions about what we know and the authority we presume we have to know them. So, this book situates pragmatics in two medieval contexts. Scholastic philosophy, logic, and semantics shaped medieval ideas about language and communication in significant ways. At the same time, later medieval pragmatic thinking and metapragmatic awareness were also embedded in other social contexts and discursive practices besides grammar and semantic theory. Medieval Christianity’s doxa and institutional authority had a direct impact on medieval pragmatics, but ideas about pragmatics were as much influenced by philosophy of language, philosophy of mind, and other intellectual concepts, especially as received from Aristotle and Augustine. Pragmatic competence is part of any linguistic community and also individual speakers’ competence and interactions. Pragmatic ideas, on the other hand, form part of the intellectual and social culture which gives language its formal intellectual presence over and above concrete practice. That is, ideas are part of social practice. In addition to intellectual culture, medieval pragmatic thinking was embedded in and sometimes constitutive of religious controversy and dissent, life writing, poetics, Bible translation and other social contexts. The history of medieval pragmatics outlined in the following chapters historicizes theory and explores how pragmatic and metapragmatic thinking intersects with and informs many sociocultural discourses, texts, and understandings. This sociocultural exploration of medieval pragmatics also shows how those ideas and practices informed and produced people’s attitudes toward agency, subjectivity, and sociability. The field of late medieval public discourse was heterogeneous and multilayered, and pragmatic theory and practice reveal how and how much dominant discourse and hegemony were contested by and with competing voices. I take a broad interdisciplinary view of the history of pragmatics as the study of how language use has been part of theorizing and resolving

Introduc tion: Where is Medieval Pr agmatics?

25

intellectual and everyday social problems. Medieval pragmatic ideas and metapragmatic awareness manifest in various contexts: philosophical and grammatical analysis, people’s linguistic and communicative repertoires, social interactions, structures and ideas of power, literary and cultural representations. These discursive contexts are intellectual, poetic, official, personal, cooperative, or antagonistic. Comparing intellectual, artistic, and controversial texts, we can critically reflect on the linguistic and social implications of different sorts of knowledge discourses, how they inform representations of identity and social status, and how they inscribe power or dissenting relations. As part of a critical history of linguistics, we also ask: For all the differences, what in that past can we reuse and how? The analyses of texts and ideas in the following chapters emplot a discursive history of late medieval pragmatics conceived as a “field of fields” (Bourdieu 1990: 66-68). In this sense, a field is not a discipline or academic subject. Rather, a field is a social space into which people are thrown and which is maintained or regarded as coherent by collective habitus, behavior, discourse, and belief, which manufacture and reproduce the conditions by which the field perpetuates itself. A crisis emerges when (some) people challenge the dominant discursive practices of a field, calling into question the assumptions and identities by which the field has maintained itself as a social formation. Through alternate or counter-hegemonic practices, groups imagine, create, or invent a socially recognizable new field whose autonomy threatens (antagonizes) an existing one.1 Pragmatic thinking depends in part on the “unthought presuppositions that the game [of the field] produces and endlessly reproduces” (Bourdieu 1990: 66). Pragmatic strategies and metapragmatic awareness enable people to reimagine and manipulate existing social interactions and relations, unpack occluded intention, and make and remake social groups and agency. That is, pragmatics is at once part of the habitus and part of the transformative potential inherent in the field. Whereas “historical pragmatics” focuses primarily on language change (e.g., pronouns of address, insults), the history of pragmatics focuses on the relations between theory and practice in intellectual and everyday discourses in one or more historical periods. As with any critical or historical inquiry, uncovering and foregrounding people’s thinking with and thinking about pragmatics depends on where we look and how and why. Looking outside language study per se does not mean we are looking beyond linguistics or beyond language and certainly not beyond human 1 On the critical and discursive potential of antagonism and crisis for producing change in social space, see Laclau and Mouffe 2000:79-131.

26 

The Medieval Life of L anguage

experience. For communication and language analysis, context is everything but sometimes hard to pin down. This book is organized into three parts. Each chapter addresses one or more key topics in the history of medieval pragmatics: meaning, intentio, sinceritas, expressive form (including grammar, ellipsis, indirectness, figurative language), equivocatio, affectus, suppositio, politeness, cooperation, dialogism, discursive power. Chapters sometimes overlap or are in dialogue with one another in that they discuss pragmatic topics in different kinds of texts or discourses or treat pragmatic ideas as they emerge among different social groups. In the first part (Chapters 1 and 2) I take up medieval grammar, philosophy, and literature in relation to semiotics and theories of language structure and pragmatics. I discuss medieval ideas about interjections and elliptical expressions as they bring to the fore pragmatic perspectives and the role of affectus in the theory of language. Chapters 1 and 2 discuss grammar and pragmatic thinking in the schools after 1050. Although they did not use the term pragmatic, Roger Bacon, Peter (of) John Olivi, and others proposed powerful and far-reaching theories of language and meaning in pragmatic and social contexts. The thirteenth-century grammarians Rosier refers to as ‘intentionalists’ pushed out the parameters of grammar and logic as received from the Late Latin grammarians and twelfth-century commentators and articulated semiotic theories of language from a pragmatic perspective. The grammar, status, and expressivity of interjections became a topic for rich theorizing and pragmatic analysis. Some thirteenth-century grammarians focused on the role of affectus, feeling, and emotion in expression and expanded the idea of verbal meaning to include both cognitive and affective signification as understood in specific contexts. Priscian (Institutiones grammaticae, 2.15) had provided some structure for the contrast between assertive sentences referring to substances and nonassertive sentences signifying mental dispositions. Interjections and other kinds of emotional expressions became a launch pad for interrogating meaning making and double articulation within a grammatical system and for rethinking signification, reference, and affect in terms of disposition and interpersonal rather than only intention-based semantics. The second part (Chapters 3 and 4) reveals how many of the pragmatic ideas in scholastic grammar and philosophy were revisited or reflected in practice in late medieval English vernacular texts by Chaucer. I address the pragmatic role of context and dialogue in Chaucer’s poetry and prose with an analysis of the functions and significations of the Middle English interjection allas. I discuss Chaucer’s Miller’s Tale as an encounter with

Introduc tion: Where is Medieval Pr agmatics?

27

grammatical theory and pragmatics more generally, for comic and satiric effect to construct identity and affective power. In the third part (Chapters 5 and 6) I discuss medieval pragmatic thinking and pragmatic awareness as they emerge in and shape narratives of dissent and social or institutional conflicts by Bernard Gui, William Thorpe, and Margery Kempe. Grammatical theory and literature were not the only places where pragmatic ideas and metapragmatic awareness were productive or contested. Thorpe and Kempe’s writing shows that people’s thinking about how language is used to define, express, control, and resist also informed how they pragmatically and metapragmatically constructed their speech or writing for social survival and asserted subjective authority and agency in asymmetric social or institutional interactions. Medieval grammarians’ and logicians’ concerns with reference and equivocatio (ambiguity, polysemy, vagueness) were reinterpreted in controversies about how heretics and nonconformists talk in institutional situations. Chapters Four, Five, and Six use critical discourse analysis (CDA), conversation analysis (CA), and Bakhtin’s theory of dialogism to elicit speakers’ pragmatic and metapragmatic understanding implicit or explicit in representations of talk between narrative characters and between medieval inquisitors and dissenters or reformers. The life writings of Thorpe and Kempe reveal sophisticated strategic uses of equivocatio and modality as metapragmatic practice. Gui’s account of heretics’ speech suggests that such pragmatic understanding was part of official knowledge. Both inquisitors and dissenters display pragmatic and metapragmatic understanding of what participants are doing, but from different standpoints as they struggle for discursive control. CA and dialogism set pragmatic thinking in a sociolinguistic frame. Institutional discourse about disobedient or transgressive speech becomes an historical source for our understanding of pragmatic and linguistic knowledge. Linguistics or grammar or pragmatics is a discourse or field situated in a social context. A history of linguistics is a history of discourse in different social contexts. The history of linguistics can, or should be, a critical practice informed by critical discourse analysis and analysis of broad social and intellectual fields as contexts. A critical history of linguistics contributes to a socially-informed and nuanced perspective on the contemporary discursive formation called Linguistics and the orders of language, knowledge, authority, and subjectivity from which it has emerged. With semiotic and textual analysis, we can retrieve and critically examine pragmatic ideas and understandings and metapragmatic awareness embedded in medieval texts composed by people from different social and discursive groups.

28 

The Medieval Life of L anguage

Medieval pragmatic theory and thinking and metapragmatic awareness have important connections with contemporary pragmatics, not least around notions of implicitness, speech acts, power, and context. But medieval grammarians did not necessarily draw the boundary between semantics and pragmatics or between linguistic and nonlinguistic vocalizations in the same way linguists and philosophers today might. While we have inherited a good deal of linguistic metalanguage from ancient and medieval grammarians, earlier grammarians and linguists used those terms in ways and for reasons which are not necessarily our own. Their analyses of speech contexts and meaning reveal a sophisticated grasp of the nuances of dynamic contextualization, making meaning on the fly, as it were. Moreover, linguistic ideas were part of people’s intellectual, cultural, and intuitive repertoires for participating in the active life of the society, whether officially, playfully, poetically, or under coercion. A critical history of linguistics doesn’t assume that all linguistic knowledge is captured in grammatical or linguistic theory nor that ‘proper’ pragmatics is relevant primarily for how interlocutors cooperate with one another to achieve a shared goal in a reasonable way or achieve the enlightened view that language in use is language tout court. One of the continuing claims in this book is that pragmatic understanding manifests the embodied connection between reason and emotion, between act and habit, between intention and reception, in interpersonal communication. Implied meaning is not necessarily always cognitive. Conversational conflict and antagonistic constructions of agency, polysemy, or coercion reveal as much about pragmatic thinking as does the analysis of successfully interpreted indirect communication or the achievement of reasoned consensus. Grice (1975) argues that communication depends on speakers finding agreement on shared goals and flexibility with respect to the designated speaker’s intentions and illocutionary purposes. It’s no argument that cooperation and shared understanding are essential strategies and values if we hope to make society more just or gain richer, more inclusive understandings of human experience. But those aims are not the be-all and end-all of pragmatics nor of social interaction or even of communication. Recognizing motivated conflict or legitimate difference also goes a long way toward beginning to create a just or cooperative society. Antagonisms and alternate hegemonies are as much part of that struggle as is cooperation. Pragmatics, antagonisms, and power emerge together, but not always with the same force in the same direction. Intellectual history and the history of linguistics can contribute to the work of critique by asking how giveness has come about and why what

Introduc tion: Where is Medieval Pr agmatics?

29

goes without saying needs to be spoken. A critical history of medieval linguistics, linguistic ideas, and pragmatics shows us some of what we have retained and also what we have left out from the past we inherit, the past we acknowledge, and the past we assume or repeat without question. The history of linguistics also shows us some of that past, the past as Other, the one we didn’t even know we could inherit. Historicizing our situation entails making the past relevant by asking what in the past is reusable and for what purposes. A critical history of linguistics, medieval or otherwise, can effectively and thoughtfully intervene if and when it contributes to a broader understanding of idea formations and their consequences as social practice. As a critique of ideology, the history of linguistics can offer a grounded knowledge of possibilities and experiences which can help destabilize repressive knowledge/power relations in institutions or in our social interactions and futures.

1

Medieval Pragmatics: Philosophical and Grammatical Contexts Abstract This chapter discusses grammar and pragmatic thinking in the schools after 1050, focusing on internal history and semiotics. Roger Bacon and Peter (of) John Olivi proposed far-reaching theories of language and meaning with pragmatic perspectives. Bacon foregrounded pragmatic understanding in his semiotics of language and proposed context and communicative effectiveness rather than formal completion as the criteria for linguistic acceptability. Grammarians’ analysis of interjections, speech fragments, and emotional expressions opened new ways of understanding meaning-making, discursive interaction, and double articulation within a grammatical system and new pragmatic thinking about signification, reference, and affect. Olivi’s pragmatic approaches to some philosophical and theological accounts of language and expression focused on speakers’ and listeners’ responsibilities and how words’ meanings and contexts can change over time. Keywords: actus significatus, actus exercitus, affectus, conceptus, Roger Bacon, Peter John Olivi, sign theory

After 1050 CE, medieval philosophers and grammarians used the traditional vocabulary and concepts of Latin grammatica as well as metalanguage and concepts derived from Aristotle and Augustine to describe and theorize language in structural and pragmatic ways. They never used the word pragmatikē or a Latin or vernacular cognate, but grammarians and philosophers reworked existing metalanguage, concepts, and theories to produce two new approaches to grammar and language. One theory, associated with the grammatica speculativa and related general philosophical grammar, adopted a framework based on Aristotle’s causes to develop a theory of autonomous language according to the modi of being, understanding, and

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch01

32 

The Medieval Life of L anguage

signifying. The modistae focused on the four orders of syntax: natural, grammatically obligatory, rhetorical, and logical (Kneepkens 1990). They sought to explain logical order, rather than word order, and congruitas with respect to semantic relations, much as modern generative and dependency grammars do. The other theoretical approach, which never really gained a name, combined Aristotle’s account of language and mind with Augustine’s theory of language and signs to confect a more functional perspective on communication and language in use. Aspects of medieval functional theories anticipate current pragmatic and functionalist theories of language and social interaction. In this chapter I discuss pragmatic thinking in mostly thirteenth-century philosophy of language with special attention to the work of Roger Bacon (OFM, 1214/1219-c. 1292) and Peter (of) John Olivi (OFM, 1248-1298). In Chapter 2 I zero in on a productive and provocative aspect of medieval grammatical and pragmatic theory, grammarians’ analyses of the interjection and ellipsis as meaningful language use outside explicitly logical order.

Three Terms and a Theory Three key theoretical concepts in scholastic thinking about language structure and usage were modus, actus, and conceptus. The theory of modi, ways of doing something, was the prevalent theoretical framework in scholastic grammar and philosophy. Some grammarians used the theoretical category actus, activity, to ground their language theory in pragmatics and use. The theory of conceptus in language and cognition gave rise to an alternative theory of affectus. All three concepts were linked with the notion of usus, but in different ways. The increasingly complex relations of logic and grammar in the schools and universities of the eleventh and twelfth centuries had significant consequences for medieval theories of language.2 In the mid-twelfth century, the theory of different modi for being, understanding, and signifying was developed to account structurally for Langage in the form of Latin. Modus had a specialized meaning in medieval grammar and philosophy. According to Alexander of Hales (OF, c. 1185-1245), Shropshire-born and University of 2 L. M. De Rijk’s Logica Modernorum: A Contribution to the History of Early Terministic Logic (1962-1967) remains the standard account of grammatical theory and the “doctrine of fallacy” in twelfth- and thirteenth-century philosophy of language. Cf. more recently, Margaret Cameron 2015.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

33

Paris educated, modus means “dispositio rei in esse terminatae” [the order of things concluding in being]. Modistic grammarians sought to explain how a logical order is the most appropriate way of representing significations by means of consignifications (cf. Kelly 2002: 54-56). Modi exist with mensura [measure], defined as “esse voluntatis in eo quod conformat voluntatem suam voluntati divinae” [the being of the will because it conforms its will to the divine will] (1924-1948: 2.1.⁋33).3 Language is controlled by the will, a concept which medieval pragmatics exploits and changes in important ways, as we shall see. The modi were correlated with aspects of essendi, intelligendi, and significandi within the ontological and epistemological triad res, anima, verbum. Speculative grammarians rendered language as a formal system and a human instrument with realist signification. The grammarians’ terminology is tricky. Significatus/significandi have several meanings: intended meaning, or making meaning, or ‘signification’ in the sense of a contextualization of the utterance beyond immediate reference or sense. Usually, significatus/significandi did not mean ‘reference’ or ‘refer to.’ Secondary signifying, consignificatus/consignificandi, usually pertained to linguistic forms or word accidents such as morphology (declensions, tense, voice, derivation). For some grammarians, consignification also included figurative or prophetic language uses (Kelly 2002: 54). The modi significandi are frames and procedures for doing something linguistically, making meaning or communicating through language guided by the will. The modi intelligendi are the frames and procedures by which listeners make sense of what they perceive materially and linguistically. How much common and procedural knowledge speakers and listeners share becomes an important epistemological question with respect to language as communicative activity. Actūs are explicitly language activities, parole rather than langue. In scholastic theory actus was divided between actus significatus and actus exercitus, roughly corresponding to contemporary speech act theory’s distinction between constative and performative utterances (Austen 1962). Beginning in the mid-twelfth century, logicians paid new attention to the formal aspects of propositions (sentences) and to the presuppositions, implications, and ambiguities individual utterances posed. As Nuchelmans (1988) and Rosier 3 Alexander’s Summa universis theologicae was probably completed by his students and colleagues. The most recent editor retitles the work Summa fratris Alexandri; see Summa universae theologiae (Summa fratris Alexandri), 1924-1948, ed. Bernardini Klumper and the Quarracchi Fathers.

34 

The Medieval Life of L anguage

(1994: 157-206) have shown, the distinction among actūs became productive for post-1200 linguistic thinking in Europe. Making use of Aristotle’s analysis of propositions in De interpretatione, logicians and then Priscian commentators and grammarians classified sentences or propositions according to their formal structure (congruitas) and sometimes also their pragmatic actus. Formally, a sentence requires a subject and a predicate to be a “complete” utterance (completus, perfectio). Priscian (fl. early 6th c.), however, had noted that a single Latin word or phrase could constitute a complete utterance in response to a question (adjacency pair) or when used as an imperative (Institutiones, 2.15; KGL 2:54). Despite Priscian, many medieval logicians set incomplete or elliptical sentences aside to focus on ‘complete’ propositions. The antagonism between complete and incomplete utterances would become productive in medieval grammar. Whether utterances conformed with the idealized model of syntax or whether they could be understood or communicative in a given context regardless of form opened the field to pragmatic theory. Grammarians pointed out that enumerating and blessing are speech acts, two types of actus exercitus. Gosvin of Marbais (13th c., fl. 1270) uses the example to neatly illustrate the difference between actus significatus and actus exercitus. The number words unus, duo, tres refer to quantities, whereas the utterance “Unus, duo, tres” performs the act of enumerating and is a “complete” sentence although it lacks a verb (Gosvin of Marbais, Tractatus de constructione, 1998: 83; cf. Magister Johannes, Sicut dicit Remigius, MS Bibliothèque nationale lat. 16618, in French in Rosier 1994: 259). Roger Bacon points out that the benediction “In nomine patris, et filii et spiritus sancti” performs a blessing but does not refer to anything. The utterance is understood as “complete” in that it enacts a blessing on the congregation with the performative verb benedico either explicit or implicit in the syntax (Summa grammatica, in Rosier 1994: 187). The distinction between actus significatus and actus exercitus prompted some medieval logicians and grammarians to expand the theory of linguistic communication to include not only utterances which are formally (grammatically) complete according to the vocalized or sentence meaning (ad sensum) but also those spoken or heard as complete according to the intellect (ad intellectum) or the speaker’s intention (secundum intentionem proferentis). Among logicians, Peter of Spain (probably OP, c. 1205-1277?) in his influential Summulae logicales, follows Aristotle and Boethius (c. 480-524) and distinguishes complete (perfecta) from incomplete (imperfecta) utterances in terms of propositional form and a theory of meaning based around idealized oratio (Summulae logicales, ed. de Rijk, 1.6-7; 1972: 3-4). Peter’s Summulae became a standard logic text at the University of

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

35

Paris and attracted several influential commentaries. Many thirteenthcentury grammarians described how an individual utterance is “perfected” (perficitur) by one or another modus. Homo currit is a complete utterance ad sensum with a specif ied subject and predicate. More pragmatically inclined logicians analyzed Lego [I read], the adjacent expression bene verbera [Well said], and single word utterances (Honestas, Heu) as complete utterances ad intellectum. They claimed the full sense of those utterances can be inferred from context and that the speaker and listener usually have the pragmatic competence and cultural knowledge to do so. Imperative, vocative, and echo responses can all be considered complete utterances from the perspective of communicative will and accepted use, even though they lack some grammatical constituents for formal or logical congruitas. In this respect, medieval pragmatic accounts of speech exceed some terministic logicians’ and speculative grammarians’ primary attention to logical order and propositions as the basis for meaning. The sensus of not only incomplete utterances but also words used figuratively rather than literally depends on the interpreter understanding the words’ possible meanings and inferring the speaker’s position or intent from the words used, how they signify, and to what they might refer. At some level, an utterance can be understood as meaningful simply because a listener or reader interprets it to be so – Ipse dixit. Communication succeeds because people accept what they interpret an utterance to mean. But successful communication isn’t necessarily the same as agreement. Understanding is never just simple or one way. The word class adverbum illustrates how in medieval grammar the pragmatics of actus exercitus were used to analyze constative and performative speech and ellipsis in terms of inferring utterance or speaker meaning from context. Typically, grammarians and logicians analyzed the adverb as functioning formally with the verb or clause in the same way an adjective functions (adjoins) with the noun. In addition to modifying the verb, adverbs could signify both logical relation and mental disposition. Therefore, the adverb is governed (regere) by the verb (e.g., Pseudo-Grosseteste, Tractatus de Grammatica, ed. Reichel, 1976: 58). However, adverbs did not sit easily in the received Latin grammarians’ analysis of parts of speech. In thirteenth-century schools many scholars were interested in syncategoreumata, which included adverbs, adjectives (totus, omne), prepositions, conjunctions, interjections, some verbs, negation, and a few other modal expressions. Logicians such as Nicholas of Paris (fl. 1230-1240) devoted separate treatises to the class of syncategoreumata (words that cannot serve as the subject or predicate of a proposition). They discussed how those words signify and how they determine or modify the

36 

The Medieval Life of L anguage

meaning of phrases or clauses. In his Syncategoreumata, Nicholas, following Priscian, states “omnis dictio significat mentis conceptum” [every part of speech signifies a mental concept], and everything that signifies a concept in the mind signifies a some-thing (ed. Braakhuis, 1979: 3.9-10; cf. 3.10-17. My translation.). Whereas categoreumata (noun, verb) signify substance, quality, or action in themselves, syncategoreumata (pretty much every other form of speech not explicitly a noun or verb) do not signify in themselves but in conjunction with other words (“aliae sunt quae per se non-significant sed in coniunctione ad alias; et tales dicuntur sincategoreumata”; ed. Braakhuis, 1979: 2.1). Similarly, Peter of Spain stated that syncategorematic words “significant res que sunt dispositiones subicilium vel predicabilium …” [signify things that are dispositions of things capable of functioning as subjects or predicates] (Syncategoreumata, ed. de Rijk, trans. Spruyt; 1992: 38-39). In Peter’s definition, syncategorematic words’ “dispositions,” meaning something like attitudes toward other things, are themselves considered res [things]. Pragmatically, interjections are especially interesting because they fall somewhere between the two classes of categorematic and syncategorematic words (see below, Chapters 2 and 3). Despite not being able to form a complete proposition on their own, syncategorematic words such as praeter [besides (exceptive sense), prep. or adv.], nisi [unless, conj.] and interjections signify conceptually or affectively because they are part of thought (cf. the influential twelfth-century anonymous Abbreviatio montana [Paris, Mont St. Geneviève], ed. de Rijk, 1967, vol. 2.2: 78). For logicians, such words play an important role in cognition and expression because they are crucial to producing propositionally meaningful utterances in that they connect substance words together. Following Priscian, Peter of Spain differentiated those adverbs which determine the status of the proposition (necessario, contingenter, impossibiliter, etc.) from those which determine the status of the verb action (velociter currit, etc.) (Summulae logicales, ed. de Rijk, 1.19; 1972: 11-12). Medieval grammarians and logicians also debated how certain words (omnis, desinet, praeter, some pronouns, etc.) could be used differently to produce different kinds of meanings. Words, they implied, are functionally categorematic or syncategorematic depending on the utterance context or situation. Syncategorematic words express relations among the parts of a discourse (nouns and verbs) and among the substances and actions signified. In a sentence or a discourse, a pronoun is used syncategorematically when it refers to a prior word, explicit or implicit in a discourse, and thus produces syntactic cohesion. Syncategorematic words also establish the speaker’s position and disposition with respect to objects and listeners (deixis).

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

37

Aquinas (OP, 1225?-1274) considered the signification of syncategorematic words such as the adverb in relation to epistemology, both divine and human. He argued the adverbum “importat modum cognitionis” [brings about/ introduces a mode of understanding] with respect to either the object known or the knower or both. If so, Aquinas wondered, how do we understand the syntactic and pragmatic relations in the phrase, “sic cognoscere aliquid sicut in cognoscente est …” [ So to know a thing as it is in the knower …], where sic functions adverbially? Does the adverb adjunct to cognoscere indicate how an Object-thing exists in itself or how a Subject-agent acts toward or knows the Object-thing? Aquinas adopts the second reading with theological implications: Si vero intelligatur secundum quod hoc adverbium sic importat modum ex parte cognoscentis, verum est quod sic solum cognoscens cognoscit cognitum, secundum quod est in cognoscente, quia quanto perfectius est cognitum in cognoscente, tanto perfectior est modus cognitionis. Sic igitur dicendum est quod Deus non solum cognoscit res esse in se ipso; sed per id quod in seipso continet res, cognoscit eas in propria natura; et tanto perfectius, quanto perfectius est unumquodque in ipso. [If however the adverb so is understood to import the mode [of knowledge] on the part of the knower, in that sense it is true that only the knower has knowledge of the object known as it is in the knower; for the more perfectly the thing known is in the knower, the more perfect is the mode of knowledge. We must say therefore that God not only knows that all things are in Himself; but by the fact that they are in Him, He knows them in their own nature and all the more perfectly, the more perfectly each one is in Him.] (Aquinas, Summa Theologiae, I, q. 14 a. 6 ad 1).

Formally, we understand the adverb depends on the verb, but pragmatically, in the context of the speaker’s intention, is the semantic or grammatical force (virtus) of the adverb sic to inflect the object phrase or the entire clause? For Aquinas, the form of the sentence (‘virtus sermonis’) alone does not determine the speaker or listener’s understanding of the sentence, because sentence meaning is underdetermined by syntax, a fundamentally pragmatic insight. In his analysis, Aquinas contextualizes the syntax by supplying external presuppositions about divine knowing and objective knowledge. Contemporary linguists resolve Aquinas’ problem with the grammatical and semantic scope of sic by distinguishing between phrasal adverbs and sentence adverbs. Medieval grammarians analyzed the syntax differently. Words within the class adverbum remained functionally unstable and

38 

The Medieval Life of L anguage

pragmatically equivocal because their syntactic roles were fluid. Sometimes, grammatical instability can kickstart grammatical and pragmatic innovation, as we shall see with medieval discussions of interjections. Alexander of Hales, by some accounts one of Aquinas’ favorite theologians, also discussed the semantics and pragmatics of adverbs when considering implications and presuppositions in language use. As a regent master, Alexander was responsible for introducing Peter Lombard’s Sentences into the University of Paris Arts curriculum and for promoting the application of new logical modes of analysis to theological and scriptural questions (Cullen 2006). In his Summa universae theologiae, a commentary on Lombard’s Sentences, Alexander discusses adverbial inferencing by considering the sentence, “Deus me lecturum mutabiliter scit.” We can read the sentence in two ways: God contingently knows me to be reading or God knows me to be reading in a contingent manner (1924-1948: 1:265a). Latin adverbs are indeclinable, so mutabiliter does not have case marking. The sentence’s formal congruitas and meaning and the relation of the adverb to other parts of the sentence cannot be determined from formal or surface rules alone. Grammatical relations among the words have to be construed using internal (cohesive) and external (coherent) contexts and relevant presuppositions. For the statement to be true, says Alexander, mutabiliter can’t refer to the subject Deus of the active verb because by definition God is immutable (external context). Therefore, mutabiliter must attach to the object me lecturum (internal context). If so, then we can expand the sentence as “Deus me mutabili modo lecturum scit,” because God knows my reading will be contingent on circumstances which I cannot foresee. Alexander’s analysis of sentence meaning ultimately relies not on grammatical form but on the interpreter’s prior knowledge and semantic inference as to the ontological and theological attributes associated with the words Deus and me and the difference between human and God’s time. Alexander distinguishes a true and a false reading of the sentence based on a semantic frame of reference which is exterior and effectively prior to the utterance. The presupposition God is immutable and omniscient sets the context for construing the grammatical linkages among the individual words, and Alexander must have assumed his projected readers shared that existential and ontico-theological understanding. He assumes readers share an understanding of the theological doxa that God is immutable, has omniscient foreknowledge, and is omnipotent. Therefore, they can link the adverb to the appropriate syntactic element, with me lecturum rather than with Deus. From a pragmatic perspective, the ‘proper’ or intended meaning as Alexander reads the sentence cannot be derived from the form of the

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

39

utterance. Rather, the preferred meaning depends on the communicative situation and theological grounds the speaker/writer and listener/reader are presumed to share. The implications of the actus significatus/actus exercitus distinction were far reaching for grammatical and pragmatic theory and indirectly influenced other thinking about discourse and performativity and the function of deixis, ellipsis, relevant context, and extralinguistic situations in the production of meaning. Some grammarians used the concept of actus exercitus to analyze the consignifications of words and expressions. Robert Kilwardby (OP, c. 1215-1279), for example, discussed consignification in terms of words’ extended meanings and inferences, including figurative language, rather than grammatical accidents (e.g., In Donati Artem maiorem III, 1984: 41.233-237; cf. Kelly 1992: 196). These counter-modi explorations of meaning as use show how medieval theory about language structure, language use, and metapragmatics in intellectual discourse often ran alongside official or everyday understandings of language use rather than directly influencing them. Grammatical theory and philosophy of language, in addition to shaping social ideas about language, emerged from some of the same underlying ideas and practices which also informed everyday medieval pragmatics and metapragmatic awareness. The third idea of scholastic language theory relevant to medieval pragmatic thinking was the distinction between conceptus and affectus. The dominant medieval theory of signification was conceptualism. Based mostly on Aristotle (De interpretatione, 1; 16a3-9) and Priscian (Institutiones, 11.7), the conceptualist theory proposed that spoken or mental words (vox) refer first to mental concepts (mentis conceptum, cogitatio, intellectus) and then indirectly to the external objects or abstract ideas to which the concepts refer directly (cf. Panaccio 1991). Aristotle’s theory of cognition in De anima, newly available in the twelfth century from Arabic sources and translated into Latin by William of Moerbeke, expanded the medieval conceptualist theory. According to Aristotle, “Not the rock itself but its image/similitude is in the soul/mind” (De anima, 3.8; 431b29-432a1). Aristotle’s text, much quoted in thirteenth-century logic and grammar, uses an unnuanced notion of mental representations as ‘likenesses,’ which had significant implications for medieval theories of language and pragmatics. The most common medieval philosophical term for “similitude” was species. Some defined a concept as a “mental word” (verbum mentis) (e.g., Aquinas, Quaestiones disputatae de potentia, q. 8, art. 1). Concepts formed in the mind mediate the relation between language and the world of concrete and abstract reference. The conceptualist theory of linguistic signification was adopted by philosophers

40 

The Medieval Life of L anguage

as diverse as Boethius, Abelard, Aquinas, and Ockham, but not always in the same way (e.g., Boethius’ In librum Aristotelis Peri hermeneias commentarii (editio secunda), Book 1; LLT-A, Bk. 1). In his philosophy of language and mind Abelard says the theory of concept constitutes an adverbial rather than a resemblance or conformity theory of understanding (cf. King and Arlig 2018: sects. 4-5). The object apprehended in the act of perception functions as a kind of adverb describing the manner in which the object is perceived. Abelard’s position is consistent with a pragmatic theory of language. Bacon, as we shall see, rejected the conceptualist theory of linguistic reference altogether, but not the grammatical framework for analyzing and theorizing about language nor its pragmatic implications. In his influential Logica, Lambert of Ligny-le-Chatel (OP, fl. 1240-1260)4 discussed concepts and the signification of terms and succinctly sums up conceptualism: Significatio termini est intellectus rei ad quem intellectum rei vox imponitur ad voluntatem instituentis: nam, sicut vult Aristoteles in primo Perihermeneias, voces sunt signa passionum que sunt in anima, id est in intellectu; intellectus autem sunt signa rerum … quia a voluntate fit unio intellectus rei cum voce, et in hoc consistit vocis impositio, ideo vox dicitur significare ad voluntatem instituentis. [The signification of a term is the concept of a thing, for which concept of a thing (intellectus rei) a vocal sound (vox) is imposed according to the will of the one instituting it, for, as Aristotle proposes in Book One of On Interpretation, vocal sounds are signs of affections (passionum) that are in a soul, that is, in an intellect, whereas concepts are signs of things … because the union of the concept of the thing (intellectus rei) with the vocal sound (vox) comes from the will, and it is in this that the imposition of a vocal sound consists] (Logica, ed. Alessio, 1971: 205.12-16, 205.26-29; Logica, trans. Maloney, 2015: ⁋1232, 1235. Translation slightly modified.).

According to Lambert, the mind connects a sequence of sounds (vox) not directly with an external referent but with a mental concept which is the immaterial representation of an external referent. What pulls the sounds, concept, and referent together is the will (voluntas). Language is a humanmade signifying system. Language’s referentiality is mediated by concepts 4 Maloney 2015 (xxxix-xl) disagrees with Libera (1981) and Alessio (1971) about Lambert’s birthplace. Maloney also believes Lambert taught logic at a Dominican school in Auxerre and later in Paris. Sometime between his residence in Auxerre and Paris, Lambert may also have tutored Prince Theobald, future king of Navarre (1239-1270).

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

41

of things real (Socrates, olives) or abstract (justice) or imagined (chimera). This is scholasticism’s soft realist ontology for language. A person’s cognitive process begins with sensory input, from which mental concepts are formed. Then a concept is directly connected with a sequence of sounds by imposition whose implicature is an external referent: “unde cum vox sit signum intellectus et intellectus signum rei et sic rei est signum. Vox que est signum signi, scilicet intellectus, erit signum significati, scilicet rei, sed immediate est signum intellectus, mediate autem signum rei” [since a vocal sound is a sign of a concept and a concept is a sign of a thing, so [a vocal sound] is also a sign of a thing. A vocal sound that is a sign (signum) of a sign, i.e., of a concept, will be a sign of what is signified, namely, of a thing, but directly it is a sign of the concept and indirectly a sign of the thing] (Logica, ed. Alessio, 1971: 206.2-6; Logica, trans. Maloney, 2015: ⁋1236). Underlying Lambert’s account is the more general scholastic and realist assumption, via Aristotle, that “language imitates nature.” The question is: How? Most conceptualists emphasize the cognitive role of mediating concepts which indirectly connect language, signs of signs, with the world. When scholastic grammarians distinguished between modus conceptus and modus affectus, they usually were trying to resolve formal and semantic problems with a conceptualist theory of signification. This was as true for speculative grammarians as for intentionalist grammarians. However, some grammarians also used the distinction to elaborate more pragmatic perspectives on language, meaning, and performativity. The difference between words or utterances signifying per modum conceptus and those signifying per modum affectus is most clearly illustrated in thirteenth-century discussions of the interjection. Early in his Syncategoreumata Nicholas of Paris states that every dictio signifies a mental concept, but with respect to concepts he doesn’t distinguish nouns from interjections or conjunctions. However, later in the text he revises and expands his notion of signification and distinguishes interjections from other parts of speech. The interjection, he says, signifies under the mode of affectus, whereas the other parts of speech signify under the mode of conceptus (Syncategoreumata, ed. Braakhuis, 1979: 132.11-12, 14-15). According to Nicholas, a word is either a sign of a thing or a sign of a sign. Nouns (exceptio [exception]) and verbs (excipere [to except]) are signs of things or what are taken as things, the abstraction ‘exception’ or the act of ‘excepting.’ Adverbs such as praeter signify conceptually as signs of signs: “hec dictio ‘preter’ significat exceptionem ut per alterum significativa. Cuius signum est quod huius ‘omnis homo preter Sortem currit’ est hec exceptiva ‘omnis homo Sorte excepto currit’.” [the word praeter signifies ‘exception’ in such a way that it is significant of it through another

42 

The Medieval Life of L anguage

[sign]. It is a sign that an exceptive proposition associated with ‘Every man besides Socrates is running’ is ‘Every man is running, Socrates excepted’.] (Syncategoreumata, ed. Braakhuis, 1979: 132.23-133.3; CTMPT, 1:180-181). We can expand Nicholas’ analysis to uncover some of its pragmatic potential. The adverb praeter qualifies the truth claim of the matrix clause or proposition (‘Everyone except Socrates …’), but it has a loose relation to the matrix clause. Interjections such as heu and alas, however, verbally manifest the speaker’s feeling of sadness at the moment of utterance and have a different but still a loose relation with any adjacent clause. Heu does not function syntactically within a clause but rather adds an affective shade or disposition to the adjacent clause. (See Chapter 3 for more detailed syntactic and semantic analysis of allas and affect shading.) As a framing example, some grammarians contrasted heu with doleo. An utterance such as dolor [sadness] or doleo [I feel sad, I suffer] signifies per modum conceptus the speaker’s thought or concept about her feeling at the moment of speaking in the present. Dolor or doleo refers to the feeling as an object or thing (res) of discourse, indexing rather than performing the feeling. Heu, on the other hand, does not refer to sadness but signifies, performs, or enacts sadness with speech (per modum affectus). Meaning is indexed rather than denoted. Interjections are produced with explicit and powerfully signifying affects. They can be considered part of embodied feelings, though not a necessary part. Conventional interjections such as heu express rather than refer to feelings, even though they originate from ratio and the intellect because they are vocalizations. In both cases, feelings and affect utterances are situated between ‘nature’ and ‘conventions’ of socially constructed behaviors. The relation between concepts and emotions or dispositions was a topic of complex analysis and debate among philosophers and grammarians. Not all grammarians agreed with Nicholas of Paris’ sharp distinction between modus conceptus and modus affectus with regard to the parts of speech. The grammarian known as Magister Johannes (13th c.), for instance, argued that signifying a concept per modum affectus blurred the distinction between concept and feeling: “Significare autem conceptum per modum affectus est signif icare conceptum rationis prout subcumbit affectionibus seu passionibus” [To signify a concept under the mode of affectus is to signify a rational concept in so far as it will succumb to/become immersed in feelings or passions] (Sicut dicit Remigius, f. 61ra; cited in Rosier 1994: 61. My translation.) In later chapters I’ll discuss how affectus and interjections shape medieval pragmatic thinking. But first, we need to work out Roger Bacon’s semiotic approach to pragmatics and interjections.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

43

Roger Bacon’s Semiotics and Pragmatics In the next two sections I argue that the philosophical and semiotic work of Roger Bacon and Peter John Olivi highlight how aspects of scholastic thinking about language contributed to medieval pragmatic theory, ideas, and metapragmatic awareness. I don’t present a comprehensive account of Bacon’s or Olivi’s semiotic and language theory. Rather, I am interested in how they deploy key concepts, metalanguage, and examples from scholastic logic and grammar to address pragmatic questions and constitute a pragmatic orientation to language and communication. Bacon’s semiotic theory of signification and language was one of the most thorough and original medieval accounts of how we use verbal and nonverbal signs to make meaning, worlds, and social relations. Bacon’s semiotic ideas were available in his Summulae dialectices (c. 1240-49), Compendium studii theologiae (c. 1292), and Summa grammatica (probably 1235-1245). But his most comprehensive work on semiotics and language somehow became separated from the manuscript of his Opus maius (Part Three, Chapter 2) of which it was originally a part. The section of the Opus now known as De signis, probably composed around 1267, survives in only one manuscript (Oxford, Bodleian Library, Digby 55, fols. 228r-244r. See Fredborg, Nielsen, and Pinborg 1978.). Much of Bacon’s linguistic writing is motivated by his project to make the study of language and languages central to the understanding of the arts and theology. At the same time, as Maloney notes, Bacon was an original thinker “in addressing the traditional ways of making distinctions and, in general, of just looking at things” (2013: 3; cf. Rosier 1994 passim). Bacon’s semiotic theory of language is relevant to pragmatic theory with respect to his definition of the sign, the category of Relation, and the concepts of impositio, equivocatio, analogia, ampliatio, and restrictio. These concepts and terms were part of the discourse of terministic logic, but by and large logicians deployed them as part of a theory of fallacies. They gave little attention to nonpropositional aspects of language use, for example, interjections, natural vocalizations, and figurative language, although some logicians did devote significant analytic attention to syncategoreumata because of how those words affected propositional form and truth value. When Bacon and other grammarians took up the tools of terministic logic to explore interjections and other non-propositional expressions, they manifested the implicit pragmatic potential of logical discourse. Bacon’s semiotics embraces the intersections of language, mind, and objects, both material and immaterial. The English-born Bacon was well

44 

The Medieval Life of L anguage

schooled in Augustine’s sign theory (De trinitate, De doctrina christiana, etc.) and the Aristotelian analysis of utterances (De interpretatione). He lectured at the Universities of Oxford and Paris on Aristotle’s Organon and Priscian’s Institutiones and was familiar with Peter Helias’ (c. 1100-c. 1166) influential Summa super Priscianum. However, Bacon reworked Augustine’s theory of sign making, which emphasized intention, and Aristotle’s analysis of the truth value of propositions in order to produce more empirical and pragmatic accounts of meaning, interaction, contingency, and relevant context in the use of verbal and nonverbal signs. Bacon’s semiotic theory challenged the consensus conceptualist view in thirteenth-century philosophy of mind. He complicated the traditional linguistic distinction between nature and convention with his novel theory of nature and linguistic creativity and his investigation of how actual language use expands and creates meaning. Initially, Augustine defined the sign as triadic, modeled on speech, and for the most part intentional. In De dialectica (c. 387 CE), Augustine defined the sign as “quod et se ipsum sensui et praeter se aliquid et animo ostendit” [that which shows itself to the senses and shows something besides/beyond (praeter, nonexceptive) itself to the mind] (De dialectica, ed. and trans. Jackson, 1975: 86). Praeter suggests the dependence of cognition on sensory apprehension and connotes spatial difference. Something understood in the mind is other than what prompted the apprehension (cf. Kelly 1975). The sign results from a transmuting alterity. What I understand from the sign is other to the immediate sensation of the thing I perceive and register in the intellect as species, which Augustine regarded as the mental representation. Rather than simply copying referents in the mind by means of the senses, sensory perception generates a semiotic transmutation from the physical experience to an idea, concept, feeling, or mental meaning. Augustine later revised this Stoic-influenced definition of the sign in De doctrina christiana (397/426 CE): “Signum est enim res praeter speciem, quam ingerit sensibus, aliud aliquid ex se faciens in cogitationem venire” (De doctrina christiana, 2.1; LLT-A, 2.1.lines 4-5) [A sign is a something which by itself makes something other/beyond come to mind, besides the appearance it offers to the senses]. Augustine’s second definition of the sign emphasizes explicitly sign-making as communicative action from the speaker’s point of view: “Nec ulla causa est nobis significandi, id est signi dandi, nisi ad depromendum et traiciendum in alterius animum id, quod animo gerit, qui signum dat” (De doctrina christiana, 2.2; LLT-A, 2.2.lines 3-4) [There is no reason for signifying, i.e., for giving signs, except to produce and convey into another’s mind what the sign-giver has in his own mind]. Augustine’s definition of the sign in De doctrina christiana is

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

45

still triadic, but now it explicitly involves a perceiver/interpreter to activate the potential sign function. Whatever signifies as a sign, not only words, is a sign precisely because of its difference. Something is perceived and cognized by a mind as a mental image or concept and meaningful yet different from the thing or phenomenon which caused the mind to form the mental image or concept. Augustine’s semiotic definition of speech involves a dialectical relation between speaker/writer and listener/reader. Reception and interpretation are as important as expression, but they aren’t completely determined by the speaker’s intention. Rather, what is understood by a signifying utterance is partly the result of the listener’s shared knowledge and construction of context, relations, and what the listener understands the speaker’s intention to be. Meaning is not simply transferred intact through a conduit from speaker to listener. In some of his earlier writings, Augustine was skeptical as to the efficacy of linguistic semiosis (e.g., De magistro, 10.32). In his later writings, however, he claimed that functionally the only way we can know anything is by understanding signs: “Omnis doctrina vel rerum est vel signorum, sed res per signa discuntur” (De doctrina christiana, 1.2; LLT-A, 1.2.line 1) [All knowledge is either of things or signs, but things/reality are learned/understood through signs]. Knowledge can be attained either through signa naturalia (natural signs) or signa data (given/conventional signs). Augustine’s distinction, derived from Stoic semiotics, depends mostly on causation as the basis for semiotic interpretation. For example, seeing or smelling smoke indicates fire is present, a footprint in the ground indicates a creature or person had passed by, and so forth (De doctrina christiana, 2.1; LLT-A, 2.1.lines 5-7). Peirce identified most such signs as indexical. They point to that which caused the signifier to come into being as a potential sign, which an interpreter actualizes as a signifier (e.g., 1940: 104-115). Bacon’s understanding of Augustine’s theory of signs was filtered through Aristotle’s De interpretatione, Metaphysics, Physics, De anima, and other texts. Working between Aristotle and Augustine, Bacon arrived at a powerful and dynamic triadic model for constructing and understanding spoken and material signs in different contexts. His semiotic theory builds on insights regarding intention and meaning associated with Abelard (1079-1142). Abelard had posited that a speaker’s intentio was the basis for understanding the signification and ethics of any utterance. In Dialectica and then in later writings on logic, Abelard argued that communication depends on the position of the speaker in relation to the utterance. Utterance meaning depends necessarily on the speaker’s prior intention. Speakers or writers

46 

The Medieval Life of L anguage

can use the same words but mean different things by them in different contexts (cf. Mews 2005). In De signis Bacon explicitly expands the role of intention to include both speakers and listeners, perhaps in line with Augustine’s later theory of the sign. The construction of signification emerges from the dynamic and constitutive relation between speech and understanding, not just intention. How, Bacon asks, does intention shape the relation (relatio) between speaker and listener and actualize potential signification? Bacon’s answer is, Dialogism: “si non sit cui actu significetur non est signum in actu sed in potentia tantum. Aliud enim est esse actu impositum ut significet quibuslibet est possibile, aliud esse actu signum” [… if there is nothing [no interpreter] to which it [a sign] signifies, it is not actually a sign, but one only potentially. For it is one thing to be actually imposed for the purpose of signifying to whatever [interpreters], but another to be actually a sign] (De signis, p. 82; On Signs, ¶1). Bacon’s definition of the sign compares with Peirce’s, which similarly emphasizes relation and actualization: “A sign, or representamen, is something which stands to somebody for something in some respect or capacity” (1940: 99); “A sign is an object which stands for another to some mind” (1991: 141). The perceiver and the interpretive context are both constitutive of actual signification. For Peirce and Bacon, a sign is not a sign in itself. The sign must have some “real connection” with that which it signifies and be interpreted as signifying (Peirce 1991: 141). For Bacon, interactional intentionalities and communicative activity as semiotic relations constitute the sign and ground his functional and pragmatic understanding of language and communication. I understand utterances, whether assertive or non-assertive, not simply by combining the linguistic units of the expression in a particular order, as in the componential or combinatorial model for sentence meaning. Rather, my ability to understand utterances depends on my inferring the utterance’s contextual relevance, intent, purpose, and source, not all of which are necessarily or explicitly articulated in the utterance and not all of which can be assigned to the speaker’s intention. The listener’s cognitive action and interpretive share are crucial to the actual sign function. Furthermore, our semiotic production and understanding of meaning involves the cognitive and interpersonal (inter)actions of speakers and audiences, how the participants identify or share interpretive contexts, relevant presuppositions, and semantic contexts, or if they do. In a given conversation or text, understandings and contexts are shared but also fluid, changeable, and sometimes divergent. That we can predict or feel confident about our understanding of an utterance or text depends on the extent

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

47

to which relevant interpretive contexts are invoked, consented to, and maintained. Whether something actually is a meaningful sign, says Bacon, depends on how the sign-maker and sign-perceiver interact in relation to that which is taken in context to be a sign. Many ancient and medieval discussions of semiotics begin with a definition of the sign, but Bacon actually begins his in strikingly pragmatic fashion by foregrounding the sign function and the signifier’s relation to an interpreter: Signum est in praedicamento relationis et dicitur essentialiter ad illud cui significat, quoniam illud ponit in actu cum ipsum signum sit in actu, et in potentia cum ipsum est in potentia. potentia. Quia nisi posset aliquis concipere per signum, cassum esset et vanum, immo non erit signum, sed maneret signum solum secundum substantiam signi et non esset in ratione signi … si non sit cui actu significetur non est signum in actu sed in potentia tantum. [A sign stands in the category of Relation and is spoken of essentially with respect to that [interpreter] to which it signifies, because it places that [interpreter] in act when the sign itself is in act, and in potentiality when it is in potentiality. Unless someone could conceive through a sign, it would be useless and empty. Indeed, it would not even be a sign, but would remain a sign only in substance and would not have the character of a sign.] (De signis, pp. 81; On signs, ¶1).

Bacon’s phrase “category of Relation” refers implicitly to Aristotle’s Categories (6a-8b), one of the foundational texts of the medieval university Arts curriculum, and to the ongoing debate about relationality in late medieval philosophy and theology. Reversing the prevailing medieval account of Being, Bacon suggests that the sign is 1) accidental rather than substantive and 2) a some-thing which causes or makes it possible for semantic and pragmatic ‘relations’ to emerge in context (cf. Peirce 1991: 141-143). The sign as a cognitive percept emerges from the relation between two or more entities as constructed by a speaker or perceiver/cognizer. Specifically, the sign constitutes a relation between some-thing (word, object, image) and some-thing else (idea, referent, experience, feeling). The relation between the two is cognitive and functional in the context or topic to which the sign is relevant. The sign as sign, that is, as a meaningful signifier, is not the material itself (word, gesture, etc.) but something immaterial which emerges as a relation between a signif ier and a signif ied in the sign-perceiver’s mind. If someone does not perceive a relation between a signifier and a

48 

The Medieval Life of L anguage

signified, intended or not by another, there is no sign. It’s just stuff. This is a fundamentally pragmatic notion of signification. (For a different but compatible account of Bacon on sign and the category of Relation, see Cesalli and Rosier 2018.) Bacon explains signification as something more social and contextualized rather than strictly expressive. In Categories (6a-b) Aristotle argued that “Those things are called relative, which, being either said to be of something or related to something else, are explained by reference to that other thing” (Boethius’ translation: “Ad aliquid vero talia dicuntur, quaecunque hoc ipsum quod sunt aliorum esse dicuntur, vel quomodolibet aliter ad aliud, ut majus id quod est alterius dicitur, aliquo enim majus dicitur, et duplum alterius dicitur hoc ipsum quod est, alicujus enim duplum dicitur.”; LLT-A, 2.216. lines 50-53). Boethius’ commentary on this passage is illuminating (LLT-A, 2.217.lines 15-18). Picking over Aristotle’s language, Boethius foregrounds the category of Relation as entailing plurality (pluralitas). Similarly, in De signis, Bacon situates the sign within the logic of predication (Relation). The sign is at once epistemic, relational, and polylogic. Relationality is the ground of pragmatics as Bacon conceives it. The signuser’s position in time and space is constitutive of the relation between the sign-user, interpreter, proposed meaning, and that which is referred to by an utterance. Referentially, in medieval terms, relations between signifiers and signifieds can be realis (the signified actually exists) or rationis (the signified does not actually exist materially or may have existed in the past but no longer exists). Boethius’ Latin translation of De interpretatione is misleading, but Bacon wouldn’t have known that. Bacon modifies Aristotle’s definition of the sign to account for those signs first present to the senses and those present to the intellect alone, including thoughts, ideas, and feelings: “Signum autem est illud quod oblatum sensui vel intellectui aliquid designat ipsi intellectui, quoniam non omne signum offertur sensui ut vulgata descriptio signi supponit, sed aliquod soli intellectui offertur, testante Aristotele, qui dicit passiones animae esse signa rerum …” [A sign is that which presented to the senses or intellect designates something according to the intellect itself, since not every sign is offered to a sensory faculty, as the common description (vulgata descriptio) would have it. One sort, however, is presented to the intellect alone, as Aristotle [On interpretation, 1.16a4-5] bears witness, who says that the soul’s experiences (passiones) are signs of things …] (De signis, p. 82; On signs, ¶2; cf. Rosier 1994: 114-115, 322n8). By saying signs can have immaterial as well as material referents, Bacon develops a theory of cognition and representation which accounts for the imagination and can be put in the service of theology.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

49

Bacon’s definition of the sign, adapted from Aristotle and Augustine, also takes issue with the dominant linguistic conceptualism in the twelfth and thirteenth centuries. He says, for instance, a loaf of bread displayed in a shop window does not signify material bread (i.e., itself). Rather, it signifies to a competent perceiver the concept or understanding that bread is sold inside the shop. Again, we see how signification is dynamic and relational, depending on the intention, the form of the utterance, and the relevant uptake. Similarly, a circle of vines or a barrel hung outside a shop indicates that wine is sold there. Perceivers make these interpretations by triangulating the signifier and signified with the social context commercium [retail?]. A single instance of material bread signifies “bakery where baked goods are sold”; a single barrel or circle of vines signifies a tavern selling wine (De signis, p. 83; On signs, ¶7). The relation between the loaf (signifier) and the concept bread for sale (signified) is neither iconic nor indexical but symbolic (in Peirce’s terms), “instituted by the intellect” but not completely arbitrary. The loaf of bread in the shop window does have some resemblance or direct relation to the sort of things on offer inside. Bacon emphasizes that immaterial sign referents such as ideas or feelings are mental or abstract but still representational: “… quae passiones sunt habitus ipsi et species rerum existentes apud intellectum, et ideo soli intellectui offeruntur, ita ut repraesentant intellectui ipsas res” [(These) experiences are the very (cognitive and affective) habits and species of things existing within the intellect, and so they are presented to the intellect alone so as to represent things outside [the soul] to the intellect] (De signis, p. 82; On signs, ¶2). By “habits and species of things” Bacon seems to mean something like mental representations or dispositions rather than direct experiences of things in the world (on scholastic habitus, cf. Peter of Spain [below]; Aquinas, ST, 2.1.1, qq. 49-54). According to Bacon, we humans are embodied, language-using, sign-making beings who come to experience, know, and feel the world through signification, linguistic and otherwise. Bacon takes a commonsense and non-conceptualist approach to reference. He claims that signs refer first to the things (concrete or imagined) they designate, not to the individual’s mental concepts of those things. This is a problematic distinction with respect to feelings and ideas but conventional with respect to everyday communication. Bacon regarded concepts as traces transferred from the intellect to memory or implicit within utterances as part of sentence cohesion. To illustrate his realist theory of signification, he uses a favorite example of medieval logicians: Sortes currit and the theory of suppositio (in particular, ‘deliberate (in consilio) equivocation’). If the action running designated by “currit” [X runs] is realistically performed by

50 

The Medieval Life of L anguage

the subject of the sentence, that agent/actant must be the actual (extensive) person referred to by the name “Socrates,” not the mental concept of Socrates which the linguistic name designates (De signis, p. 129). Bacon and other logicians’ arguments on this point raise questions about the conceptualist theory of linguistic reference: What relation, if any, does the mental concept have to the external entity referred to or denoted? and Is the speaker’s mental concept in any way comparable or equivalent to the mental concept inferred by the listener from the utterance and the listener’s posited context? If the majority of words have no resemblance (iconicity) to what they refer to, then aren’t the concepts people produce as thoughts more indirect than we might f irst think? As we shall see, medieval pragmatic theory and metapragmatic practice addressed these problems with the conceptualist theory of signification in interesting, sometimes dissenting ways. These examples of kinds of utterance are standard in philosophical logic. Pragmatic approaches to language use and meaning redeployed medieval logic’s examples. In fact, Bacon’s Sortes (or homo) currit example is one of three linguistic questions which medieval logicians and grammarians after 1150 regularly used to introduce pragmatic topics or perspectives. The other two questions are: “To what does the word canis refer?” and “Is the expression of a relationship, for example, Father/Son, true or meaningful if one part of the relation no longer exists or never existed?” These and similar examples circulated in many post-1150 logical and grammatical texts. Bacon uses all three in De signis. His semiotic analysis and theory of grammar and pragmatics thus participates in a wider scholastic discourse on language and reason. The Sortes currit sentence introduces the question of whether an utterance refers to a verbal, conceptual, or extraverbal referent (three kinds of suppositio). Most medieval analyses of the example sentence adopted a physical if not altogether commonsense approach to the sentence’s reference: the actual person Socrates is running. But individual logicians produced more complex analyses based on supposition theory which explored the implications of agency and the status of verbal signs in utterances. The second question, “To what does the word canis refer?” introduces a different kind of polysemy and context in language and understanding. The canis question was most likely available to medieval logicians as an example of verbal equivocation from Boethius’ De divisione (ed. Magee; 1998: 8-9) and his commentary of Aristotle’s Categories. Depending on the context and speaker/listener relation, Latin canis can refer to a four-legged animal, a sea creature, or a star. Boethius referred to such polysemy as “accidental (a casu) equivocation,” which he restricted to individual words, as opposed to

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

51

ambiguous expressions (De divisione, ed. Magee; 1998: 8-9, 46-47). Both have pragmatic implications. Polysemy and equivocation are always potential in the dialogism of speech interaction and the situatedness of understanding (cf. Voloshinov 1973; Bakhtin 1994: 88-122; Chapter 3 below). Canis and equivocation were linked in logicians’ theory of fallacies. Peter of Spain’s Summulae logicales was a key source for logicians and grammarians when discussing implications of equivocatio with respect to being, signification, and predication. Significatio (sense within the grammatical system) is prior to suppositio (use in particular contexts) in that it is the representation of a some-thing by means of a word in accordance with convention and the grammatical structure (“est rei per vocem secundum placitum repraesentatio”). Suppositio depends on significatio (Summulae logicales, ed. de Rijk, 6.2-3; 1972: 79-80). A term can signify univocally or equivocally. Thus, Peter argues, “Ratio enim entis, secundum quod dicitur de substantia, est ens per se; secundum autem quod dicitur de aliis novem predicamentis, est ens in alio. Et ita predicatur secundum diversas rationes. Et ideo non predicatur univoce, sed potius equivoce aut multipliciter” [The analysis [i.e., understanding] of being (ens) in accordance with which it is said of substance is being itself (ens per se), but [the analysis] in accordance with which it is said of the other nine [Aristotelian] categories is being [occurring] in something else (ens in alio). In this way, being is predicated in accordance with different analyses, and so it is not predicated univocally but rather equivocally or in multiple ways.] (Summulae logicales, ed. de Rijk, 2.20; 1972: 25. My translation.). Peter thus distributes being and semantic meaning across the categories of predication. A word’s potential or equivocal signification is made discrete and meaningful by its use in particular contexts of utterance and understanding. Peter’s example is Latin canis (four-legged animal, sea creature, star [ed. de Rijk, 2.20, 7: 28, 7.30; 1972: 25, 98-100]). Interpretation is an activity, and utterance meanings are understood and constructed in context by “different analyses” (diversas rationes). Peter, like Augustine, usually restricted equivocatio to individual words. But his and similar accounts of equivocatio, significatio, and suppositio contributed to post-1200 theories of language and pragmatics, including Bacon’s. Logicians’ analysis of equivocatio also informed what many intellectuals, for example, Walter Map and Bernard Gui, considered to be equivocal or vague discourse, especially with negative implications (e.g., Peter of Spain, Summulae logicales, ed. de Rijk, 7.26-39; 1972: 98-105). The third question in terministic logic relevant to medieval pragmatic discourse is whether the expression of a relationship is true or possible if one part of the relation never existed or no longer actually exists. The question

52 

The Medieval Life of L anguage

was often posed in medieval logic texts by exploring the relational force in the referential use (supponere) of the words pater and filius: Can someone be called a father if no son in fact currently exists? (e.g., Abbreviatio montana, ed. de Rijk, 1967, vol. 2.2: 93; Peter of Spain, Summulae logicales, ed. de Rijk, 3.17-19; 1972: 34-35]) The terms pater and filius are semantically related in that each establishes the existential conditions for the other. Bacon states the semantic connection between pater and filius pertains more to the category Relation than to that of Substance: “… substantia patris manet quando filius est mortuus et non relatio paternitatis” [… the substance of a father remains even after his child is dead, but not the relationship of paternity] (De signis, p. 81; On signs, ¶1). From the point of view of everyday family relations and lived experience, Bacon’s argument seems a bit hard headed, but he does have a point. With respect to pragmatic-semantic understanding, the meanings of father and son are relational more than componential. The meanings of the individual words are linked, yet changes in the temporal position (father or son has died) or realis (no father or son exists) of the extensive referent will change the implications of the term in some respects. Again, the context of the utterance is constitutive of its meaning. Bacon’s analysis implies that meaning belongs first in the real world. People talk, interact, make meaning, communicate. They agree, argue, and repair their speech. Meanings change over time, or at least they can change. Bacon, Peter of Spain, and other medieval analysts of language considered relationships such as father/son to be referentially real and therefore predicable and able to be signified in so far as the components actually exist. Conceptually, the relation father/son exists. But in particular instances physical circumstances can destabilize that conceptual meaning. The examples of Sortes currit, canis, and pater/filius formed part of a logical and discursive matrix which Bacon and other grammarians and philosophers used for different or alternative pragmatic, semiotic, contextual, or semantic arguments. Bacon adopted the metalanguage and many of the principles and examples found in standard terministic logical texts such as Peter of Spain’s Summulae logicales, but he reread them, uncovering the semiotic potential within terministic logic and turning that discourse in a more pragmatic direction. We can extract from Bacon’s De signis a typology of signs with significant implications for pragmatic thinking: 1 signification by inference vs. signification by similitude (two different kinds of apprehension) 2 “data” [given] or conventional signs produced intentionally or “ad placitum” [according to pleasure, by choice or will] 3 signs generated “naturaliter” [naturally]).

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

53

Bacon rethinks the received classification, most likely from Augustine, of natural and given or conventional signs. Natural signs, he argues, are immediately produced by the intellectus but come about unintentionally due to the state of natura as simply what is without reason or free will. Natural signs are generated “instinctu naturali et impetu naturae” [by natural instinct and the force of nature] (De signis, p. 83; On signs, ¶8). Like the Stoics, Bacon distinguishes natural signs according to iconicity, probability, and causation. Some signs are natural in that they signify something “propter conformitatem et configurationem unius rei ad aliud in partibus et proprietatibus, ut imagines et picturae et similitudines …” [because of the conformity and configuration of one thing to another in parts and features, for example, images and paintings, and likenesses …] (De signis, p. 83; On signs, ¶5). Peirce called this type of sign an icon. Other signs are natural in that they achieve signification through the perceiver’s recognition of causation, “sicut vestigium est signum animalis et fumus est signum ignis” [as a footprint is a sign of an animal and smoke is a sign of fire] (De signis, p. 83; De signis, ¶6). Peirce called this second type of sign an index. Bacon identifies a third kind of natural sign which signifies in relation to the present, past, or future “hoc quod < aliud > necessario vel probabiliter infert” [with respect to consequence and necessary implication]. Examples of this third kind of natural sign, Bacon says, are: for signification in the present “ut in leone et alio est signum fortitudinis, et sic scientis est posse docere” [so a lion signifies strength and so the ability to teach signifies being knowledgeable], for signification based on an inferred past “sic habere lactis copiam ad nutriendum infantem est signum partus in muliere” [so a lactating woman signifies she has recently given birth], and for signification inferring a future “sic aurora est signum ortus solis” [so the coming of dawn signifies sunrise] (De signis, p. 82; On signs, ¶4). Bacon’s third kind of natural signs raises several questions as some of his examples seem to be motivated by causal inferences rooted in science, logic, or cultural symbols. Peirce, for example, might analyze the lactating woman abductively as an indexical (inferential) sign based on the interpreter’s knowledge of human biology. Some of Bacon’s other examples of natural signs seem to depend on inferences or conventions deriving from social norms, stereotypes, or metaphors rather than strictly natural phenomena. The lion as a symbol of strength derives in part from a longstanding cliché. The lion/strength relation is a figure of the sort Peirce might label a symbol, a socially constituted convention of signification rather than a kind of natural relation. Some of Bacon’s natural signs seem to be based on social experience, likelihood, or stereotyping; for example, the positive feeling of being a

54 

The Medieval Life of L anguage

mother signifies love, or excessive decoration in one’s dress signifies pride and lasciviousness, or someone who often wanders around suspiciously at night implies he’s a thief (De signis, p. 82; On signs, ¶4). Such signs are based not on natura as such but on social understanding, inferences, and probable consequences in a contingent if/then relationship. Bacon’s examples of the loving mother and the nighttime wanderer rely on broad social and normative inferences similar to those in contemporary contexts, such as references to so-called “natural” maternal love or the racial or social profiling in law enforcement or the neighbors’ worry that teenagers hanging out on the corner at midnight are up to no good. Bacon’s examples of his third kind of natural signs also complicates how he deploys the concept of natura. For Bacon, natural signs are not willed or imposed by the intellect, at least not completely so. They are perceived as inherently and necessarily part of or caused by natural processes. However, some of his examples derive from inferences based on social understanding, conventions, metaphors, and stereotypes. The question of what is and isn’t a natural sign and how such signs can have pragmatic signification becomes more complex when we consider actual animal vocalizations. In Summulae dialectices (SD), for instance, Bacon calls animal or human vocalizations “natural” behaviors. Commenting on Aristotle (De interpretatione), he argues that “Vox significativa naturaliter est quae ordinatur significandum, ut gemitus infìrmorum et omnis vox ferarum vel sonus” [An utterance signifying naturally is one that is directed by nature to signify, like the groans of the sick and each utterance or sound of wild animals] (SD, ed. Libera, 2.1.25; 1986: 223; Maloney 2009: 60). As with other medieval grammarians and logicians, Bacon’s distinction between linguistic and nonlinguistic vocalizations is important for pragmatic theory and understanding, because vocal signification includes but is not exhausted by grammar. Bacon describes natural signs by deploying natura broadly to include not only what we might readily call natural but also what, as I have argued, are significations based on socially constructed understandings, however much they might be accepted as necessary or automatic (consequences) in human experience (convention, context). Bacon says these natural signs as signs are motivated by cause, likeness, or necessary implication or consequence yet are still produced by the intellect, that is, by inferential and circumstantial, if not always strictly logical, reasoning. Although Bacon wants to maintain natura as being without intention, natural signs establish pragmatic relations of signification even without sign-makers because they emerge through human inferencing and reasoning and because they are

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

55

highly context-dependent. Natural signs are only comprehended as signs within a context which ignites interpretation and human understanding, that is, if someone perceives them as signs. Otherwise, they’re just stuff. Bacon’s typology of signs becomes still more complicated and provocative when he identifies different types of given (data) or conventional signs imposed “ad placitum” or “secundum placitum” [according to our pleasure/ choice/will] (for the translation of “ad placitum,” see Engels 1963). Natural signs “dicuntur, quia ex essentia sua et non ex intentione animae signi rationem recipient” [are called [signs] because they receive the rationale of the sign from their very essence and not from a soul’s intent] (De signis, p. 82; On signs, ¶3. My translation.). Natural signs are signs by physical or material necessity. Given signs are directed by the soul or mind with the faculty of intellect (“signa ordinata ab anima”), a definition which reflects the influence of Aristotle’s De anima on Bacon’s semiotics. Given signs are motivated by human conventions of understanding deployed by sign-makers. Given verbal signs are integrated by humans into a grammatical system through impositio and constructio. Impositio, probably first used by the Latin grammarian Varro (116-27 BCE) to identify how a vox becomes a sign for something else (De lingua latina, 7.32), became a key concept in medieval logic and semiotics. Bacon characterizes linguistic impositio as analogous to baptism. Bacon’s idea of impositio, also referred to as institutio, begins with a speech act (“I baptize you …”). First imposition constitutes a language’s initial (original) semantics and signifying potential, the foundational connection of words and things as linguistic convention. But because language is a social institution, there’s more than extensive reference at stake. Because of his complex definition of natura, Bacon’s typology of signs starts to blur the distinction between animal vocalization in general and human speech. When Bacon connects natural signs and some given or conventional signs with natura, he makes an important semiotic intervention in the medieval conceptualist view of how utterances signify. Can a sign signify without deliberation (intent)? Comparing animal and human vocalizations might answer that question if we know what animals and people ‘are.’ In Summulae dialectices Bacon asks whether animal sounds are natural signs. He suggests that some animal sounds can be communicative within the context of a particular species’ distinct information system. That is, some animal vocalizations can be meaningful in ways which suggest a kind of inherent knowledge and even a kind of intention: “Et hoc possumus videre manifeste, quia gallina aliter garrit cum pullis suis quando invitât eos ad escam, et quando docet eos cavere a milvo” [And we see this clearly because the hen clucks differently to her chicks when she calls them to eat

56 

The Medieval Life of L anguage

and when she teaches them to beware of a hawk] (SD, ed. Libera, 2.1.22; 1986: 222; Maloney 2009: 59). We don’t know how much Bacon based this view on his own or others’ personal observations, but we do know that he was keenly interested in knowledge of the physical world and strongly advocated for empirical evidence as the basis for statements about nature. For humans, Bacon argues some conventional (data) vocalized signs are produced not by deliberation but by force of nature. Sighs, laughs, and screams are semiotically linked to emotions or dispositions in particular contexts. In his sign typology, human sighs, laughs, and groans are “natural” human behaviors which come from the intellectus but are produced without (more or less) deliberation within a discursive context. If we agree such vocalizations make meaning in context, they can be regarded as if they are uttered deliberately and therefore can be considered as if they are language. From a pragmatic perspective, sighs, crying, giggles, and other vocalizations are meaningful or at least interpretable in so far as they are understood to be so by listeners or interpreters who adopt a cognitive frame and construe a situation in which the vocalizations are understood to signify. This is context inferencing and the perceiver’s interpretation of discursive relevance. The notion of pragmatic relevance and the role of context will become increasingly important as we consider various medieval pragmatic and metapragmatic perspectives, both in theoretical and textual contexts. Medieval theorists did not use the term “relevance” as such, but some of their discussions of appropriate meaning and the listener’s role in communication bear comparison with contemporary relevance theory. Relevance is not generalizable. It’s particular, situational, and relational. Sperber and Wilson (1996, 2012) have proposed a theory of pragmatics based on relevance and inferencing as an alternative to Grice’s theory. They argue that the goal of speech interaction is to “enlarge mutual cognitive environments.” Sperber and Wilson can be criticized for relying too much on an informationprocessing model for describing human cognition and for focusing on individual cognitive processes to the exclusion of broad social contexts and cultural factors. Nonetheless, their relevance theory of pragmatics calls attention to the importance of “attention” and the way interactions can be understood as “successful” or “meaningful” even if the participants do not necessarily share some presuppositions or goals, as Grice’s theory requires. Conversations can be understood to be meaningful or productive, even successful, even if participants are not directly cooperating with one another. Like other pragmatics theorists, Sperber and Wilson assume that meaning is underdetermined by linguistic forms. Pragmatic contexts for meaning and communication develop or are confirmed over the time of the

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

57

interaction and are based on participants’ inferences of motives and social situations in addition to their interpretations of utterances themselves. As Sperber and Wilson state, “The cognitive environment of an individual can be modified by providing a single piece of new information, but it can equally well be modified by a diffuse increase in the saliency or plausibility of a whole range of assumptions …” (2012: 87). A particular textual meaning is not always restricted to a single overarching and stable context, as conversation experience and conversation analysis (CA) make clear. Successful relevance inferencing means making space for meaning to emerge or shift, especially if participants do not share presuppositions and expectations. Talk’s fluidity and how relevant talk is produced and understood, the basis for Bakhtin’s “dialogism,” become important for medieval uses of pragmatic ideas and metapragmatic awareness, as we will see later when we look closely at grammatical accounts of interjections and elliptical speech and narratives of heresy examinations. The material description of vox [significant sound] was a staple of Latin grammarians and medieval philosophers, although different writers used different descriptors: “Vox est percussio aeris per linguam, que per quasdam partes gutteris (que arterie vocantur) ab animali profertur” (e.g., Boethius, In librum Aristotelis Peri hermeneias commentarii, secundo editio, 1; LLT-A, 1.4.line 18). Priscian argued that words constitute either vox articulata or vox litterata, speech or writing (speech’s other) (Institutiones, 1.1; KGL 2:5; cf. Kelly 2002: 11-17). But neither Boethius nor Priscian nor the modistae gave much attention to animal sounds, sighs, and laughs and definitely excluded them from consideration as language. Bacon, however, construes both words and sighs as meaningful vocalizations (vox) produced by living creatures, but they have different intentional potentials. Materially, both words and sighs are “percussio aeris per linguam.” However, Bacon’s sign typology and his theory of natural signs complicates the grammatical definition of vox. As Stephan Meier-Oeser points out, Bacon’s “reason for distinguishing two modes of natural signifying … is, on the one hand, an equivocation of the concept of nature, meaning ‘substance or essence of something’ (substantia sive essentia cuiuslibet), as well as ‘force acting without deliberation’ (virtus agens sine deliberatione) (De signis, 1978: 85f.) and, on the other hand, the insight that, contrary to what holds for the natural signs in the first sense, in the case of the latter [data, given signs] there is always a sign-giver, not only someone taking something as a sign” (2011). Bacon implies we interpret groans, sighs, and laughs as if they are language or words, so we seek to understand them as we understand utterances.

58 

The Medieval Life of L anguage

When Bacon characterizes some natural signs as being produced by “virtus agens sine deliberatione,” he gives a pragmatic slant to paralinguistic vocalizations. Taking virtus in the context of virtus sermonis, we can read Bacon to be redeploying a theoretical concept important in exegesis for more general semiotic, semantic, grammatical, and pragmatic purposes. The phrase virtus sermonis, adapted from Aristotle via Isidore (“vis verbi vel nominis”; Etymologiae, 1911: 1.29.1), was typically used to appeal to the ‘natural,’ ‘original,’ or etymological meanings of words or to establish the meaning of an utterance based on linguistic form and given semantics alone (autonomous language), apart from any additional context which might alter the meaning of the words (cf. Courtenay 1984: esp. 112-116; Goubier and Pouscoulous 2011; also below, pp. 69-70, 141, 165). In rhetoric and grammar, virtus sermonis was sometimes contrasted with usus loquendi, which implies both a greater degree of individual semantic reassignment and also the possible expansion (amplificatio) or restriction (restrictio) of meaning according to the speaker’s will in specific speaking contexts. Virtus sermonis implicitly acknowledges the equivocation inherent in linguistic and semantic structure and thus the importance of managing the interaction of linguistic forms and meanings and internal with extraverbal contexts and audience inferences. The concept became especially controversial in debates over theological discourse and logic. In 1340 the Arts Faculty of the University of Paris condemned the so-called Ockhamist principle that propositions from authoritative sources could be considered false with respect to virtus sermonis (Courtenay 1984). Bacon’s use of the concept of virtus sermonis relies on a theory of language rather than on an institutional policy of authoritative discourse. Latin virtus principally means ‘force, power,’ a some-thing necessarily inherent in whatever exerts the force, or perhaps a force of nature but without deliberate intent. Nature is a virtus (OF and ME virtu/vertu), a life force which exerts power and energy within matter to bring about life. That is how Chaucer uses the word at the beginning of the Canterbury Tales: “Of which vertu engendred is the flour” (General Prologue; Canterbury Tales, 1.4). Words’ grammatical structure and conventional semantics generate meaningful discourse within a speech community, but words’ ‘potency’ and humans’ will mean that speech can and does exceed conventional or formulaic usage. Certain kinds of vocalizations or utterances, especially interjections and meaningful giggles or sighs, prompted Bacon and other twelfth- and thirteenth-century grammarians to rethink the nature-convention distinction received from traditional Latin grammar as well as the clear distinction between conceptus and affectus as modes of understanding.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

59

As I said, Bacon’s sign typology also complicates the theory of given (data) verbal and nonverbal signs. Interjections are a liminal case in grammar because they partake of both natural and conventional elements. Given signs always entail a sign-giver, but some, like interjections, are not always fully controllable or deliberately used in discourse by the utterer. In addition, Bacon introduces into his theory of linguistic semiosis an openness which acknowledges the difference between speaker and listener and therefore a potential gap between a speaker’s intention and the utterance itself and between utterance meaning and a listener’s intention and understanding of the speaker’s intent/meaning. Pragmatic theory and practice are spread over that gap. The dialogism of speech is activated, or potentially activated, by that gap. Approaching the Other in a speech context opens the encounter to the possibility of multiple intentions and responses, not just the speaker’s, and to epistemic contexts which effect understanding, not all of which are given or known in advance. Although most natural signs do not entail a sign-giver, they still signify within interpretive, intellectual, and communicative contexts, but with a stronger sense of necessity or just there-ness because, well, that’s how things are. In Bacon’s semiotic theory, things, behaviors, and events are not significant in themselves but become signifiers (sign-vehicles) when they are perceived and understood as having or making meaning. Bacon’s view is similar to Peter Helias’ distinction in his Summa between congrua voce and congrua sensu. A comprehensible utterance as sign requires inferencing, contextualization, and uptake, not just an intending speaker. Nothing is a sign outside use and context of situation. In fact, Bacon implies, anything and everything can be a sign if used or taken to signify in some way to someone. Recall Peirce’s similar definition of a sign or representamen: “something which stands to somebody for something in some respect or capacity” (1940: 99). Signification is relational and contingent. Bacon’s semiotic theory sets out a partial vocabulary for medieval pragmatics in terms of intention with degrees of deliberateness and relational meaning. Recall his examples of the bread loaf and the circle of vines and the barrel hoop hung outside a shop. The loaf, circle of vines, and barrel hoop become indexical signifiers when situated in a space and embedded in social practice. Objects signify rather than serve their original function because they are made to do so by intenders and perceivers relying on social conventions, audience expectations, and several layers of social and discursive praxis which make such correlations habitual: taverns are where wine is served, people want wine; coopers make barrels, people fill them; bakers bake bread, people want to eat; commodity trade relates sellers and

60 

The Medieval Life of L anguage

consumers. Again: “Unless someone could conceive through a sign, it would be useless and empty. Indeed, it would not even be a sign” (De signis, p. 81; On signs, ¶1). When the vines are intentionally formed into a circle to make a tavern sign, the social context in which they acquire meaning as a shop sign precedes and exceeds the sign reader. The sign maker and sign reader both participate in signification under modus habitus, a regular, repeated way of doing something physical and/or mental in relation to a particular thing in a social context. Habitus names how life is lived. Medieval philosophers and semioticians borrowed the idea of habitus from classical philosophy. Cicero was a common source. Habitus, he says, is “not given by nature but acquired by effort and industry,” that is, culturally learned yet conveniently regarded as almost transcendent and permanent (De inventione, 1.25.36). Aquinas used habitus more comprehensively as “not the terminal boundary of a power, but the disposition of a power to an act as to its ultimate term” (ST, 2.1.54 Art. 1, Reply Obj. 3). (The genealogy of habitus as systems of durable vs. changeable dispositions is yet to be worked through. In the meantime, see Sapiro 2015.) It’s debatable whether Bacon would agree with Grice that ‘rational’ discourse always depends on participants adopting the Cooperative Principle and related Maxims. Bacon’s notion of intention is broadly semiotic and pragmatic. He posits the perceiver as an equally intentional agent in a context of explicit, natural, and habitual intentions and acts. The possibility of communication through signs, especially language, depends on speaker and listener using the same or similar habits of understanding. But as we shall see, not all pragmatic interactions are founded necessarily on cooperation or a single speaker’s intention. From a semantic and pragmatic perspective, Bacon’s account of reference and the contexts and circumstances for signifying entails more than material or physical conditions: Non enim sequitur: ‘Signum in actu est, ergo res significata est,’ quia non entia possunt significari per voces sicut et entia, nisi velimus dicere quod esse quod necessario requiritur ad significatum non est nisi apud intellectum et imaginationem. [For ‘A sign actually exists; therefore a signified thing exists’ does not follow [or, is not a valid inference], because nonexistents [non entia] can be signified by vocal sounds just as beings [entia, existents] can be, unless we do not wish to say that Being/Essence necessarily required for signification does not exist in what is understood (i.e., mental represented) in the intellect and the imagination] (De signis, p. 82. My translation.).

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

61

Bacon uses the compelling example I mentioned earlier of the father/son semantic relation to illustrate how verbal signification is relational, whether conceptually or in actuality: “sicut substantia patris manet quando filius est mortuus et non relatio paternitatis” [… so the substance of a father remains even after his child is dead, but not the relationship of paternity] (De signis. p. 81; On signs, ¶1). For Bacon, semiotic relations are circumstantial, conventional, and palpable. The father/son example brings home in a powerful way Bacon’s concept of the sign as first of all a relational and mental phenomenon and then as a vocalization or physical entity. The example might seem incoherent or contradictory with respect to the definition of material and immaterial existents, but Bacon is doing some subtle reasoning here. In Bacon’s semiotic, substance is prior to relation which is prior to all other categories of being in scholastic philosophy (De signis, pp. 107-108). Bacon wants to maintain a realist ontology for signification, but he also wants to give real status to mental objects and experiences so as to account for feelings, imagined things, mental representations, and abstract ideas. Moreover, as the emotionally harsh example of the father and his dead son suggests, Bacon argues that concrete significations of terms (word-thing relations) are neither fixed nor absolute nor eternal. They can and do change over time and/or space as the circumstances and contexts of signifying relations and as participants’ aims, power, knowledge, and agency change. I can still refer to the person known as Socrates by name even though he is dead, but as a speaker and knower I cannot be in the same existential and signifying relation if that physical person no longer exists as when I say “Socrates” and intentionally refer to someone who exists. A speaking subject does not always inhabit the same position with respect to existing physical and mental things. But then neither does a listener. This slippage between realist material and realist immaterial significations has important implications for medieval pragmatics, especially with regard to emotional expressions, metaphor, and imaginary concepts. In classical and medieval grammar, the structuring of words as significatio was conceived as a process from impositio to constructio. In the act of imposition, speakers assign meaning to a vocal unit (word, phrase) and turn a purely vocal sequence into dictio (meaningful speech). This linguistic transformation of sound into meaningful or signifying sound derives from Bacon’s insistence that the listener must understand or take a vocalization as meaningful sound for there to be a sign. Bacon’s account of impositio lets us distinguish between vocalization in general, which applies to humans and animals, and speech (vox articulata), which applies to humans. However, the distinction becomes less clear when we consider certain kinds

62 

The Medieval Life of L anguage

of expulsions of air (e.g., moans, giggles, interjections) which may or may not be considered as utterances. In pragmatic theories of language, meaningful human vocalizations include more than ordinary or conventional words. The second structuring of words is constructio, the ordering of words in a clause or sentence according to grammatical accidents (case, tense, plural, etc.). Constructio is defined formally as congrua voce [congruent speech] and in a more semantic sense as congrua sensu (e.g., Priscian Institutiones 2.15, KGL 2:53; Helias, Summa super Priscianum, 1993: 835; cf. Kelly 2002: 165-69). Ultimately, congruitas [agreement] is implicated in both significatio and consignificatio. Conventional significations are produced according to the play of the mind and social habitus. In the twelfth and thirteenth centuries grammarians hotly debated whether congruitas as well-ordered speech was strictly a formal matter or whether a construction’s “correctness” also depended on other modes of signifying and the content of the utterance. Bacon clearly argued for the latter side of the question (cf. Amsler 2011: 70-80). The concept of impositio was crucial to post-1050 philosophical analyses of meaning. Sometimes, it was also inflected by a pragmatic approach to language and signification. Bacon’s semiotics of language is one such inflection. The term and concept were adapted from ancient and Late Latin grammar and philosophy, especially via Boethius’ translations of and commentaries on Aristotle’s Categories and De interpretatione. This again highlights the relations between medieval pragmatics and the logic curriculum. Boethius’ logical writings and Aristotelian commentaries were important resources for medieval logicians and core texts in the developing scholastic university curriculum. (For an overview of Aristotelian translations and commentary traditions in the later Middle Ages, see The Cambridge History of Later Medieval Philosophy [1982].) Boethius used the concept of impositio (his term is positio, a Latin calque from Greek) as a near synonym for identifying those signs as given (data), arbitrary, or established “by convention” (ex consuetudine) as opposed to “natural” signs. Impositio is the principal act which establishes the relation between words and things (semantics) and then between words and words inflected by accidents (grammar). Imposition is a double process. Linguistic signification is grounded in both reason and convention. In his second commentary on the Categories (c. 509-511 CE), Boethius outlines the two modes of imposition and signification: Prima igitur illa fuit nominum positio, per quam vel intellectui subjecta vel sensibus designaret. Secunda consideratio, qua singulas proprietates nominum figuras que perspicerent, ita ut primum nomen sit ipsum rei

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

63

vocabulum: ut, verbi gratia, cum quaelibet res homo dicatur. Quod autem ipsum vocabulum, id est homo, nomen vocatur, non ad significationem nominis ipsius refertur, sed ad figuram, idcirco quod possit casibus inflecti. Ergo prima positio nominis secundum significationem vocabuli facta est, secunda vero secundum figuram: et est prima positio, ut nomina rebus imponerentur, secunda vero ut aliis nominibus ipsa nomina designarentur. Nam cum homo vocabulum sit subjectae substantiae, id quod dicitur homo, nomen est hominis, quod ipsius nominis appellatio est. Dicimus enim, Quale vocabulum est homo? et proprie respondetur, nomen. [Therefore, that was the first imposition of nouns/names, by which [the name-givers] denote/designate things subjected (subjecta) to the intellect or the senses. The second consideration [took place] when they [the name-givers] noticed the formal properties of nouns/names, so that the name first [imposed or given] is truly (ipsum) the word for the thing; as, for example, whatever thing is called homo. And so the word itself, that is, homo, is called a noun, [and] it [i.e., the word class noun] does not refer to the meaning of that name but to its form, because it can be inflected for cases. Therefore, the first imposition of the name/noun is made according to the [lexical] meaning of the word, and the second [is made according to] the form; and the first imposition is the imposition of names/nouns on things, and the second [imposition is] when those names/nouns are denoted by other names/nouns. For when homo is the word (vocabulum) for a thing, that which is called homo, the name/noun pertains to a person, because it is the appellative form of that name/noun. So we say, What sort of word is ‘homo’? and the correct response is, Noun.] (In Categorias Aristotelis libri IV 1; LLT-A, 1.159.lines 8-13. My translation.)

Medieval grammarians and logicians also referred to first imposition as significandi and second imposition as consignificandi. By the thirteenth century, Boethius’ “appellatio” was usually replaced by suppositio. In his De divisione (c. 505-509), Boethius demonstrates how to use differentiae for logical reasoning by analyzing first and second imposition as part of a ‘correct’ philosophical definition of the term nomen (translating Aristotle’s onoma). A common name, he says, is “‘vocum significativarum secundum positionem sine tempore aliae sunt quarum pars extra aliquid significat,’ hoc pertinet ad orationem, ‘aliae quarum pars extra nihil significat,’ hoc pertinet ad nomen, nominis enim pars nihil extra designat” [“Of spoken sounds significative by imposition (secundum [im]positionem), without tense, some have a part that independently signifies something,” [and] that pertains to an expression (oratio); “others have a part that independently signifies nothing,”

64 

The Medieval Life of L anguage

[and] that pertains to the [form of the] noun]. In contemporary linguistics, this is the distinction between a word and its grammatical forms or between a word and a non-word syllable. A few lines later, Boethius rephrases the definition and expands the notion of imposition: “nomen est vox significativa secundum placitum sine tempore, cuius nulla pars extra significativa est separata” [a noun is a spoken sound significative according to human choice [and] without tense, no part of which when separated is significative of anything else] (De divisione, ed. Magee; 1989: 36-37. My translation.). First impositio implicates human reason, deliberation, and arbitrary choice as the conditions for establishing the relation between language and meaning. Second impositio assigns words as signifying their accidental properties (inflections, derivations, etc.) and as combining with other words in syntax (anaphora, agreement, etc.). Second imposition generates syntax, cohesion, linguistic self-reference, and ultimately metalanguage. Boethius excludes other kinds of natural vocalizations (laughs, grunts, interjections) as not being linguistic expressions (nomina); De divisione, ed. Magee; 1989: 34-35). In thirteenth-century universities, the canonical logic texts by Aristotle, Porphyry, and Boethius and more recent texts by Peter of Spain and Lambert of Ligny-le-Chatel provided the terms, concepts, analyses, and examples for logic, predication, and argument. But when read with a pragmatics or semiotic orientation, as Bacon does, those texts afforded different questions and analyses from those posed by Aristotelian or terministic logic. In the early twelfth century, some thinkers, notably Abelard, had already proposed different approaches to signif ication and language. In Paris, philosophers and grammarians began to critically evaluate the properties of terms (logica modernorum) using more complex grammatical analysis while maintaining a focus on the proposition or clause as the primary unit for analysis. The influential twelfth-century Abbreviatio montana distinguished between “vox … significativa naturaliter” [vocalization … naturally significant] signifying “naturales animalium affectus” [natural animal dispositions] (e.g., mooing, barking, person groaning) and “vox … significativa ad placitum” [vocalization … significative by convention (‘by choice’)], that is, speech. The latter “secundum rationabilem institutionem hominum aliquid significat, ut ‘homo’” [signifies something in accordance with what human beings have rationally established [by imposition], for example, [the noun] homo]. According to the anonymous author, “oratio … perfecta … perfectum sensum generat in animo auditoris, ut ‘Socrates legit’” [a complete (perfecta) utterance/clause generates a complete meaning (proposition) in the listener’s mind]. An “oratio … inperfecta” does not generate a proposition, so the Abbrevatio writer effectively sets aside incomplete

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

65

utterances, including grammatical phrases (homo albus), as not logically relevant (ed. de Rijk, 1967, vol. 2.2: 78-79. My translation.) But if terministic logic is concerned with natural language, this manoeuver is suspect. The next generation of philosophers and grammarians, reflecting on these gaps in earlier logical texts, explored the status and signification of incomplete or elliptical utterances and the relation between natural and conventional vocalizations. Despite terministic logic’s focus on identifying and correcting fallacies in argument, some grammarians and philosophers discovered more positive approaches to various forms of utterance, not just so-called complete propositions. Interjections, adverbs, participial constructions, cries, and other kinds of expressions instigated new, alternative grammatical and pragmatic analyses. The concept of impositio was also scrutinized with new theoretical frameworks. Some logicians took a new approach, not always pragmatic, to impositio as the establishment (institutio) of signification and as the signifying mode by which meanings are associated with words or words with things and the mode by which verbal accidents are established as grammatical conventions (agreement, idioms) (e.g., Peter of Spain, Summulae logicales, ed. de Rijk, 6.2-3; 1972: 79-80). Following Boethius, twelfth- and early thirteenth-century logicians such as Peter of Spain and grammarians attended to first imposition, convention, use, and accidental signification when analyzing how utterances are intended or received as meaningful. They reemphasized impositio as grounded in human acts of will (voluntas) or deliberation, reason (ratio), and intent (ad placitum). Lambert of Lignyle-Chastel’s definition in his Logica is typical of this view of language as a human-made product: “… et licet sit naturalis et intellectus rei et vox similiter, quia a naturalibus principiis formatur, tamen quia a voluntate fit unio intellectus rei cum voce, et in hoc consistit vocis impositio, ideo vox dicitur significare ad voluntatem instituentis” [… although both a concept of a thing (intellectus rei) and a vocal sound are similarly natural because they are formed by natural principles, a vocal sound is nevertheless said to signify according to the will (ad voluntatem) of the one instituting it, because the union of the concept of the thing with the vocal sound comes from the will, and it is in this [action] that the imposition of a vocal sound consists] (Logica, ed. Alessio, 1971: 205.26-29; Logica, trans. Maloney, 2015: ⁋1235). Lambert again emphasizes the role of will (choice) and reason in the act of imposition when he discounts natural sounds (presumably groans, sighs, and so forth) as relevant to logic: “Sonus sic dividitur. Sonorum alius vox, alius non vox. De sono autem non voce nihil ad logicum” [Sound is divided in this way: one sort of sound is vocal sound (vox), the other is nonvocal

66 

The Medieval Life of L anguage

sound. Nonvocal sound is of no interest to a logician.] (Logica, ed. Alessio, 1971: 6.34-35; Logica, trans. Maloney, 2015: ⁋30). Even though Peter of Spain argues, like Bacon, that natural hum-animal sounds (moaning, barking, groaning) signify, he agrees with Lambert that such vocalizations are not relevant to logic (Summulae logicales, ed. de Rijk, 1972: 1-2). Lambert goes on to distinguish two senses of will: as something free (“libera”) and changeable (“transmutabile”) or as something “recta ratione considerata” [considered by right reason] and “non est principium transmutabile” [not especially changeable]. Human will is relevant to the imposition of sound on concepts only in the second sense, reasoned choice, which creates vox significativa (“et hoc modo voluntas est principium imponendi voces ad significandum, et non primo modo”) (Logica, ed. Alessio, 1971: 7.34-8.2. My translations.). Peter of Spain argues that “vox significativa ad placitum” is produced by first imposition (Summulae logicales, ed. de Rijk, 1972: 2). Second impositio, he says, again following Boethius, is constituted not by will, ratio, or choice but by the elements of a word’s composition, as for example when the preposition in is added to mortalis to derive immortalis, negating the root the bound morpheme is attached to (Summulae logicales, ed. de Rijk, 7.53; 1972: 111). For Lambert, Peter of Spain, and many other logicians, what makes air under pressure (percussio aeris per linguam) become cognitively and affectively meaningful is human will and the power of speakers and listeners to create verbal signification with concepts. Bacon took the concept of impositio further in a pragmatic direction by distinguishing between primary and secondary imposition in a different way. Impositio, he said, is the renewable creative act central to the life of language. Double imposition (semantic and grammatical) is repeated informally, tacitly, as ongoing reimposition. Bacon’s theory of reimposition and signification depends implicitly on a theory of relevance. By describing imposition as open-ended, Bacon creates a complex framework to account for individual intentions and microsocial or altered circumstances for reference and meaning in different contexts, even within a single conversational interaction. For instance, the name Socrates has one relevance context and referent when Socrates is alive and another when he is dead (cf. Cesalli and Rosier-Catach 2018). This analysis picks up the temporal logic embedded in terministic logic (cf. Uckelman 2013) while opening a much wider context for signification and social interpretation. Bacon’s account illustrates how speech works as contextualized and symbolic social interaction. Moreover, as Kelly points out, Bacon’s theory of ongoing impositio “plays up salient features of the new significatum by juxtaposing it with the old. Thus, modus significandi ut is the familiar rhetorical figure of translatio” (Kelly 2002: 26;

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

67

cf. Rosier 1994: 135). Bacon’s pragmatic orientation to speech and meaning foregrounds the dialogism and dialectics of speech as situated social interaction, speech as meaningful within one or more relevant contexts of situation (cf. Halliday 1978: 28-31; Chapter 3 below). One implication of Bacon’s theory is that if a word’s signification depends on a relevant context and if reimposition is how meaningful language is created and maintained, then it’s possible for a word to lose signification at some point or to acquire new signification with a new audience at a different time. Bacon characterizes ongoing impositio as the creative force in linguistic action. Imposition is so fundamentally a part of everyday usage, so much a part of the language-making habitus, that we are scarcely aware of how often we impose and reimpose meanings on sounds to maintain or renovate discourse. Bacon links the question of reimposition to thinking differently about equivocatio as a positive creative aspect of language use: “Similiter quo occultatur ista aequivocatio sicut multiplex alia est propter ignorantiam modorum diversorum imponendi quoniam nos tota die imponimus nomina et non avertimus quando nec quomodo …” [Likewise, why this equivocation, like other sorts of ambiguity, is concealed from us is explained by [our] ignorance of the different ways of imposing, since we impose names all day long and pay no attention to when or how …] (De signis, p. 100. My translation. Cf. De signis, pp. 131-132.) Equivocation, ambiguity, metaphor, and other figurative language were and are key topics in the philosophy of language. But the status of equivocation in theory and practice is controversial, and Bacon’s theory moves against the logical mainstream, medieval and modern, which seeks propositional disambiguation above everything. Augustine in De dialectica (Chapter Ten) itemized the semantic and discursive implications of equivocatio: from art (ars), from use (usus), and from art and use together. A word is equivocal from ars if it belongs to more than one descriptive linguistic category. Tullius, for example, is a noun or a dactyl, depending on the descriptive frame. A word is equivocal from usus if in a given context it can signify more than one thing (recall medieval logicians’ canis example). Augustine’s example is the name Tullius, which can refer to the orator, his picture or statue, his collected writings, or his corpse (compare the contemporary English “Have you read your Shakespeare today?”). Equivocation destabilizes relevant context of situation and discourse. In some contexts, said Augustine, equivocation can border on lying, when the speaker deliberately intends to say what is not true in order to deceive others (De mendacio, 3.3; Contra mendacium, 10.24). Twelfth- and thirteenth-century logic texts described kinds of equivocatio as part of their theory of fallacies in argument (de Rijk, 1962: 49-71).

68 

The Medieval Life of L anguage

Such instabilities of meaning are inherent in the semantic network of a language. Contexts and intentions can be used to disambiguate meanings or, alternately, to activate potential meanings beyond immediate or conventional denotation or community norms. For many philosophers, logical analysis aims to disambiguate terms in utterances for clearer, more transparent meaning authorized by the speaker’s intention. Say what you mean, the advice goes. Bacon, however, regards impositio and equivocation more positively and productively. He gives several reasons for why we continue to tacitly reimpose meanings in the context of both special and everyday speech: needing or wanting to express our meaning by coining a new word (neologism), using a word differently or in nonconventional fashion (literally or figuratively, to make an analogy or metaphor, in contrast to conventional use or genre expectation), or creating a new word by modifying an existing word and giving it a new meaning (e.g., deriving a verb from a noun or switching word class through usage). Bacon says literal, univocal reference is only one kind of language use and not the benchmark against which all kinds of use should be evaluated. Helias and Kilwardby, among others, also identified figurative uses of language as a special case of meaningful speech via congrua sensu without grammatical correctness. Bacon’s theory of impositio and will revises key aspects of received semiotic theory in the thirteenth century and established a strong link between theory and practice in medieval pragmatics. What Bacon calls secondary imposition, tacit imposition or re-imposition, covers a great deal of conceptual territory, especially how a word’s meaning gets shifted by translation, analogy, equivocation, context, ambiguity, innovation, poetic creativity, discursive negotiation, and reasoned argument. In fact, Bacon’s distinction between primary and secondary imposition, set against a ground of accepted use and ‘given’ (data) semantics and grammar, opens up a large and fruitful field for pragmatic theory and discourse and for recognizing the fluidity of linguistic activity. Language as used does many things: speech sustains and creates identities, creates social relations, produces new knowledge through new language use or new connections with given knowledge, reworks or violates (flouts) signifying conventions, especially linguistic ones, to create novel understandings or challenge existing or dominant ones, represents what and how we perceive the world as the world. Imposition and re-imposition constitute a pragmatics of maintaining but also challenging established social order and knowledge discourse through intentional or ad hoc linguistic praxis. We will explore how Bacon’s semiotic theory, while not explicitly cited by other writers, is reflected in the pragmatics of dissent and the dialogism of speech in later medieval texts.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

69

In addition to reorienting the logical topic of equivocation toward more creative and affirming uses, Bacon’s theory of re-imposition is bound up with analogy, equivocatio, ampliatio, and restrictio. As modes of meaning making, imposition, equivocation, ampliation, and restriction were major topics of discussion and debate among thirteenth-century logicians but less so among grammarians, except for impositio. Robert Kilwardby and Bacon are exceptions (cf. Rosier 1994: 96-100). These linguistic meaning-making activities entail the decisions or will of individual speakers or speech communities. Regarding the extension of verbal meaning by analogy, Bacon explicitly agrees with Avicenna (c. 980-1037; Logic and Metaphysics) and Algazel (c. 1058-1111; Logics) and ultimately with Aristotle (Metaphysics): “quod voces significant multa et quasi infinita per hanc viam” [that vocal sounds signify many things and along these lines an almost unlimited number of things] (De signis, p. 116; On signs, ¶103). The open-endedness of Bacon’s theory of impositio gives it pragmatic potency. Scope and analogy are productive tools for thinking and speaking, open-ended, flexible, and dependent on the active perceptions, conceptualizations, and contextualizations of speakers and listeners. They are also key to ampliation and reimposition. In logic and law, equivocation has been (and still is) often construed as distracting ambiguity, imprecise speaking, or outright evasion. Bacon, however, describes equivocation as a fundamental linguistic condition which enables people to use language for different purposes and intentions because meaning and forms are unstable. In Bacon’s semiotic theory, equivocations in language exist on a continuum from absolute difference between meanings (complete difference between two or more meanings assigned to a vocal sound; e.g., canis to denote an animal, sea creature, or star, or the name Socrates to signify a living person or a nonexisting person or a type of name) to those uses of a root word which agree absolutely but differ with respect to a mode of signifying or property (e.g., legere [to read], legens [reading] (participle), lectio [a reading] (noun)). Absolute difference relates mostly to lexical semantics, while modal difference relates to grammatical constructions. Both contribute to constructing discursive meaning in particular utterances. Earlier, we discussed the example of canis, but Bacon notes that many equivocations result from words with related meanings. For instance, the meaning of sana [healthy] shifts depending on how the word is linked with a subject noun when used in a predicate: e.g., “Urina est sana” [Urine is healthy], “Medicina est sana” [Medicine is healthy], “Dieta est sana” [Eating well is healthy] (De signis, p. 84; On signs, ¶40). Bacon takes this example from Peter of Spain (Summulae logicales, ed. de Rijk, 7.28; 1972: 98-99), who takes it from Boethius’ notion of “deliberate (a consilio) equivocation” in his commentary on Aristotle’s

70 

The Medieval Life of L anguage

Categories. Each expression signifies a kind of “healthiness” (sanitas), but the semantic modality shifts from indicating healthiness (urine) to bringing about healthiness (medicine) to maintaining healthiness (eating well). In different verbal contexts, sana signifies a means, an end, or a state. Many medieval logicians considered such potential multiple (equivocal) meanings for a given term as leading to fallacies in argument. Similarly, but more caustically, inquisitors argued that speakers who appeared to speak equivocally were deceptive and insincere (see Chapters 5 and 6 below). From a pragmatic perspective, however, depending on the speaker’s intention, equivocation in utterances can be interpreted variously as a discursive strategy, a problem, or a motivation for recursion, repair, semantic clarification, or deictic marking. It all depends on intention, uptake, and most important, context. As manipulations of semantic scope, ampliation and restriction were interesting topics for medieval logicians, but Bacon considered them key for re-imposition. Logicians discussed ampliation and restriction in terms of broadening or narrowing the scope or semantic valence of a subject term in a proposition, usually by means of a predicate, or vice versa. Bacon mostly agrees with logicians’ account of ampliation and restriction in predicate logic as received from Aristotle and Boethius, but with one major addition, Will (voluntas): Unde causa efficiens huius ampliationis et restrictionis est anima rationalis, quando ex bene placito suo fiunt, et potest, si vult et quando vult, facere eas sive cum praedicato vel subiecto tali quod vocatur ampliativum vel restrictivum, sine eis, … et in veritate facimus hoc, quia sine talibus praedicatis vel subiectis transsumamus nomina entium ad non entia et ad species vel similitudines rerum et ad ipsas voces et ad omnia quae volumus. [So the efficient cause of these ampliations and restrictions is the rational soul, since they are constructed as one likes (ex bene placito suo fiunt), and it is possible, if and when one wants to, to construct these [words] whether by predicating or subjecting, such that it [the word] is called an ampliative or a restrictive, or [to do so] without them. … In truth we do this because even without such predicates or subjects we transfer the name of beings to nonbeings and to the species of likenesses or things and to vocal sounds themselves and to all the things we want to.] (De signis, pp. 114-115. My translation.)

Bacon’s argument here is not some kind of Pollyanna pragmatics but a strong assertion of human free will (voluntas, ex bene placito) linked to reason (ratio) as the creative force for acts of making language and meaning. Bacon

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

71

then clarifies the pragmatics and situational context of re-imposition: “Et ideo quod amplient et restringant hoc non est formaliter vel effective nec virtualiter loquendo sed per concomitantiam et occasionaliter, et quantum ad evidentiam” [So, [to say] they ampliate and restrict is not to be speaking formally or causally (effective) or virtually, but rather by concordance and because of the situation (occasionaliter) and with respect to the evidence] (De signis, p. 115; On signs, ¶99). Language and meanings are situationally dynamic, fluid, even plastic, creative and ever moving toward an open horizon of meaningful speech. People use language intentionally to communicate and respond to given circumstances or experiences, to create or change circumstances and understandings, to imagine alternatives, or to engage one another’s attention and so forth, as the occasion and our will prompt us to do. The word domus might be used to refer to, i.e., be intentionally imposed on, the plan for a house or the completed (vera) house or both at a more abstract level, depending on the speaker’s intention and the discursive context of the utterance (De signis, p. 109; cf. De signis, p. 132). Nothing about the vocal sound requires it to be imposed on either referent, only the intention and the understanding of the speaker. Complementarily, the listener’s understanding actualizes the sign relation as vox articulata at the moment of speaking or reading. Implicit in Bacon’s analysis is the notion of double articulation, whereby the speaker’s “intention” to produce a series of sounds (vocalization) is merged with the “intention” to signify something or perform something vocally, on the one hand, and on the other, the listener’s “intention” to construe meaning by means of the perceived vocalization in terms of common understanding or individual inferencing. Imposition and equivocation lead Bacon to rethink the status of mental representations and concepts we form from vocal utterances. Toward the end of De signis (pp. 134-135) Bacon reinterprets the key grammatical term conceptus in a more overt pragmatic manner, emphasizing that the act of interpretation is constitutive of signification. Recall that Bacon rejected the consensus (Aristotelian) view that spoken words refer first to concepts in the mind which are likenesses of things in the world or imagination. Priscian stated that a part of speech signifies a concept of the mind (mentis conceptum) according to the pleasure (will) of the listener (Institutiones 11.2; KGL 2.552). Bacon rereads Priscian’s mentis conceptum as not a noun phrase with conceptum but a participle phrase with the past participle of concipio/-ere. Bacon redirects Priscian’s authoritative text to support a more pragmatic but nonetheless realist argument: speech signifies not the species of things but “res concepta et intellecta in quantum huiusmodi” [things conceived and intellected [in the mind] in some way or another] (De signis, p. 135. My translation.).

72 

The Medieval Life of L anguage

Ampliation, re-imposition, analogy, and other procedures for adjusting or creating meaning depend on speakers’ and interpreters’ decisions, and Bacon offers a capacious and active account of the will (voluntas, ex bene placito): “… voluntas superat naturam in hoc quod est ad opposita et diversa. Et quod natura facit uno modo, voluntas facit pluribus, et ideo per libertatem voluntatis potest nunc vox sibi nunc alii imponi” [… will prevails over nature with respect to what is opposite and different, and what nature does in one way will does in many. So, through the freedom of the will a vocal sound can be imposed now for itself and at another time for something else] (De signis, p. 92; On signs, ¶32). Whereas the modistae positioned the will as determining the appropriate modus significandi in an act of speaking, Bacon theorizes will as the creative agent motivating language use and understanding by both speakers and listeners. The will is intimately connected with intentio for both speakers and listeners. Bacon’s view was partly shaped by Aristotle’s De anima, which had a significant influence on twelfth- and thirteenth-century grammar and philosophy, and Abelard. Abelard argued that the meaning or consequence of an utterance is determined by the speaker’s intention at least, if not more than by the strict form of the words spoken. Abelard asserted that because one has the moral responsibility to do what is right in the eyes of God, a person is guilty of sin even if he does not act on his sinful thought. Bacon’s account of the will expands on Abelard’s in that both speaker’s and listener’s wills are involved in any communicative act. Will (voluntas) is constitutive in that an interpreter’s choice or intellectual act creates the signifying relation: “Unless someone could conceive through a sign, it would be useless and empty. Indeed, it would not even be a sign” (De signis, p. 81; On signs, ¶1). Bacon’s semiotics of language places articulation and vocalization under the sign of voluntas, including both participants’ intentions, expression, and reception. Communication may depend on speaker and listener sharing some presuppositions and context, but understanding in itself does not entail shared presuppositions. The dialogism of speech entails a pragmatic frame. While relying on a realist theory of reference, Bacon’s semiotics of language and pragmatic theory of sign-making crucially depend on the freedom of the will of the speaker and listener. Both are interpretive agents.

Peter (of) John Olivi: Pragmatics and the Will to Speak Roughly contemporaneous with Bacon, Peter (of) John Olivi (OF, c. 1248-1298) was also an important, often original thinker in theology and philosophy.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

73

Olivi’s work contributes important ideas to our understanding of late medieval pragmatics and cognitive theory. Olivi’s provocative, sometimes controversial writings, especially on the will, voluntary poverty, and religious reform, were influential within the Franciscan Spiritual reform movement and among lay dissenters and reformers in southern France, especially Waldensians. However, because Olivi’s writings were condemned by his order and suppressed, his work did not have a wide circulation in the thirteenth century. Whereas Bacon situated language and pragmatics within a general theory of semiosis, Olivi developed a pragmatically inflected theory of mind, interpretation, and expression in the context of what he called the “communi intentio” [common understanding (in the sense of ‘intended meaning’)]. The concept matches up with the idea of social norms of meaning, accepted meaning, or dominant meaning. Some of Olivi’s recent interpreters have referred to a “certain intrusion of pragmatics” in his account of the signification of names (e.g., Mora-Márquez 2011), but I don’t think a pragmatics perspective was all that extraneous to Olivi’s account of verbal signification. Moreover, as we have seen, medieval logic texts can be and were sometimes read as affording pragmatic perspectives on philosophy of mind and philosophy of language, perspectives which the authors of those texts did not necessarily pursue on their own. I will argue that pragmatic understanding was a constitutive part of Olivi’s account of the nature of socially constructed meaning in language and of his theory of the autonomous human will as a precondition for faith. Olivi’s absolute insistence on the freedom of the will (voluntas) connects pragmatics with intention and communicative context and accounts for the creation of meaningful utterances in dynamic speech situations and registers. Olivi’s thinking and writing about language suggest several intersections of grammar, logic, and theology in the late medieval period. After studying in Paris (1267-c. 1272) he mostly taught in Franciscan study houses in southern France. (On Olivi’s life and career, see Burr 1976; Boureau and Piron 1999.) Not a grammarian but a philosopher and theologian, Olivi composed a series of logical questions (Quaestiones logicales) as well as commentaries on Lombard’s Sentences and various scriptural books. He also composed other sets of quaestiones, in particular the controversial Quaestio de usu paupere, whose strict understanding of the Franciscan vow of poverty put him at odds with the order’s mainstream branch. Olivi’s quaestio Quid ponat ius vel dominium, part of his Quaestiones de novissimis ex summa super IV Sententiarum, offers a strong defense of the will as a human right as well as an account of the priority of voluntary signs. As a theologian, Olivi held

74 

The Medieval Life of L anguage

provocative views on human freedom, free will (liberum arbitrium), and radical or apostolic Christian virtue. His work was severely censored by his order, which limited his influence on later philosophers. Nonetheless, Olivi’s ideas on the will and belief were very influential among dissenting Waldensian lay people in Languedoc, which further harmed his reputation in the eyes of his order. Some historians of philosophy regard Olivi’s views on cognition, speech, and interpretation as constituting a strong thirteenth-century version of positions later associated with nominalism (e.g., Mora-Márquez 2011; Pasnau and Toivanen 2018). In his commentaries on the Sentences and his Quodlibeta (c. 1289-1295; ed. Defraia 2002), Olivi developed a theory of attention, intention, and utterance with provocative implications for pragmatics and ethics. Like Bacon and other thirteenth-century grammarians, Olivi’s orientation toward language and pragmatics is Augustinian and semiotic. He also relied on Abelard’s theorizing about intention in terms of utterance and meaning. Olivi’s theory of intention primarily depends on his idea that the will is radically free (liberum arbitrium). Only that which is thought or done completely voluntarily can be considered truly free. For Olivi, freedom of the will is a fundamental condition for being human and for having faith. Free will is about choosing (recall Bacon’s use of the term “ad bene placitum”) and depends on that which determines itself and is not determined by some other cause or motivating force. It’s hard to think of a medieval theologian who didn’t subscribe to a theory of free will as a condition for true faith, but Olivi also relates the theory of will with a theory of language and communication. That is, free will for Olivi establishes a person as autonomous with respect to their intellect, reason, and intention as well as with respect to their relation with God. Freedom of the will is foundational for Olivi’s theology and for his pragmatic theory of signification and communication. Olivi analyzed utterance and speaker meaning in terms of communicative and discursive contexts. He argued intentions can be explicit (direct) or implicit (indirect, oblique). Intentions are analytically discoverable, but also integral to a speaker’s and a listener’s apprehension and understanding. In a quasi-phenomenological move, Olivi added the important act of “attention” (aspectus) to his account of intention’s role in perception and cognition. Earlier, Abelard had also worked out an account of the process of intention in cognition. We understand speech or any experience, encounter, or object only if we have first turned our conscious attention to that some-thing and then seek, an act of will, to engage or participate with the object or behavior in a physical, affective, or cognitive way (cf. Passnau and Toivanen 2018; A. Martin 2020). Some phenomenologists refer to this initial movement toward a

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

75

some-thing as intending. Acts of speaking, touching, or thinking are all based first on acts of attention, noticing or turning to. The will voluntarily notices or moves toward some-thing – an object, event, word – and then intends to see, touch, speak to, or hear it. In so doing, the will ignites the intellect to produce meaning in relation to that some-thing. Acts of speaking, then, involve two intentions. The wills of both speaker and listener are involved in any communicative utterance. Both free will and social norms and conventions are at the heart of Olivi’s theory of speech and signification. Olivi’s theory of free will might seem at first to conflict with norms and conventions, but those social contexts and conventional meanings, communi intentiones, do not determine utterance meaning absolutely, just as the form of the utterance does not necessarily determine utterance meaning. Rather, says Olivi, the communi intentiones are part of how human society is organized and functions by means of common presuppositions or accepted meanings. But communi intentiones also include flexible contexts and understandings which individual participants can decide to adjust or change in a given situation to suit other intentions and meanings. The motives for introducing or imposing such changes are what determine the speaker’s position. Voluntarism always has an ethical edge. For example, in his important Quid ponat ius vel dominium Olivi writes: Ad illud autem de duplicitate pro periurio, dicendum quod, ex quo hec vel illa vox est communiter et a communi intentione seu beneplacito hominum ordinata ad tale quid significandum, non potest quis propria auctoritate ea uti ad aliam significationem absque simulationis dolo, nisi suam intentionem clare exprimat auditori. Unde ratio determinate significationis ipsarum vocum non solum sumitur ab actuali intentione loquentis, set etiam a communi et a habituali intentione hominum illius lingue. Unde bene dicitur quod, sicut actualis significatio vocis est in actuali intentione loquentis et in actuali apprehensione audientis, sic eius habitualis significatio est realiter in communi et habituali intentione hominum. [Now, as to duplicity by perjury (duplicitate pro periurio), one must say that because this or that word is commonly and from the common intention or agreement (communi intentione seu beneplacito) of people ordained to signify such a thing, no one can use it on his own authority for another signification without being deceitful (simulationis dolo; lit., ‘the deceit of insincerity’), unless he clearly expresses his intention to the listener; so the reason for the determinate signification of utterances is taken not only from the actual (actuali) intention of the speaker, but also from the common and habitual/dispositional intention of the speakers of that

76 

The Medieval Life of L anguage

language. So, it is right to say that just as the actual signification of the word consists in the actual intention of the speaker and in the actual apprehension of the listener, in the same way its dispositional signification really (realiter) consists in the habitual and common intention of people (in communi et habituali intentione hominum).] (Quid ponat ius vel dominium, Ad argumenta primae partis; 2016: 8. My translation.)

Read from a pragmatic perspective, this passage shows Olivi directly confronting the possible disjunction between a speaker’s intention and community presuppositions and conventions in terms of the ethics of speech and the problem of sincerity vs. lying (sinceritas vs. mentiens). As contemporary conversation analysis has shown, our discursive maneuvers can be extraordinarily complex from utterance to utterance in a single exchange, let alone across exchanges and textual contexts (Schegloff and Sacks 1973; Levinson 1983: 284-370; Sidnell 2010). Sinceritas is one of the persistent topics in medieval pragmatics and metapragmatics, because it emerges from the analysis of intention and because it directly engages with the status of social interaction. Sincerity is also a question where speech, power, and authority intersect in medieval discourses, whether in everyday interactions or heresy examinations. Demonstrating good faith in an interaction is a principled presupposition. Sincerity and good faith must be continually maintained, whether assumed or demonstrated, in talk or the communicative event will be disrupted or contradicted by one or both participants. Good faith is the first presupposition in Habermas’ and Grice’s ideal speech situations, and sincerity establishes the ground for their theories of pragmatics and communication. But, as we shall see, not all speech situations are ideal, not all utterances are sincere, and not all interactions are cooperative, although that doesn’t necessarily mean they are entirely uncooperative, insincere, or lying either. Many medieval narratives reflect on the pragmatic consequences of assuming a speaker is acting sincerely or insincerely and how that interpretive frame affects the construction of meaning. Olivi’s account of discourse in Quid ponat ius posits two levels of signification for an utterance and three participants in a speech event: “actualis significatio vocis est in actuali intentione loquentis et in actuali apprehensione audientis” and “habitualis significatio est realiter in communi et habituali intentione hominum.” His distinctions among “actualis intentio loquentis,” “actualis apprehensio audientis,” and “habitualis signification” expand medieval logicians’ analysis of equivocatio as a fallacy to include, as did Bacon, both speaker and listener intentions in the construction of meaning from vocalizations. In Olivi’s account, utterance or discursive meaning

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

77

is not restricted absolutely to the speaker’s explicit intended meaning. In Quid ponat ius Olivi seems to be arguing that the meaning of an utterance is not a one-way transfer from sender to receiver. Rather, meaning emerges between speaker and listener dialogically as part of verbal give and take. Two participants, two intentions. Just as importantly, meaning is something both speaker and listener can have some control of. If only the speaker in a given situation has free will, how could a listener be described as anything but a passive receptacle? Furthermore, individual interactions are always located in a broad social context of dialogism, frames, institutions, and meanings. For Olivi, constructed meaning in a particular moment is shadowed by a more general “habitualis significatio” identified with a general social intention and set of understandings, which are ideally what a speaker and a listener share and use to interpret and evaluate utterances. Such common intentions have much to do with the presupposition of sincerity in communicative situations. Olivi acknowledges that we sometimes know experientially that participants in an oral or written exchange may not share the same presuppositions, but as the Quid ponat ius passage makes clear, he believes participants can repair or augment those mismatches with more or better or cooperative speech and metacomment. Olivi’s positive and pragmatic analysis of mismatched speech interaction is unusual in medieval philosophical discussions of discursive presuppositions and assumptions. As we have seen, logical texts (e.g., by Peter of Spain and Lambert of Ligny-le-Chatel) usually discuss equivocatio, amphibol(og)ia, and ambiguitas as fallacies or problems with argument which must be corrected with clearer, less ambiguous speech and analytic thinking. The logicians focused on individual propositions and did not usually account for situations as they occur in everyday interactions and dialogue where participants do not share relevant communi intentiones, where participants are not willing to acknowledge they do not share enough relevant communi intentiones, or when their respective presuppositions and common meanings are mismatched due to asymmetrical power relations. Other medieval discourses (poetic narrative, heresy inquiries) did address such conflicts and struggles and the question of equivocation, revealing how different social groups articulated pragmatic understanding and metapragmatic awareness as part of their public and group identities. Olivi’s notion of speech virtue, that is, acting in good faith, may resemble Grice’s Cooperative Principle (Grice 1975) in a certain sense, but not completely. Olivi’s idea of collaboration seems to grant full voluntarist participation to both speaker and listener, whereas Grice models “rational” collaboration around a speaker’s intention and participants’ common goal

78 

The Medieval Life of L anguage

or what Mey calls the “Communicative Principle” (Mey 2001: 68-71). Like Grice, Olivi affirms the importance of semantic norms in the speech community and the risk posed by ambiguous, vague, or disorienting kinds of speech, but he includes both speaker and listener as active participants in speech interactions. The speaker or writer can implicitly or explicitly claim an intention for an utterance or text, but an attending listener or reader does likewise from the point of view of her construction of meaning. The speaker and the listener individually enact an “attention” (aspectus) that initiates an “intention” (intentio) as part of their discursive interaction. Olivi’s account implies that both participants’ intentions are in relation to the communi intentiones of the community. For Olivi, the underlying pragmatic value is individual responsibility, which depends on free will. Speakers are responsible for intending and articulating their meanings. When a speaker’s utterances deviate from community norms as the audience understands them, the speaker needs to make his different interpretive contexts explicit for the listener. Listeners can, responsibly, talk back to the speaker and express their understanding or confusion. Olivi’s free will is the condition for dialogism. Speakers with free will are obliged to interact with good faith and to seek clarity in their discourse, because for Olivi and medieval philosophers and theologians generally human free will is the power which enables someone to conform their behavior to God’s law and wisdom. Olivi’s notion of speech interaction bears some similarity to speech act theorists’ analysis of Flouts. In contemporary speech act theory, when a speaker is obviously failing to observe one or more Gricean maxims, they should indicate that they are intending to communicate in a nonnormative way, whether metaphorically, sarcastically, or otherwise. What Olivi does not make clear, at least in Quid ponat ius, is whether recognizing such flouts is only up to the speaker or whether listeners too can and should make such mismatches known at the time of the interaction. What is clear for Olivi is that effective communication entails social cohesion. Speakers must speak in good faith, sincerely, and listeners must attend to the speech they perceive and interpret also with good faith while relying on shared contexts. If needed, the speaker has some obligation to help listeners adjust their attention and adopt a new presupposition in a particular interaction. Although Olivi’s pragmatic view of interaction has some similarities to Grice’s, Olivi emphasizes the dialogism of pragmatic interaction rather than just the speaker’s responsibility for violating a communicative principle. Olivi’s view sits alongside Bacon’s notion of re-imposition, whereby speakers generate novel meanings and listeners need to adopt a new orientation and presupposition to make the utterance meaningful and successful from their and the speaker’s point of view.

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

79

Olivi’s sincerity principle might seem obvious or go without saying – If everyone or everyone else is lying or not speaking in good faith, then how can we believe anything? – but his theory has far reaching implications for medieval pragmatic thinking. The degree to which a speaker and a listener’s intentions overlap constitutes part of the domain of shared understanding (communi intentiones) within a given speech context. Olivi argues that the domain of shared understanding is not only cognitive but also depends on speakers’ affective and “habitual” relations and presuppositions. A listener or reader’s intention and imposition is an act of pragmatic inferencing to flesh out her understanding of perceived behavior and utterance. Reciprocally, the interpretation of another’s intention is also a precondition for establishing or questioning communicative good faith and trust. While medieval logic did not directly address uncooperative interactions, medieval narratives did, especially when violations of community norms were embedded in the struggle for agency and discursive control or power in potentially threatening situations. (Chapters 3 through 6 explore how Olivi’s theory of human free will as a pragmatic condition is sometimes reflected in narratives by Chaucer, Bernard Gui, Margery Kempe, and William Thorpe.) Olivi rejected the external existence of universals, and like some other scholastic philosophers, adopted a conceptualist epistemology. However, he takes the theory of the mediating concept in a more distinctly agentive direction with respect to the interpreting subject. In response to the question, “Does a name signify present, past, and future (i.e., universally)?” he argues: Respondeo quod licet nomina rerum sint imposita ad significandum res, non tamen sunt imposita rebus nisi prout sunt apprehensae ab intellectu et sub illo modo sub quo intellectus vult res se intellectas significari. [I reply that, even if the names of things are imposed in order to signify things, they are not imposed on things except insofar as they are grasped by the intellect and under the mode under which the intellect wants to signify the things understood.] (Quaestiones logicales, ed. Brown, q.1; 1986: 338. My translation.)

Whereas Bacon argued that utterances refer directly to res, physical or abstract things, Olivi reasons that a thought is the formation of a mental utterance which is the concept itself. The thought/concept does not refer to a separate concept in the intellect; that is, “our mental word is our actual thought” (Tractatus de contractibus, 6.2.1, cited in Pasnau and Toivanen 2018). Aquinas similarly considered a concept to be a “verbum mentalis.” Olivi relates mental language or representation to an ethics of pragmatics.

80 

The Medieval Life of L anguage

He claims that significations are keyed not to abstract universals but to the perceptions of reality as apprehended by the intellect. When Olivi argues that signification constitutes how the intellect wants to signify the things understood (“intellectus vult res se intellectas significari”), he adopts a strong agentive, voluntarist position on the question of how one produces sincere utterances. Volo collapses into intendo. Olivi argues that concepts are not intermediaries of understanding but acts of thought – conceptualizing – actually attending to and interpreting the world. Despite his strong subjective theory of interpretation, Olivi still takes a matter-of-fact realist approach to reference. Considering the standard logical example sentence Homo currit (one of the metadiscursive puzzles for medieval pragmatic thinking), Olivi argues as did Bacon that “Cum enim dicitur ‘Homo currit’ non intendimus dicere quod conceptus quem habeo de homine vel homo prout est in intellectu nostro currit aut quod humanitas abstracta currit, sed potius quod res quae est homo concretive acceptus currit” [when we say ‘a person runs,’ we do not intend to say that the concept of a person I have or that the person as he is in our intellect [as a mental representation] runs or that abstract humanity runs, but rather that the real thing that is a person concretely attended to runs] (Quaestiones logicales, ed. Brown, q.4; 1986: 347. My translation.) Because he does not afford mental concepts separate status in understanding, Olivi can adopt a more realist-sounding approach to reference. However, Olivi’s theory of reference and signification is more complex with respect to actual utterances. In response to the question “Does an equivocal name signify simultaneously everything it signifies?” Olivi answers provocatively: Dicendum quod significatio nominis dupliciter accipitur: primo scilicet quasi habitualiter; secundo in actu aut exercitio, sicut fit cum ex intentione significandi actualiter profertur aut cum in certa narratione est scriptum. Primum modum habet ex impositione; secundum vero ex applicatione ad quae iam est et erat impositum. [It must be said that the signification of a name (noun) is taken in two ways: first, dispositionally (habitualiter), as it were; second, actually (in actu) or in the act of being uttered (exercitio), as it occurs when [the name] is actually uttered because of an intention to signify or when it is written in some discourse. [The name] has [signification] in the first mode from its imposition; and in the second [mode] from its application to the things to which it is now and was imposed.] (Quaestiones Logicales, ed. Brown, q.7; 1986: 353. My translation.)

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

81

Olivi’s initial question would seem to open the looming abyss of equivocatio. Signification relies on established or common understanding, but in actual speech possible meaning (in potentia) becomes actualized, located, or concretized by particular words combined in particular utterances in particular contexts. First imposition occurred at some indeterminate earlier time, while subsequent impositions (“actualiter”) occur at different times in the context of different apprehensions, different intellects, different circumstances. In Olivi’s voluntarist framework, semantics and language use are social and collective, but they take hold in discrete uses in diachronic and dialogical contexts. Olivi elaborates on this point in terms of prophetic speech and lying. If a word has multiple possible meanings, can all those meanings really be in play whenever the word is used in a particular context? The quick answer is, No, because potential meaning is narrowed and determined as ‘actual’ meaning in a particular context, whether verbal or extraverbal. A word’s meaning can change over time (recall Bacon on the relational meaning of father/son), but when it is actually used in a particular context, the word may only signify one or some of its possible significations. However, Olivi qualifies this “context-determined semantics” notion by considering how God’s prophets speak at particular times with particular intentions. Although the prophets’ speech is historically contextualized, their words codified in scripture continue to signify in different contexts over time as God’s word. God speaks “plures sensus” through the prophets, so the words can be interpreted literally, “mystically,” and in other ways over time by the faithful. Bacon and other grammarians and logicians also considered the relation between truth statements and diachronic changes of existents or the conditions of existents, challenging the argument that all truth claims are absolute and unconditional. For instance, Nicholas of Paris discussed the example of closing and opening one’s hand during the time one says “My hand is closed. Therefore, my hand is not open.” The two statements would seem to be generally synonymous, but, Nicholas argues, from a referential semantics point of view the conclusion (“Therefore, my hand is not open”) does not follow from the premise (“My hand is closed”) in the context of closing and opening one’s hand. As Nicholas explains it, the circumstances of the utterance’s referent have changed over time (transcasus), so the concluding proposition is irrevelant, even though taken out of context the statement would seem to be true as it is the negative inverse of the original premise, which is accepted. Transcasus, says Nicholas, is “transmutatio veritatis enunciationis secundum transmutationem temporis” [the change of the truth of a statement in accordance with the change of time] (Syncategoremata, ed.

82 

The Medieval Life of L anguage

Braakhuis, 1979: 169.22-170.4; CTMPT, 1:210). A similar principle regarding meaning over time is embedded in Bacon’s theory of semiotic relationality. Again, the pragmatic implications of logic are manifested by temporal contingency. Olivi uses a temporal rather than a logical framework to describe how verbal equivocations often arise over time as language and contexts change. He recognizes that equivocal speech can lead to misunderstanding, but his example of prophetic speech suggests he is more interested in the possibilities for continued communicativity of speech and text over space and time. An utterance’s ambiguity or multiple possible meanings can be negotiated or resolved if interpreters rely on the tacit or explicit cooperation and agreement presupposed in a discourse community, that is, if speakers, listeners, and interpreters engage in dialogue about their collective understanding. Given a powerful voluntarist theory such as Olivi’s, cooperation can override the restrictions of first imposition. New or changed contexts, says Olivi, always take precedence over original imposition. When God speaks through his prophets, a plurality of meanings (“plures sensus”) is potential for listeners and readers. The prophets’ words are interpreted (actualized) over time according to different modes of signifying (historical, mystical, etc.). As pragmatic thinking, Olivi’s description of verbal equivocation opens the possibility there can be multiple textual meanings of a given word or phrase in theological and poetic contexts, whether or not they are governed or initiated by first imposition or speaker intention. (For somewhat different approaches to Olivi’s semantic theory, see Mora-Márquez 2011; Rosier-Catach 2004: 160-166.) Thirteenth-century intentionalist grammarians and philosophers and even some modist grammarians can be said to have adopted varying degrees of a pragmatic perspective on semantics and signification. They combined grammatical and terministic logical concepts – impositio, significatio, equivocatio, actus significatus, actus exercitus – with certain moral or psychological values – voluntas, ad bene placitum, liberum arbitrium, sinceritas – to emphasize how speaker and listener attention, intention, and agency are constitutive of communication. These grammarians also appealed, again to varying degrees, to shared understandings and common intentions as foundations for speech communities and interpretive contexts. Most important, these philosophers and grammarians, especially Bacon and Olivi, recognized in different ways the dynamic, historically constructed possibilities for semantic changes in interpretive and communicative contexts. Their different approaches to equivocatio, semantic difference, and communicative norms suggest the grammarians and philosophers

Medieval Pr agmatics: Philosophical and Gr ammatical Contex ts

83

were developing questions within a deceptively stable consensus model of linguistic meaning. In fact, they reaccented terministic logic discourse in a semiotic and pragmatic direction. Bacon’s pragmatically-oriented semiotics of signs and speech foregrounds ongoing re-imposition as the tacit discursive practice for generating new meanings and new understanding. Olivi’s pragmatic perspective explicitly links equivocatio and mismatched talk not with logical fallacies but with individual speakers’ will to speak and intend meaning in ways which may depart from community presuppositions or with how the shifting position (deixis) of prophetic speech over time determines different meaning contexts for interpreters. The utterance’s continued meanings are shaped and remade by different audiences’ attention and reinterpretation of speech in various social and religious contexts and discursive communities. The meaning of scripture does not depend on the words alone. We can also read Bacon and Olivi’s pragmatic thinking as confirming the idea of the autonomous soul as part of language in use. Talk is distinct, contextual, and dialogical, constitutive of social belonging. As Mora-Márquez remarks: “Olivi explicitly and radically rejects the possibility that things or their properties affect in any way whatsoever the human soul.” Olivi does, however, point to how the soul and will are situated in time. Mora-Márquez continues: “As a result, the logical features of both human knowledge and human language depend to a great extent on the intentional modalities of those who think and talk. … [Olivi] also rejects the idea that universality as a logical feature of language is immediately dictated by real properties of external essences, thus putting forth the only known thirteenth century anticipation of fourteenth-century nominalism” (2011: 68, 70). Bracketing Mora-Márquez’s claim that Olivi is the thirteenth century’s sole protonominalist, I believe Bacon and Olivi’s theories of discursive interaction maintain subject agency and willed, creative motivations for speech from two somewhat different pragmatic perspectives. Both foreground the contingency of speech interactions. Both theorize how individual meanings can implicate and/or exceed conventionally shared understandings and meanings. Like Bacon’s semiotics, Olivi’s theory of will, attention, and intention reworks Augustine’s sign theory in terms of pragmatic perspectives on speech, meaning, and society. Unlike Bacon, Olivi also elaborated in his theory, more explicitly than did many of his contemporaries, an ethics of pragmatics grounded in a social concept of sincerity and responsibility.

2

Interjections: Does Affect have Grammar?5 Abstract This chapter first explores early grammarians’ accounts of the interjection and then elaborates on thirteenth-century grammarians’ analysis of interjections, communication rather than formal grammar, and the role of affectus, feeling, and emotion within grammar and semantics. The grammarians stretched the idea of verbal meaning to include both cognitive and affective signification as understood in specific contexts. Priscian (Institutiones, 2.15) provided grammarians with a framework for contrasting assertive sentences referring to substances with nonassertive sentences or interjections signifying mental dispositions or affects. These grammarians situated grammar and usage within interpersonal speech contexts. Keywords: interjection, medieval grammar, grammatical theory

The role of interjections in grammatical theory has a checkered past (on ancient grammar, see Colombat 2016). In contemporary linguistics, Biber calls interjections “inserts” (Biber et al. 1999), while Culpepper and Kytö distinguish word-form interjections from nonwords or “pragmatic noise” (2010: 199). When thirteenth-century grammarians became interested in the status of interjections and irregular speech, they ended up articulating pragmatic arguments about language and communication. Interjections posed provocative but unresolved theoretical questions about the nature of utterance, about the boundary between linguistic and nonlinguistic vocalization, and about linguistic and pragmatic meaning more generally. Some medieval grammarians and philosophers explored how natural and 5 An earlier version of this chapter was published as “Affectus in Medieval Grammar,” in Before Emotion: The Language of Feeling, 400-1800, ed. Juanita Feros Ruys, Michael W. Champion, and Kirk Essary (New York and London: Routledge, 2019), 26-37. Republished with permission.

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch02

86 

The Medieval Life of L anguage

conventional interjections functioned as speech phenomena and were communicative even though those expressions seemed to exceed the field of grammatica as earlier Latin grammarians had conceived it. Bacon shows this in his semiotic and grammatical work. Other grammarians associated, intellectually at least, with Robert Kilwardby composed treatises on the interjection which often elaborated pragmatic perspectives on grammar, meaning, and linguistic form different from the theory of language worked out by modistic grammarians. From the Middle Ages to the present, interjections and similar kinds of vocalizations have enticed linguistic exploration by challenging existing linguistic categories. Irène Rosier (1994) refers to a group of mostly thirteenth-century grammarians and philosophers as “intentionalists” with respect to language use and meaning. But the grammarians she discusses deployed many of the same linguistic and logical concepts and metalanguage as the modistae and, more distantly, Abelard, especially around the concept of intentio. What separates these different philosophers and grammarians is the theoretical framework or presuppositions they adopted or the outcome they were looking to. Both intentionalist grammarians and modistae were by-and-large realists with respect to universals, whereas Abelard was a quasi-realist at best. Both the intentionalists and the modistae began with Augustine’s statement that “omnis doctrina vel rerum est vel signorum. sed res per signa discuntur” [all knowledge is about things or signs, but things are known through signs] (De doctrina christiana, 1.2.2. My translation.). The modistae aligned grammatical theory with Aristotle’s causes (material, formal, efficient, final) to account for the relations of speech, mind, and reality. They repositioned grammatica as a speculative science (scientia, theory) because the goal of their theorizing was to understand and explain universally necessary structures and processes (modi) by which the mind connects up speech and reality (res). Modists’ grammatical explanations included linguistic analysis and contextualized information and concepts drawn from other disciplines, especially logic, physics, and philosophy of mind galvanized in part by newly available Aristotle texts, especially De anima, Physics, and Posterior Analytics and commentaries by Avicenna, Algazel, and others. They privileged syntax over the lexicon, modus conceptus over modus affectus, explanation over description, and to an extent, structure over function (for overviews of the modistae and late medieval grammar, see Rosier 1983, 1999; Kelly 2002). Intentionalist grammarians, on the other hand, focused more on language in use (usus) and explored the intersections of grammar and logic with a communicative perspective. Their emphasis on intentional meaning was more in line with Abelard’s earlier theory of intentio with respect to

Inter jec tions: Does Affec t have Gr ammar?

87

the signification of utterances and actions. Intentionalist grammarians discussed syntax and form in relation to speakers’ intentions and listeners’ inferences, and they paid attention to the roles of semantics, especially equivocation and nonliteral usage, in language theorizing. As we saw in Chapter 1, some grammarians, especially Bacon, adopted semiotic perspectives on language and attached great importance to the role that contexts, situations, and inferencing play in the creation of meaning in speech. They privileged communicativity and the dialogism of speech interaction rather than formal grammatical congruitas or perfectio. As a result, the traditional but relegated Latin word class interiectio and certain syncategorematic words became the theoretical site for many of their pragmatic investigations. In this chapter I briefly outline how Latin grammarians thought about the interjection as word class, affective utterance, and pragmatic signifier. Then I focus in detail on the pragmatic implications of grammarians’ analysis of the interjection and how their attention reorders linguistic theory. The degree of continuity or lack of continuity we recognize between early Latin grammar and late medieval grammar depends in part on the significance we give to the development of pragmatics in theory and practice. The word class interiectio intersected with the concept of affectus (feeling, disposition) in that grammarians sought to account for aspects of language structure and use not explicitly included in grammatica’s traditional metalanguage and theory of morphology, parts of speech, and syntax. The question, Are interjections even linguistic at all? – Helias didn’t think so, but PseudoKilwardby (Paris, 13th c.) and Bacon did, for different reasons – reveals much about how pragmatic thinking and metapragmatics informed late medieval understanding of language use, meaning, and affect in theory, religious discourse about orthodoxy, and literature. Prior to 1200, Latin grammatica accounted for interjections in several ways. Boethius established the vocabulary for feeling with his influential translation of Aristotle’s De interpretatione, using passio or motus as equivalent to affectus (“motus animi,” “passiones animi”). In De divisione Boethius equates natural affects with interjections: “sunt enim quaedam noves quae dolores designant, aliae quae animi passions naturaliter quae nomina non sunt, ut interiectiones” [For there are certain spoken sounds that designate states of grief, others that naturally designate passions of the mind and yet are not names, e.g., interjections] (De divisione, ed. Magee; 1989: 34-35). Boethius describes interjections as natural vocalizations and therefore not strictly ‘linguistic’ vocalizations. Many Latin and early medieval grammarians discussed the interjection as a pars orationis [part of speech] in terms of the relation between language

88 

The Medieval Life of L anguage

and feeling. Latin grammarians repeated or paraphrased existing definitions and materials from Donatus’s (4th c. CE) Ars minor and Ars maior or earlier Roman grammarians such as Remmius Palaemon (1st c. CE). Their accounts of the interjection relied on a matrix of key terms and analytic concepts: affectus, motus, erupta, passiones, incondita, and (in)completus. These grammarians also introduced innovations into grammatical description, including using affectus meaning ‘disposition, mood, emotional state’ as a term in technical grammar, especially when discussing the interiectio. In contrast to Greek grammarians, the Latin grammarians began to regard adverbs and interjections as separate parts of speech, the latter expressing the disposition or attitude of the mind or soul. Donatus’s definition was standard in elementary grammars and later commentaries: the interjection is “Pars orationis significans mentis affectum voce incondite” [a part of speech signifying the disposition of the mind with a disordered/irregular voice] (Ars minor, 9; ed. Holtz 1981: 602. Translations in this chapter are mine unless otherwise noted.). The adverb incondita here suggests an irregular vocalization outside what is considered normal or measured in terms of ratio or oratio. Donatus connected the interjection as expression with the Stoics’ and Latin grammarians’ standard array of emotions: happiness, sorrow, surprise, and fear. In his Ars maior Donatus omitted reference to disordered speech, provided an etymology of interiectio based on syntactic usage, and connected individual interjections with a slightly different array of emotions: “Interiectio est pars orationis interiecta aliis partibus orationis ad exprimendos animi adfectus: aut metuentis, ut ei; aut optantis, ut o; aut dolentis, ut heu; aut laetantis, ut evax” [The interjection is the part of speech inserted into other parts of speech to express the disposition of the soul: either fearful, as ei; or desiring, as o; or sorrowful, as heu; or joyful, as evax] (Ars maior, 2.17; ed. Holtz 1981: 652). Many Latin grammarians use similar definitions of the interjection but with certain differences. Servius in his otherwise important commentary on Donatus adds little to the analysis of the interjection. Diomedes (4th c. CE) omitted the reference to feelings, stating “Interiectio est pars orationis affectum mentis significans voce incondite” [the interjection is the part of speech signifying the disposition of the mind/soul with disordered/irregular speech]. He said interjections can signify affectus either by convention or by attaching to a verb (Diomedes, Ars grammatica, 1; KGL 1:419). Probus (1st c. CE) called attention to how the interjection expresses “animi motum per suspirationem” [movement of the mind by deep breathing (lit., sighing)], where the forceful expression of air becomes a vocalization and marks a particular sequence of sounds as an interjection rather than as another part

Inter jec tions: Does Affec t have Gr ammar?

89

of speech (Instituta artium; KGL 4:146). Probus and others make explicit the motus theory of emotions in classical and medieval thought, whereas Diomedes leaves unclarified the semantic relation between expressing and signifying. In a move which would become important in the thirteenth century, Charisius (4th c. CE), citing Palaemon, restricted affectus in his definition to what is instinctive or natural in vocal expression: “interiectiones sunt quae nihil docibile habent, significant, tamen adfectum animi, velut heu eheu hem ehem eho hoe pop papae at attatae” [interjections cannot be taught; nevertheless, they signify the disposition of the mind, such as heu (alas), eheu, hem, ehem, eho, hoe, pop, papae, at, attatae] (Charisius, Ars grammatica, 2.16; KGL 1:238). Some Latin grammarians and Isidore (c. 560-636) explicitly linked the interjection with affectus, characterizing a particular kind of utterance or feature as expressing an emotional state of mind or disposition and specific to an individual language. In the fifth century, Pompeius and Cledonius, commenting on Donatus, applied the function of the interjection and the concept of affectus to discourse rather than just parts of speech: “quidquid affectum exprimit, interiectio est …” [whatever [utterance, vocalization] can express a mental mood is an interjection] (Cledonius, Ars grammatica; KGL 5:26); “quidquid potest animi motum exprimere …, interiectio dicenda est” [whatever can express a motion of the mind … is called an interjection] (Pompeius, Commentum; KGL 5:281; cf. KGL 5:97-98). Pompeius contrasts the neutral constative “mortuus est Vergilius,” an assertion which does not entail an emotional load, with the expressions “Va mortuus est Vergilius” (joy, surprise) and “Heu mortuus est Vergilius” (sorrow) which do entail an affective load. In Pompeius’ analysis the interjection functions disjunctively as a sentential or utterance frame rather than as a lexical item (word) combined with other lexical items (cf. Chapter 3 on interjections as shaders). Pompeius’ analysis also suggests that any utterance expressing a disposition or mood can be considered an interjection or to have interjective force. Isidore’s influential encyclopedia Etymologiae sive origines (c. 625) repeats Late Latin grammarians’ association of affectus with primarily non-cognitive vocalizations, interjections, and emotional, sudden, or disordered speech. However, Isidore adds that interjections are language specific and resist translation: “Interiectio vocata, quia sermonibus interiecta, id est interposita, affectum commoti animi exprimit, sicut cum dicitur ab exultante ‘vah,’ a dolente ‘heu,’ ab irascente ‘hem,’ a timente ‘ei.’ Quae voces quarumcumque linguarum proprie sunt, nec in aliam linguam facile transferuntur” [The interjection is so called because it is inserted, that is, interposed, into utterances [and] expresses the disposition of an agitated/excited mind, for

90 

The Medieval Life of L anguage

example, when vah is said for/because of joy, heu for sorrow, hem for anger, ei for fear. These utterances are proper to individual languages [and] cannot be easily translated into another language.] (Isidore, Etymologiae, 1.14). Christian Latin grammarians sometimes cited Racha (Matthew 5:22) as a Hebrew interjection – It’s actually Aramaic, roughly equivalent to English Fool! – as an example of an interjection’s untranslatability. Donatus, Servius, Diomedes, and other Latin grammarians argued the speaker’s emotional or mental state (“affectus mentis”) can make an utterance’s phonological or syntactic features irregular (“incondite”). However, every interjection by definition can appear as a sudden outburst and cut through an otherwise cohesive linguistic constituent or can stand outside constituents altogether. The interjective utterance, natural or conventional, can be communicative even if it violates the ratio principal that meaningful speech is rational, rule-ordered, and conventional. However tentative, the Late Latin grammarians’ discussions of interjections and affectus enlarged the traditional scope of grammatica to include pragmatic, indirect, or affective meaning. Strong emotions such as joy, sadness, or fear can cause a person to speak with explosive vocalizations or forceful accent or use elliptical syntax or shift the typical stress pattern or articulation of certain words. Fear, anger, and other strong feelings can cause the speaker to utter conventional interjections or seemingly non-linguistic sounds or ungrammatical speech. All these forms of borderline grammatical utterance are taken as revealing or enacting the speaker’s interior state of mind or feeling. Interjections, says Diomedes, exceed grammatical constraints; they are “quae affectus potius quam observationes artis inducant” [[those parts of speech] which represent affect more than follow (grammatical) art/rules] (Diomedes, Ars grammatica 1: KGL 1:419). What made, and makes, interjections linguistically interesting is that they are understood as word-like, meaningful, yet in many respects irregular. For the Latin grammarians, interjections slip around the presumed boundary between linguistic and non-linguistic utterances while conveying meaning inferred in various contexts. Interjections provoke a pragmatic explanation. While Priscian’s account of the interjection and affectus did not differ substantively from those of earlier Latin grammarians, his discussion was more elaborated, if somewhat disorganized, and was supported with literary examples from Virgil and Terence. He claimed the interjection is a separate word class “quia videtur affectum habere in se verbi et plenam motus animi significationem, etiamsi non addatur verbum, demonstrare” [because it seems to contain/render emotional disposition in the word itself and show the full meaning of the movement of the mind, even though it is not added/

Inter jec tions: Does Affec t have Gr ammar?

91

attached to the verb] (Priscian, Institutiones grammaticae, 15.7; KGL 3:90). Priscian distinguishes the interjection from the adverb according to the grammarians’ description of the adverb as adjacent to the verb (“eo quod sit iuxta verbum”). Interjections, he says, express or signify in themselves. They embody or perform (“in se verbi”) mood or emotional disposition (affectus) rather than depend on another element in the clause. At one point, Priscian seems to conflate affectus and passio when he refers to “voces, quae cuiuscumque passionis animi pulsu per exclamationem intericiuntur” [speech sounds which having pushed out/expressed some passion of the mind are interjected [into the utterance] through loud calling out (exclamationem)] (Institutiones, 15.7; KGL 3:90). Like other Latin grammarians, Priscian associates the interjection and affectus with emotional disposition or behavior which deviates from ‘proper’ or normal vocal accentuation (exclamatio) or imitates inarticulate (that is, non-linguistic) sounds, such as giggles or expressions of disgust (e.g., Institutiones, 1.5, 15.7; KGL 2:20, 3:91). Giggles and expressions of disgust pragmatically challenge the formal grammatical notion of an ‘utterance,’ as we saw in Bacon’s semiotics of language (and as we will see in Chaucer’s Miller’s Tale). In Priscian’s account, affectus in speech, especially interjections, is somatically and psychologically transformative as well as expressive. Priscian’s account of the interjection was influential, if not always directly, in that his grammar was a core text for later medieval grammarians and commentators. Carolingian grammarians sometimes used the concept of affectus to address interiority and intentionality. In doing so they introduced pragmatic aspects of implicature and inferencing in their grammatical discourse. Carolingian commentators suggest that interjective or incomplete (elliptical) utterances – clauses lacking an explicit subject or main verb – connote emotional affectus or disposition in ways which other parts of speech or complete utterances – clauses including a subject noun phrase and a predicate verb phrase (NP+VP) – do not. In other words, these commentators took grammatical deviation as performative signifiers of strong affect, just as some grammarians argued that figurative language signified by deviating from grammatical and (literal) semantic norms to create poetic or symbolic meaning or theological truth. Carolingian grammarians connect interjections and ellipsis with affectus. In medieval grammar affectus is never identified with neutral or empty emotionality. Smaragdus (9th c. CE), for example, connects affectus and the interjection to make explicit the role of affectus in the mediation of interior disposition and exterior verbal expression: “‘Interiectio’ proprie dicitur vox confusa de mentis archano prolata, quae tantum ad hoc profertur in

92 

The Medieval Life of L anguage

publicum, ut interioris hominis lucide demonstret affectum” [The interjection is rightly called disordered speech for having revealed the hidden mind, which, to the extent it is expressed publicly by this [part of speech], clearly shows the disposition/emotional state of the inner person] (Smaragdus, Liber in partibus Donati; CCCM 68: 233.). Smaragdus substitutes the Latin near synonym confusus for the Latin grammarians’ “incondita,” perhaps borrowing from Pompeius (Commentum; KGL 5:281). Moving away from a strict word class description, Remigius of Auxerre (9th c. CE), in his influential commentary on Donatus’s Ars maior, argues that any utterance can be interjectivized by inserting a conventional interjection, word (Heu), or phrase (Deo gratias) into a statement or question (Commentum Einsidlense in Donatis artem maiorem; KGL 8:266). The primary function of all interjective utterances, he says, is to express “affectus animi” (KGL 8:265). Remigius makes two points regarding affectus which become important in later medieval grammatical theory. First, affectus in speech reveals the speaker’s interior emotional state through the nature or quality of the vocal articulation as well as through the use of a particular kind of word. Second, despite their sudden, explosive eruption in speech, interjections are like other kinds of words in that they are controlled by the will or volition (“Affectus dicitur desiderium, i[d est] voluntas”; KGL 8:266). However, Remigius leaves unclear whether interjections are determined by ratio. In Chapter 1, we saw how beginning in the mid-twelfth century in Paris and Oxford grammarians and philosophers rethought the traditional Latin word class framework for linguistic analysis by considering grammar in relation to logic and philosophy of mind. Philosophers such as Abelard and William of Champeaux analyzed grammatical questions using and critiquing dialectic and logical frameworks from Aristotle and Priscian. With the distinction between categorematic and syncategorematic terms, they distinguished a decontextualized word’s meaning from its function and functional meaning in a sentence. This is an especially significant bit of pragmatic reasoning, not unlike what the Stoics had done centuries earlier. Stoic philosophers argued that in practice sentential or intended meanings were not bound by a componential theory of the sentence. Substantive words (noun, verb, adjective, pronoun, participle) signify essential properties of things, whereas syncategorematic indeclinables (adverb, preposition, conjunction, interjection, vocative) signify relations among things. In the early thirteenth century, two new approaches to grammar emerged, the modistic and the “intentionalist,” in commentaries on Priscian and grammatical treatises. Many of the latter especially presented a vigorous pragmatic perspective.

Inter jec tions: Does Affec t have Gr ammar?

93

The structure and function of the interjection and the theory of affectus were important for both modistic and intentionalist grammatical theory, but for different reasons. After 1275, Helias’s commentary on Priscian’s Institutiones became extremely influential for later grammarians and some philosophers of language. Helias consolidated a certain logical view of sentence form and congruitas, but he also destabilized the notion of grammaticality to argue that some kinds of sentences were still communicative even if they did not fulfill the criteria for complete syntax. Generally, Helias adopted Priscian’s definition of the sentence, “Oratio est ordinatio dictionum congrua, sententiam perfectam demonstrans” [A sentence is congruently ordered speech indicating a complete thought/meaning] (Institutiones,2.15; KGL 2:53; cf. 3:108-109). Helias denied that interjections are a distinct word class, although he did accept some conventional interjections as partes (e.g., Latin papae, heu). Helias argued that if interjections proper are “primitiva” and signify “naturaliter,” then the interjection “non esse partem orationis per se” [is not a part of speech per se]. Commenting on the interjection, Helias anticipated Bacon’s theory of impositio ad placitum, if only to rebut it avant lettre. He proposed that interjections such as papae which signify “ad placitum” [arbitrarily, according to convention] should be understood as a subclass of adverb (Summa super Priscianum, 1993: 807-808). Helias echoes Augustine, who had claimed “Interiectio non pars orationis est, sed affectio erumpentis animi in vocem” [the interjection is not a part of speech but a movement (affectio) of the mind bursting into speech] (Regulae, KGL 5:524). Augustine’s authority among thirteenth-century grammarians and his denial that the interjection is a part of grammar made relevant the question of how grammar is defined and what are taken to be the limits of grammar with respect to meaning. Unlike Augustine and Helias, other thirteenth-century grammarians claimed the interjection is not only a part of speech but central to grammar because it is the word class which most directly, essentially, signifies or makes available mental or emotional experience to an audience. The author of the Priscian commentary mistakenly attributed to Jordanus of Saxony (OP, c. 1190-1237), for example, distinguished two kinds of linguistic signification of reality with respect to interjections and other parts of speech: “Interiectio signif icat rem existentem in anima secundum suam substantiam, sed aliae partes significant rerum similitudines” [The interjection signifies reality in the mind according to its substance, but other parts [of speech] signify the likenesses of things] (Notulae super Priscianum Minorem, ed. Sirridge 1980: 22. Sirridge convincingly refutes Grabmann’s attribution of the commentary to Jordanus.) Jordanus’ phrase “rem existentem in anima”

94 

The Medieval Life of L anguage

refers not to concepts or representations of external things but to interior affective experience as a some-thing (res). Jordanus seems to be arguing that the interjection is a performative part of the emotional experience which the vocalization self-reflexively manifests as meaningful in discourse and which other speakers recognize or attend to and associate with a meaning. Jordanus implies interjections are part of the grammatical and conventional semantic structure of whatever language speakers acquire or learn and such vocalizations are performative and socially recognizable parts of emotional experience. By the mid-thirteenth century, the concept of affectus was playing a larger role in some grammarians’ rethinking of the relation of linguistic form to cognition, disposition, and communication in grammar and pragmatics. Thirteenth-century intentionalist grammarians adopted the framework and terminology of the modes of being, signifying, and understanding, but they gave more weight to the contexts of utterance and the respective roles and shared assumptions of speaker and listener for determining the meaning and evaluating the acceptability and grammaticality of particular utterances (cf. Rosier 1994: 11-12). Robert Kilwardby, Roger Bacon, and the grammarians known as Magister Johannes (OF, 13th c.) and Pseudo-Kilwardby (13th c.) responded to Priscian and by implication Helias by asking: Do interjections signify the way other kinds of utterance and word classes do? What role does affectus play in different kinds of utterances? Is a moan or a giggle a kind of utterance? (On the authorship of the Commentum super Priscianum Maiorem attributed to Kilwardby, see Rosier 1994: 126-131; Fredborg et al. 1975: 12+-17+.) Earlier, we saw how Bacon and Jordanus answered the first question from a semiotic perspective: Interjections are meaningful and communicative as performative and embodied signification, not reference. Magister Johannes in his treatise Sicut dicit Remigius on complete and incomplete syntax analyzed interjections morphosyntactically and pragmatically in relation to affectus. He emphasized the interjection’s communicative meaning and function rather than its grammatical form and function as a word or phrase. In his commentary on Priscian Maior (Institutiones, Books 1-16), PseudoKilwardby developed a theory of affectus and mental language as conceptual and as the motivation for acceptable and meaningful vocalizations, even if those utterances are not always congrua grammatically. For these thirteenth-century grammarians, the traditional word class interiectio and affectus prompted them to focus on the pragmatic construction of meaning, the relevance of context/situation, and vocalizations as communication or performance of feeling. The supposed referentiality of speech was complicated by the expression and understanding of feelings

Inter jec tions: Does Affec t have Gr ammar?

95

and attitudes. Usually, these grammarians, reworking materials from the Latin grammarians, argued that both primary, natural interjections (grunts, sighs, moans) and secondary, conventional interjections (heu, ei, evax, papae, o) signify the speaker’s interior state of mind or disposition and consignify affects of hurried motion, urgency, crisis, fear, sadness, or other strong feelings. Groans and instinctive cries of pain or pleasure are not language as such, but natural signs, as Bacon argued, vocalized bodily expressions which imply meaning through situation, affect, and context of use, knowledge shared or presumed to be shared by speaker and listener. According to Magister Johannes, conventional Latin interjections such as heu function as signifying utterances which communicate experience or feeling “per modum affectum,” whereas other parts of speech signify “per mentis conceptum” (Sicut dicit Remigius, Ms. Bibliothèque nationale lat. 16618, fols. 48va, 49ra-rb, 51ra; transcribed and in French translation in Rosier 1994: 60-65). These grammarians imply that natural and conventional affective understandings are universally human. For Pseudo-Kilwardby, signifying under modus conceptus is to signify something under the mode of truth, whereas signifying under modus affectus is to signify something “under the mode of the Good or its opposite, which moves the soul to action” (Commentum super Priscianum Maiorem, CUL, Peterhouse 191, fol. 107v; transcribed and in French translation in Rosier 1994: 198). In this passage Pseudo-Kilwardby attributes affectus to ethical decisions or dispositions rather than to truth and falsity, to nonassertive rather than assertive discourse (reflecting the influence of Aristotelian logic, especially De interpretatione, on Latin grammatical theory). Heu expresses my sorrow as an emotional sign per modum affectus, even if the expression is a conventionally recognizable Latin word, whereas doleo [I suffer/lament/am sad] predicates the way I feel per mentis conceptum. The interjection externalizes and performs my feeling (actus exercitus) as an embodied expressive act, whereas the predication refers to my feeling (actus significatus) through discrete linguistic signification (Nuchelmans 1988; Rosier 1994: 157-206). Robert Kilwardby in his Super Priscianum Minorem interprets the interjection as both speech act and interpreted act by distinguishing between production and reception. When spoken, the interjection expresses the speaker’s affectus and signif ies naturally from the point of view of the speaker. But when received by the listener, the interjection is interpreted as a concept and signifies conventionally (ad placitum). In other words, from a pragmatic perspective the listener understands the utterance as meaningful without necessarily experiencing the affectus of the speaker (Commentum super Priscianum Minorem, Vatican Urb. lat. 298, fol. 9rb,

96 

The Medieval Life of L anguage

transcribed in Rosier 1994: 59). An interjective utterance is a communicative act, and it situates feeling within a communicative act which depends on both a speaker and a listener’s understandings of context. But the speaker and listener are not necessarily having the same mental experience or sharing the same context or cognitive understanding with the same words. In Kilwardby’s account, modus affectus entails emotional experience (affectus, passio) as behavior and feeling on the speaker’s part and conceptual experience on the listener’s through language. Primary and secondary interjections are kinds of interjective utterance, produced “subito” [suddenly] in extreme or tense situations “secundum rem” [according to reality] but without intention or deliberation. Recall that Bacon referred to both kinds of utterance as natural (“instinctu naturali et impetu naturae”) from the speaker’s point of view, but conventional (“ex institutione,” “ad placitum”) from the listener’s point of view in so far as the utterance is understandable and considered acceptable in speech interaction according to the speech community’s norms and usage (De signis, p. 84; On Signs, ¶9-10; cf. Bacon, Communia naturalium, in Bacon, On Signs 2013: 120-121). That’s Kilwardby’s point, too. Urgency, alarm, and other contextual aspects of interjective meaning are consignifications. Interjections, as kinds of natural signs, are produced with only partial deliberation or voluntary intent, yet they are still cognizable through concepts. With respect to discourse, affectus is strong emotional speech, expressive and meaningful in context to someone in some way. Affective speech can be described as both expressive and meaningful beyond what is literally articulated in conventional syntax. Affective or interjective speech is situational, indexical, pragmatic. Bacon had used virtus as a technical term to do some significant theorizing about language and meaning. Other grammarians did so too. According to the Priscian commentary attributed to Jordanus of Saxony, affectus and conceptus are not the same force or power (virtus) of the soul (Notulae, ed. Sirridge 1980: 22). Pseudo-Kilwardby, citing Aristotle’s De anima, says that “virtus enim appetitiva et imaginativa concurrunt ad formationem vocis” [appetitive and imaginative virtus (capacities) work together in forming vox] (Commentum super Priscianum Maiorem, ed. Fredborg et al. 1975: 10-11). Deploying the standard cluster of metalanguage to describe interjections and affectus, he says that modus affectus includes utterances produced suddenly (subito) which nonetheless express the speaker’s understanding of or relation to immediate objects or events because such exterior, embodied utterances are controlled by the will (voluntas) (Commentum, 1975: 11, 15). Pseudo-Kilwardby’s argument suggests that whatever is uttered is ultimately willed and intentional (because controlled by the intellect) rather than

Inter jec tions: Does Affec t have Gr ammar?

97

sometimes natural or instinctive. However, he qualifies this by adding that word and word-like utterances are not only physical acts like the bodily eruptions (erupta) of natural signs but also part of a grammatical system of lexical impositions, conventional signs under the mode of affectus. Vox is produced by a combination of natural and intellective forces. Pseudo-Kilwardby’s theory of the interjection, affectus, and linguistic signification depended on not one but two somewhat different theories of the sermo in mente (word in the mind, mental language), both of which imply a voluntarist approach to pragmatics. In Version One, sermones in mente comprise the set of mental concepts, meanings and thoughts considered universal for humans and variously denoted as “passiones quae sunt in anima,” “similitudines,” or “species intelligibiles” (Pseudo-Kilwardby, Commentum, 1975: 53, 54, 61; cf. Rosier 1994: 129-30). These mental concepts are “similitudines” of the extramental objects and events they refer to (something which Abelard had vehemently denied). In Version Two, the intellect through its “virtus … appetitiva et imaginative” intentionally produces a sermo in mente (or sermo interior), isomorphic, but not iconic, with the “sermo exterior” vocalized as speech or inscribed as writing (PseudoKilwardby, Commentum, 1975: 10). Pseudo-Kilwardby’s conceptualist theory of signification presents two different kinds of mental representations, one based on iconicity, the other on equivalence, both of which he says anchor the relation between utterance, including interjections, and signification. The two models he proposes suggest how the interjection and affectus continued to challenge and change received grammatical and logical theories of utterance and signification in the later Middle Ages. Some thirteenth-century grammarians found the concept and theoretical potential of affectus relevant for understanding not only interjections but also other kinds of so-called elliptical, figurative, or irregular speech which have interjective force. Recall how Bacon and Magister Johannes argued that because interjections are produced under modus affectus without conscious intention or deliberation, they are natural, not conventional, expressions, however word-like the utterances might appear. Bacon complicated further this new complication in grammatical theory by distinguishing between natural signs such as a sunrise, thunder, or smoke and natural interjections which function as speech without being strictly deliberate or even words. Moreover, in Bacon’s account, natural interjections are exemplary but not the only cases of urgent, sudden, emotionally charged, or affective speech. Crisis, urgency, sudden or unexpected joy, sorrow, fear, pain – such feelings and emotional experiences can also motivate or give rise to elliptical syntax (clauses lacking a subject or predicate) and figurative or nonnormative

98 

The Medieval Life of L anguage

utterances. Such affective speech often involves creative uses of language, what Bacon referred to as secondary or ongoing re-impositio. The contexts or presuppositions listeners use to infer meanings may be immediate and local or more general. Magister Johannes pushed Bacon’s and Pseudo-Kilwardby’s analysis of the interjection further in a pragmatic direction by claiming that elliptical or incomplete sentences can have interjective force. Earlier we saw that many twelfth- and thirteenth-century logicians had contained the topic of complete (perfecta) and incomplete (imperfecta) or elliptical utterances by setting incomplete utterances aside altogether (e.g., Abbreviatio montana, ed. de Rijk, 1967, vol. 2.2: 78-79) or by simply saying that cognitively elliptical speech is superseded or rewritten mentally as a ‘proper’ complete utterance. However, Magister Johannes, following Priscian (Institutiones, 15.7; KGL 3:90), argues that because interjective utterances are disjunct, they need not be constructed as a complete clause with explicit or even implicit subject NP + predicate VP in order to be meaningful or understood. Elliptical speech, he argues, conveys a first meaning as a command, warning, or other speech act and a second consignification as disposition or affectus. Magister Johannes gives the example of someone yelling, “Aqua. Aqua.” in front of a burning house. Villagers hearing the cry immediately understand and bring water to fight the fire (Sicut dicit Remigius, MS Bibliothèque nationale lat. 16618, fol. 47ra, transcribed and in French translation in Rosier 1994: 255). The elliptical cry is understandable when people inference the meaning from the external context and social convention and respond with action, that is, with pragmatic uptake. (Chaucer imagined a similar scene, with explosive effect. See Chapter 4.) When someone in the street or on the dramatic stage cries “Heu mihi.” [Woe is me] or “Heu mortuus est” [Alas, he is dead], Latin speakers use both linguistic and external contexts to infer the person or character’s distress at having been betrayed, disappointed, or saddened. Audiences might grasp some of the affect from physical gestures, but not the contextual details. Pragmatically, the elliptical expression “Aqua. Aqua.” is an urgent speech act (command), whereas “Heu mihi.” is a performative utterance which is a public part of the speaker’s feeling of betrayal as well as an indexical signifier of that feeling. Magister Johannes claims that interjections and interjective utterances such as “Aqua. Aqua.” are complete syntax in context. Similarly, he says, the concluding dialogue of the Mass consists of an exchange of utterances, all of which are syntactically perfectio ad intellectum. The utterances by the celebrant and congregation are connected as adjacency pairs or deictics. The words, gestures, speakers/listeners’ intent, and extraverbal context,

Inter jec tions: Does Affec t have Gr ammar?

99

all contribute to make the collective speech a successful performative despite the explicit lack of certain words. Likewise, emotional utterances such as the interjections Heu or Papae or the simplex Bene (equivalent to “Yes, yes,” “Well done,” or “Good, keep going,” depending on the context) are “perfectio ad intellectum,” if not “perfectio ad sensum” (Sicut dicit Remigius, Ms. Bibliothèque nationale lat. 16618, fol. 47ra, transcribed and in French translation in Rosier 1994: 255; cf. Rosier 1994: 182-87). Magister Johannes contrasts perfectio with the grammatical concept of congruitas in Priscian and the modistae’s texts. Whereas the modi significandi determine formal congruity, the modi intelligendi determine completion in the sense that that which is being talked about is comprehended or understandable. Congruitas belongs to grammar, whereas perfectio belongs more to pragmatics and communication. Another mid-thirteenth-century grammarian whose work is associated with Kilwardby, the anonymous author of the Quaestiones de interiectione, took this semantic and pragmatic analysis further still. (On the manuscript relations among Kilwardby, Pseudo-Kilwardby, and Quaestiones de interiectione, see Rosier 1994: 287-288.) The Quaestiones author argues the interjection signifies both a concept and an affect (Quaestiones, 2.1.1.2-4, 2.1.1, Sol, Ms. Cambridge Peterhouse 191, transcribed and in French translation in Rosier 1994: 291-292). Understanding and the expression occur as if or almost simultaneously, so that the interjection functions linguistically as a part of speech and signifies a concept per modum affectus, with the result that the utterance becomes part of the affective experience: “When an exterior thing (res) causes joy, grief or admiration in someone, it produces in that person a feeling (passionem) which affects the soul of the one who perceives the object or event, and [the feeling] becomes its affect, for example, joy, sorrow or admiration” (Quaestiones, 2.1.1, Sol., MS Cambridge Peterhouse 191, transcribed and in French translation in Rosier 1994: 292). Then, when the soul is affected in this way, the affect is suddenly and at the same time recognized by the intellect (ratio). In the same instant as the affect is perceived by the intellect, the concept without deliberation is expressed as a vox. The (near) simultaneity of feeling (affectus) and vocalization (prolatio) without deliberation or premeditated intention results in an interjection or elliptical clause with interjective force. In urgent or extreme situations, the affect experienced immediately (immediate) leads to an emotionally charged utterance, thus effectively overrunning any explicitly separate stage of concept formation in the intellect. These thirteenth-century grammarians’ analyses of pragmatic speech emphasize how everyday use is communicative in particular contexts even

100 

The Medieval Life of L anguage

when the surface words do not fulfill all the formal criteria for grammaticality or direct reference. Meaningful speech is underdetermined by surface form. As a natural utterance, the Latin interjection Papae [Wonderful, Strange, depending on context] signifies a real and meaningful movement (motus) of the soul. We can expand the utterance as a complete clause (e.g., “Papae, quid video?” or “Papae, iugularas hominem.” [Terence, Eunuchus, line 426]), but when spoken by itself the interjection conveys a heavier emotional or urgent affective load. From a pragmatic perspective, less is more semantically in elliptical speech when used in particular circumstances. It’s also important to note that many of Magister Johannes’ examples of everyday interjections and affective discourse appear to be drawn from Terence’s Latin comedies. Grammarians may have found Terence’s dialogues a convenient source for ready-to-hand colloquial Latin conversation and everyday exchanges. In these thirteenth-century grammarians’ accounts, conventional interjections express an emotional concept per modum affectus (heu, alas) or, as some argue, interjective utterances are part of the feeling itself, performative embodied language and feeling in action. Unlike interjections, the other partes signify per modum conceptum and predicate feeling (doleo, dolor). Heu or alas are said to perform the experience of suffering or lamenting (actus exercitus), whereas doleo signifies by predicating an experiencer who suffers (actus significatus). Incomplete or elliptical utterances also attract affective force and recognizable meaning in extreme or urgent contexts. More than other grammarians, Bacon considered natural interjections (moan, giggle, sigh) as almost equivalent to words in their expressivity and as meaningful alongside words in particular contexts, even if those vocalizations are not susceptible to grammaticalization. What kind of utterance is a giggle? becomes a provocative question for grammatical theory. Medieval analyses of natural and linguistic vocalizations were not inherently pragmatic. Rather, it depended on the theoretical perspective a grammarian adopted. The modistic grammarian Boethius of Dacia (13th c.), for example, used the dolor/doleo/heu example to articulate the relation between the signifying concept ‘pain’ and the different modi comprising syntactic forms and individual parts of speech which consignify ‘pain’: noun (dolor), verb (doleo), participle (dolens), interjection (heu), and so forth (Modi significandi, ed. Pinborg and Roos, 1969: q. 14). The modistae elaborated a theory of grammatical form with realist signification. The thirteenth-century grammarians discussed here adopted a different, more pragmatic perspective on language based on use, context, and intention rather than formal grammar.

Inter jec tions: Does Affec t have Gr ammar?

101

The pragmatic grammarians’ theoretical starting point also highlights some kinds of linguistic phenomena which the modistae’s framework ignores or dismisses. The grammaticality of interjections and elliptical expressions expands the notion of what is included in the field of grammatica. Pragmatic grammarians’ analysis of interjections and nonlinguistic vocalizations depend on interpreters inferencing utterance meaning from sentential, discursive, and extralinguistic contexts. Linguistic contexts, because individual words have conventional meanings and because a given utterance might include information relevant to the interjection’s affective force. Extralinguistic contexts, because the circumstances in which an utterance occurs provide the relevance or at least motivation for both speaking and interpreting the expression. A cry for help (“Aqua. Aqua.,” “Out. Harrow.”) anticipates a real response, not just a personal mental interpretation. Of course, interjective expressions can be misinterpreted, which confirms that such expressions are not transparent or intentionally self-evident. Rather, their meaning depends on listeners not only relying on the spoken words (ad sensum) but construing the expression as relevant and meaningful in some context (ad intellectum). Depending on the context or situation, any expression can have interjective force or affective meaning. It all depends on listeners inferencing and interpreting an utterance’s signification per modum affectus from context, presuppositions, and interactions with speakers.

3

Allas Context Abstract This chapter explores in detail Chaucer’s uses of the interjection allas in various prose and poetic texts. Chaucer’s many uses and contexts of allas suggest how the functions and significations of the Middle English interjection were not stable but rather shift according to who is speaking, where, to whom, in what narrative or textual situation. From the perspective of Bakhtin’s theory of dialogism, Chaucer’s poetic practice gives some concreteness to the grammarians’ theories of interjections, affect, and semantics in interpersonal and semiotic contexts. Keywords: interjection, Chaucer’s poetry, context, dialogism (Bakhtin)

Pragmatics is a theory of language based around user-oriented perspectives on what we say, write, and do. An analysis of pragmatic ideas, practices, and metapragmatic awareness should try and account for not only speech as utterance but also the situations and contexts in which speech is produced. But what is a context? Earlier, I discussed how semioticians and grammarians analyzed the importance of context for interpreting implied or pragmatic meaning. Context can be immediate, situational, or general. Context can be participant expectations or speech/text type. The relations between speech and situation are not causal or determinate but pragmatic. Some are direct and referential; others, indirect and associational. Some are explicit; others implicit. Aspects of what an utterance ‘means’ or ‘does,’ its implicatures, references, signification, entailments, presuppositions, attitudes, perlocutionary force, and so forth, emerge through the interaction of speaker and listener and also from expression or discourse’s relation to one or more contexts or situations, not all of which are linguistic or atemporal. In political and social media discourse, ‘context’ or the claims one’s speech is ‘taken out of context’ are often controversial or manipulated. Historians, literary critics, and linguists, on the other hand, regularly appeal to putting events or words ‘in context.’ Either way, speech does not exist independently

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch03

104 

The Medieval Life of L anguage

of context. Context is all there is for speech to be meaningful. The problem is, what or whose context enables an interpretation, and can there be an authoritative or ‘proper’ context? The medieval concept of virtus sermonis pressures the idea of situational context by establishing Langage itself as the principal context for any utterance. The relation between virtus sermonis and particular social or textual contexts then becomes a matter of negotiation or conflict. Context is defined differently in different disciplinary … contexts. In intellectual history or literary criticism, context is often described as historical, intellectual, literary, or generic. In linguistics, it’s different. Narrowly defined, linguistic context draws on the available options in a language system for producing intelligible utterances. However, pragmatic context is more than the linguistic or grammatical options within a language system. Context is often posited as outside the something which is being contextualized, for example, the occasion of the utterance, but like a sign, context is part of a relation, not a given, co-constructed by a speaker and a perceiver based on presuppositions and inferences. What constitutes the ‘speech event’ or ‘interaction’ is not clearly separate from the ‘context’ in which the event or interaction takes place. Moreover, a relevant context may include circumstances or immediate situation, institutional setting or expectations, discursive presuppositions, and so forth. Discourse or pragmatic contexts depend on participants knowing how to use language to do something directly or indirectly in the world. Hence Austin’s title How to do Things with Words (1962). Depending on where and how we look, we realize some contexts are bigger or more authoritative or more productive than others. Context can be local or global, micro or macro. A micro context includes the immediate speaking situation and the mode of interpersonal exchange and communication, whether informal face-to-face conversation, written text, public sermon, official trial, simultaneous translation, video broadcast, and so forth. Macro context includes institutional or cultural settings, meaning, and social discourses which motivate or legitimize particular speech interactions, for example, genre, register, social status, cultural norms (e.g., patriarchy), language status, and institutions of power and authority (e.g., clergy, judges, teachers, government). Macro contexts include both dominant and alternative ideological formations. Typically, historicism takes ‘context’ to mean broad situations, activities, and often non-discursive elements which function as ‘determinations’ of the meaning of a particular utterance or action. That is, historical context creates meaning for what is somehow assumed to be otherwise meaningless speech. (Is that really possible?) Cultural contexts establish or give positions for participants to

ALL AS CONTEX T

105

take and expectations for meaningful speech in an interaction. Students and teachers know where they are supposed to sit in the classroom. Whether participants fulfill the expectations of their given positions reflects the extent to which a dominant institutional context is being maintained or being challenged or modified. Context can seem to be more or less stable, as habit, convention, or common sense, but not everyone fulfills the expectations for a speaking role at all times. Medieval logicians and grammarians discussed context in several ways, some more textually oriented than others. Abelard claimed the appropriate context for interpreting individual words is the sentence or proposition in which the word(s) appear. Gilbert of Poitiers (c. 1080-1154) developed a broader contextualist theory of interpretation focused on written discourse. In his exposition on the Psalms Gilbert argued that the scope of the utterance to be interpreted varied from a single sentence to the entire psalm. He rejected the posited parallelism between what is said and what is referred to. Instead, he highlighted how the relation between the speaker’s intention and the expression shifted depending on whether the appropriate context for interpreting an utterance was the topic “substantive being” (quod est) or “the quality which makes the concrete being what it is” (quo est); for example, homo or humanitas (Marenbon 2013). Gilbert went further and claimed the different interpretive contexts depend on different kinds of readers. More perceptive readers will not stop with denotation or reference, but where appropriate, will push beyond what is said by virtus sermonis to discover the underlying conditions and abstractions which provide the ground or warrant for meaning. A particular interpretation depends on the reader’s ability to recognize the appropriate context implied by the text’s intention. Peter of Spain discussed context in terms of sentence context and in terms of habitus and dispositio. He argued habitus and dispositio are species of quality, but that in humans the two states are not the same. Habitus is something difficult to change (permanentior), for example, knowledge or virtues (justice, chastity). Dispositio, on the other hand, is a state which is easier to change or something which changes more frequently, such as heat and cold or sickness and wellness (Summulae logicales, ed. de Rijk, 3.21; 1972: 36). Affectus and emotional behaviors would seem to fall somewhere between habitus and dispositio as interpretive contexts. How to interpret the relation between text and context has also been a consistent topic in twentieth- and twenty-first century language philosophy, linguistics, and literary criticism. Cleanth Brooks and other formalist critics have appealed to the “pressures of context” as the ground for interpretation. According to Brooks, a given textual context transforms or ironizes how

106 

The Medieval Life of L anguage

parts of a poem are related to each other: “The context endows the particular word or image or statement with significance” (Brooks 2007: 800). By context, Brooks means the other words or parts of the poem or at least those the reader or critic thinks are relevant, especially metaphor and image. If Roman Jakobson were to read the same poem, he would probably come up with a different set of relevant poetic features which establish ‘the context.’ Brooks’ way of putting the point (“endows”) makes context, which has an indefinite referent, the agent which gives language something. Notice too that Brooks says “significance,” not meaning. That is, textual or linguistic context makes some language salient and poetically loaded (Brooks’ word). In Brooks’ formulation, poetic context produces connotation and true intent rather than denotation. For Bakhtin, however, what counts as a textual utterance varies depending on the sort of analysis or interpretive view of textual relations a reader chooses. Therefore, so does the context of the utterance. As poststructural literary critics, discourse analysts, and some pragmatists have observed, pragmatic context is dynamic, ongoing, always under construction, and therefore sometimes unstable or contested (e.g., Mey 2001: 39-66; Derrida 1982: 307-330). The macro context of a conversation in a culture may be conventional, but participants’ perception of what is going on in the talk can change over the duration of the talk. The same is true for a text. So, the potential fluidity of context can shape how participants respond to or assess what is required for the talk to continue. Grice posits the Cooperative Principle as a foundation for meaningful interaction, but his model presumes that a given context is consistent over the course of an episode of talk. Context is more complicated than that. Schiffrin (1994: 364-369) critiques speech act theory and Gricean pragmatics for overly ‘compartmentalizing’ context of utterance and for separating cognitive elements from contexts of situation with respect to how utterances are produced, analyzed, and interpreted. Both speech act and Gricean approaches treat context as “knowledge” in the sense of shared presuppositions and social rules for utterance, but neither gives focused analytic attention to the context of situation as such. Conversation analysis and ethnography of communication, on the other hand, enlist much more detailed exploration of the interpersonal and socio-political situations in which exchanges occur. CA and CDA will be especially important for structuring our readings of pragmatic interactions in medieval narratives (Chapters 3 through 6). CA and CDA recognize that which context or feature of context is relevant at a particular moment in a discursive situation can vary with respect to how a speaker or listener

ALL AS CONTEX T

107

understands what is going on in the discourse as well as with respect to what knowledge the speaker or listener thinks is relevant and how she interprets the situation of the discourse or utterance. Moreover, a spoken or written text can switch orientation or focalization, genre features or register, presupposition or conditions, and become a hybrid text which challenges the reader to adjust her interpreting or reading expectations. Schiffrin notes, “contexts can impose themselves upon texts (or aspects of texts) to different degrees and in different ways” (1994: 384; cf. 1994: 366-367), but it’s important to keep in mind that contexts do not create themselves. Moreover, as we have seen, participants in an interaction do not necessarily accept or adopt the same contextual conventions for their speech at all times. Context is as much a construction as the talk or text itself. Halliday, following Malinowski, distinguishes micro and macro contexts in terms of context of situation and context of culture. Context of situation, he says, is the immediate situation in which an utterance and interaction take place. The wider context of culture is “the ideological and material culture” which “serve[s] as the source of speech situations and situation types” (1978: 27-31, 68-69). Halliday unpacks the complexity of “situation” with reference to his tripartite model of field (ideational, i.e., experience, from represented action to subject matter), tenor (interpersonal relations in an exchange), and mode (textual organization, including genre). The three nodes form the micro context or situation of an utterance (Halliday 1978: 142-145). Critically, we can rethink the relation between macro and micro contexts in terms of how social habitus is manifested in particular acts and interactions in so far as they maintain or challenge knowledge/power relations in the dominant social order. A critical analysis of pragmatic context, micro and macro, recognizes the discursive situations that maintain or reproduce a society’s dominant strategies for making meaning as well as those other strategies and modes of pragmatic resistance which can emerge in social discourse. When we consider pragmatic context as constitutive rather than wallpaper, some links between medieval grammarians and modern semiotics, critical sociolinguistics, and contemporary deconstruction become more evident. Medieval grammarians argued that interjections and elliptical expressions were meaningful, even if not grammatically “complete.” They made the case for how successful communication depends on speakers (or writers) and listeners (or readers) adopting a relevant context for utterance meaning. The problem comes when speaker and listener adopt different contexts to make sense or don’t share a set of assumptions and knowledge. The grammarians typically focused on relevant contexts of situation, that is, immediate micro contexts such as a burning house (“Aqua. Aqua.”) or

108 

The Medieval Life of L anguage

an unfortunate turn of events (“Heu”). In semiotics and sociolinguistics, Peirce and Bakhtin argue that linguistic expression is meaningful when and because we construe it in a relevant context, but not necessarily the one intended. In other words, context determines where meaning is located for the speaker and the listener. There is no meaning outside a social context because there is no expression outside a context. For Peirce and Bakhtin, discursive context is always social, immediate and macro, which means that control of an utterance context can become contested. Peirce emphasizes how recognizing or discovering meaningful speech depends on abductive processes of inferencing and presupposition. Bakhtin argues the social context of speech is diverse and stratified, principally along lines of authority, prestige, voice, and agency. In “Signature Event Context” Derrida’s theory of dissemination takes us further into “the problem of context” and explores “why a context is never absolutely determinable, or rather in what way its determination is never certain or saturated” (1982: 310). While Derrida’s reading of Austin and pragmatics contrasts with Gricean pragmatics, it realizes some of the wider implications of Austin’s idea of linguistic performativity. Derrida shows that the “concept of writing,” its marked absence, iterability, proneness to drift, and potential for “citational grafting,” manifests the unboundedness and openness effectively implicit in all discursive contexts, especially textual ones: “it is only in a context determined by the will to know, by an epistemic intention, by a conscious relation to the object as an object of knowledge within a horizon of truth” that any speech or writing can be meaningful or capable of signifying. For Derrida, pace Brooks, “there are only contexts without any center of absolute anchoring” (1982: 320). It’s all context. Medieval grammarians and Bakhtin give more weight to speaker intentionality than do Peirce or Derrida, but all strongly critique a linguistic or literary pragmatics which restricts meaning to the notion of an intended, univocal, determinate context for any utterance meaning, whether for a single utterance, a conversation, or a romance narrative. Interjections prompt several questions about context in linguistic and textual formations. Speech is embodied one way or another, and interjections are some of the most affectively loaded kinds of vocalizations in any language. We saw (Chapter 2) that many medieval grammarians characterized interjections as actus exercitus rather than as actus significatus. Interjections do not in themselves communicate a message or a feeling, nor is the context for interjective speech separate from the more general contexts for any speech, spoken or written. Interjections can have different functions and pragmatic meanings in different contexts. They are a linguistic part of an affective performance and mark a speaker’s disposition or attitude toward

ALL AS CONTEX T

109

a referent or addressee, or they can serve as discourse markers to organize and evaluate text topics and information. As meaningful performatives, interjections are oriented to both micro and macro discursive contexts (e.g., register, politeness code, speaker or listener disposition), but it depends on how the expression is made salient as to which context takes priority. Linguists have often argued that interjections and similar affective expressions are common in spoken discourse, but as we shall see, they are also common in many literary texts and genres. Interjections in texts function differently depending on who is speaking in what context.

Allas: A Case for Context Discourse markers such as interjections present a rich case with which to explore the role of context in medieval pragmatic theory and practice. We have already discussed how some grammarians theorized language and meaning in relation to the status of interjections. Looking more closely at medieval textual and spoken uses of interjections can thicken the thread linking medieval grammarians with contemporary critical theorists of language and pragmatics. Given that all our evidence for medieval pragmatics is textual, we can compare how spoken and written interjections function only through the filter of medieval writing. But given that, we also need to attend to the concrete textual situations in which interjections occur. We need nuanced and close reading of micro and macro pragmatic contexts if we are to gain a deeper understanding of pragmatic or discourse markers in medieval texts. Irma Taavitsainen and her Helsinki colleagues have worked with a large dataset and analyzed the use of English pragmatic markers of emotion and interpersonal relations from the Middle Ages to the present. Based on evidence from the Helsinki Corpus for Old, Middle, and Early Modern Englishes, Taavitsainen argues that personal affect, “the expression of subjective emotions, feelings, moods and attitudes,” is encoded linguistically in interjections (including expletives and curses), first-person pronouns, private verbs of emotional states, and direct WH information-seeking questions (Taavitsainen 1997; cf. Taavitsainen 1994, 1995 and Jucker and Taavitsainen 2013: 64-69). In a study of Chaucer’s Canterbury Tales (1995) Taavitsainen concludes that Chaucer uses involved, emotive utterances in different literary genres to express different emotional states. Interjections are some of the most involved, emotionally loaded kinds of utterances in oral and written interactions. Taavitsainen claims the distribution of these

110 

The Medieval Life of L anguage

and other feeling expressions varies in Middle English literature primarily according to genre (macro) context. She and Jucker go so far as to claim that Chaucer’s use of interjections and other exclamations in his writing “developed genre-specific vocabulary” (2013: 61). For Taavitsainen, genre is the determining or most important context for textual meaning. Biber et al. (1999) similarly analyze interjections and pragmatic markers primarily in terms of genre and register. My approach to Middle English discourse or pragmatic markers of affect, context, and textual organization is different. When analyzing context and distribution, frequency counts, but so too do concrete instances of how interjective expressions work in pragmatic contexts. In the rest of this chapter, I present a case study of the Middle English interjection allas in Chaucer’s (c. 1340-1400) writing and explore its relation with other interjections and pragmatic markers in various textual contexts. First, I outline the grammatical and pragmatic uses of Middle English allas as an affective expression and discourse marker. Then I discuss Middle English allas and other pragmatic markers in micro and macro contexts in Chaucer’s texts. (Allas is the most common Middle English spelling found in Chaucer manuscripts.) As will become clearer, genre is only one of many contexts that present themselves, and sometimes not the most important one for a spoken or written discourse. Interpreting pragmatic meaning and how it emerges means we need to account for both micro and macro contexts of utterances and the relations between them in any interaction. Moreover, we need to deploy a flexible notion of the context for any given usage and let the textual situation guide our reading. As an English interjection, allas is formally ambiguous, neither nonhomophonic nor a natural vocalization. It is both primary and conventional, part of the grammatical system but semantically empty with respect to reference. Allas is different to many other primary interjections in that it has a range of conventional connotations and implications associated with it. The MED gives the definition “An exclamation expressing grief, pity, shame, apprehension, or the like.” The OED defines the word as an archaic interjection “used to express grief, pity, regret, disappointment, or concern.” The most recent edition of the American Heritage Dictionary (2011) adds “compassion” and “apprehension of danger or evil.” However, the utterance doesn’t signify or embody a particular feeling; it participates in pragmatic signification in specific discursive or textual contexts. That is, as medieval grammarians noted, expressing allas does not always refer to a particular feeling, but rather is part of an embodied feeling, an exteriorization into language of an internal state which is inferred only from context. A range

ALL AS CONTEX T

111

of possible feelings are associated with allas, as the dictionary definitions suggest. Pragmatically, who is speaking, to whom, how, and in what context are key parts of interpersonal pragmatic meaning and constitutive of the dialogism of textual situations. Many of the uses of Middle English allas in the texts discussed here occur as direct dialogue or reported speech; some occur in narrator or authorial discourse. Whether allas is located in direct, reported, indirect, or narrational speech also inflects its significance and function as a pragmatic marker. To twenty-first century ears, English allas may sound familiar but perhaps stagey, old fashioned, overly dramatic. We might imagine the expression accompanied by a gesture such as a limp forearm across the forehead or a woeful sigh. In context, the conventional English utterance is associated with sorrow, despair, or regret and equivalent to the Latin conventional interjection Heu, regularly cited by Late Latin and medieval grammarians as a signifying vocalization of sorrow. One of the most recognizable Middle English occurrences of allas is the Wife of Bath’s lament, “Allas! Allas! That evere love was synne” (3.614. Riverside Chaucer editors’ punctuation.). More on her speech later. As a conventional interjection, the form and general senses of allas were relatively stable from 1100 to 1500. Common in Middle English speech and writing, use of the interjection has declined steadily since 1800 (cf. Jucker and Taavitsainen 2013: 68-69). The two predominant spellings were alas and allas, occasionally helas. The scribes of Chaucer’s texts regularly spell the interjection allas. Middle English allas was adopted in the thirteenth century from Old French ha las (ModF helas, ‘Ah’ + ‘wretched’ or ‘weary’). The Anglo-Norman Holkham Bible Picture Book (c. 1320-1330) spells the French word “allas” as well (1971: 20). The word is first attested in English around 1230 in Ancrene Wisse (according to the MED and OED). The expression appears with varying frequency in a wide range of Middle English prose and poetry texts: Pearl, Purity (but not Sir Gawain and the Green Knight), Piers Plowman, Chaucer’s Canterbury Tales and Troilus and Criseyde, South English Legendary (c. 1270-1280), Cursor Mundi (c. 1300), Wycliffite English Bible (c. 1380-1400), Anglo-Norman Gui de Warewic (early 13th c.), Metrical Chronicle of Robert of Gloucester (c. 1270-1300), Malory’s Morte d’Arthur (1485) and numerous Marian poems and hymns (based on CMEPV data). Uses of allas in Middle English texts occur in a great number of genres, especially fabliau, romance, popular drama, sermons, Biblical narratives, hymns to the Virgin, and personal letters, but not, so far as I can tell, official documents and records. Among long texts in Middle English, Chaucer’s poetry, Malory’s Morte d’Arthur, and the fifteenth-century York cycle plays

112 

The Medieval Life of L anguage

record the highest total uses of allas (based on CMEPV data). Given these multi-genre distributions, Middle English allas does not easily conform to the generalization that interjections are primarily spoken, informal, conversational kinds of utterances. Middle English allas shows the complexity of medieval pragmatics in practice. The expression works as an interjection or as a discourse or pragmatic marker or sometimes both in grammatical and discursive contexts. The uses of Middle English allas can be grouped as follows. Most examples are from Chaucer’s texts: 1. solo allas: the conventional word is uttered by itself as a complete expression, or is combined with a noun (“Allas, my wyf!”; CT, 1.3522), phrasally with to or for (Allas for sorrow), or clausally followed by a that clause (“allas, that I was bore!”; CT, 6.215) (cf. Magister Johannes on perfectio ad sensum). Middle English allas combines with verbs of speaking (crien, saien, quod) as an explicit and attributed expression of grief or pity: “… ful ofte he seyde, ‘allas!’” (CT, 1.1073). In many contexts, solo allas functions to shade the immediately preceding or following phrase or clause with negative affect: “Allas, when shul my bones been at reste” (CT, 6.733). Shading occurs when a word or phrase distributes an associated attitude or affect to an adjacent construction, and the precise feeling or attitude is interpreted from context. Solo allas also functions as a negative evaluation or judgment of the immediately following or immediately preceding phrase or clause: “I was aboute to wedde a wyf; allas!” (CT, 3.166), “Allas, Pompeye, of th’orient conquerour, / That Fortune unto swich a fyn thee broghte!” (CT, 7.2693-2694). 2. reduplicating allas: the word is immediately repeated to connote urgency or intense emotional distress (cf. medieval grammarians’ “Aqua. Aqua.”): e.g., “Allas! Allas! That evere love was synne” (CT, 3.614), where the construction functions as an affective intensifier. Middle English reduplicating allas also serves other discursive functions, especially as a meditative trigger to evoke empathy at Christ’s crucifixion in English hymns to the Virgin and in the late fifteenth-century Marian poem De arte lacrimandi. Prosopopaeia B[eatus] Virginie, many of whose stanzas end with the refrain “Alas, alas, come wepe with me” (BL, MS Harley 2274, fols. 35a-46b; cf. McNamer 2010: 126-127). The Latin Bible provides models for double, even triple allas. The Wycliffite English Bible, following the Latin text, uses allas in Old Testament books to heighten expressions of despair: e.g., “And the kyng of Israel seide,

ALL AS CONTEX T

113

‘Alas! alas! alas! the Lord hath gaderide us thre kyngis to bitake us in the hond of Moab’” (4 Kings, 3.10; WB/F-M, 2: 241; Vulgate version: “Heu heu heu, congregavit nos Dominus tres reges …”). 3. discontinuous allas: the word is repeated at the beginning and end of a passage to frame the enclosed discourse with negative affect or evaluation: Allas, mankynde, how may it bitide That to thy creatour, which that the wroght And with his precious herte-blood thee boghte, Thou art so fals and so unkynde, allas? (CT, 6.900-6903) Allas, Fortune! She that whilom was Dredeful to kynges and to emperoures, Now gaureth al the peple on hire, allas! (CT, 7.2367-2369) Alternately, allas can be repeated discontinuously as an affective cluster throughout a passage or speech to encode negative affect in the text or highlight the speaker’s distress (e.g., TC, 3.802-808; CT, 1.2771-2776). 4. formulaic allas: the word combines or clusters with other Middle English interjections, especially weylaway and harrow: “Cryinge, ‘Allas, and weylaway!’” (HF, 170); “Allas and weylawey!” (MT, 1.3602). The interjective clusters “‘Out! Help! Allas! Harrow!’ he gan to crye” (CT, 4.2366) and “Harrow! Allas! Heere lith my felawe slayn!” (CT, 7.3045) are Middle English formulas for raising the public alarm.

Formulaic allas raises the question whether Middle English allas is synonymous with weylaway or harrow. Some contextual evidence suggests the three expressions are synonymous in Middle English. In the utterance “So weilaway, whi nyl myn herte breste?” (TC, 4.580), weylaway seems to substitute for solo allas. Sometimes, allas and weylaway combine together to form a variant of discontinuous allas which frames a particular passage with negative affect or evaluation: e.g., “Allas! Custance, thou hast no champioun, / Ne fighte kanstow noght, so weylaway!” (MLT, 2.631-632); “And he seyde, ‘Allas, Fortune! and weylaway!’” (CT, 7.2445). Some Chaucer glossaries (e.g., Davis et al. 1979) annotate formulaic allas with a lexical sleight of hand, which obscures the linguistic complexity. In the Hous of Fame, Aeneas carries Anchises on his back as they are escaping Troy, “Cryinge, ‘Allas, and welaway!’” (HF, 170). Davis defines weylaway as alas!

114 

The Medieval Life of L anguage

(complete with the editor’s punctuation), but he does not include allas itself in his glossary, perhaps because the interjection is lexically continuous to Present Day English. (Davis might have a point. Anecdotally, between 1 January and 15 May 2020, alas appeared in at least one article per week in the Guardian (UK) online.) The Riverside Chaucer editors also gloss weylaway as alas, complete with the exclamation point (e.g., CT, 1.938). Occasionally they even gloss harrow as alas (e.g., CT, 6.288). So, in Chaucer glossaries allas is taken as the anchor term for a cluster of pragmatically similar interjections. However, not all contexts are equal. Reduplicating allas is never used in Middle English texts to raise the alarm or urgently ask for help. Neither is weylaway. The formula allas and weylaway is always used in Middle English texts to express distress and a sense of loss, perhaps in a more conversational register (see #5 below). 5. flavor allas: given the word’s conventional consignifications of sorrow, regret, or despair, Middle English allas can be used performatively to broadly suggest a tone or attitude for a particular experience or state rather than performatively signify a particular feeling. The Wife of Bath’s lament, “Allas, allas! That evere love was synne!”; CT, 3.614) functions in this manner in addition to functioning as an affective intensifier, similar to shading but more generalized (see #1, above). We can imagine, for example, allas allas pronounced with a slower cadence and more lengthened vowels in the second syllables, similar to the way exciting O is pronounced when clause initial as a marker of topic shift or apostrophe. Pragmatically, the reduplication and the cadence add a flavor or shade of regret or loss to the utterance it precedes.

Many linguists have interpreted interjections as more informal and conversational kinds of spoken language, slightly more word-like than vocalizations such as Ah (“… he bleynte and cride, ‘A!’ / As though he stongen were unto the herte”; CT, 1.1078-1079) or Tehee (CT, 1.3740). However, Middle English solo, discontinuous, and formulaic allas occur not just in dialogue but in many textual contexts, in many of Chaucer’s poetic and prose texts, and in many genres, including fabliau, romance, dream vision, sermon, and other genres. Allas signals a negative disposition or state, whether solo or syntactically connected with another utterance. In different contexts allas functions to call attention to the speaker’s negative feeling about or evaluation of the circumstances which form the discourse topic. Discursive shading is a key pragmatic and syntactic feature of interjections and discourse markers, including allas. The distinction between the interjection as a word class and discourse markers becomes blurred

ALL AS CONTEX T

115

when we consider allas in various contexts. Interjections are said to be semantically and propositionally empty, but as we have seen in medieval and contemporary theory, that linguistic analysis, while true for individual words, misses the discursive and pragmatic point. Propositionally and lexically empty, yes, with respect to semantic reference, but pragmatically performative and dispositional and therefore significant and discursively meaningful as part of a context. Clustered allas, whether reduplicating or discontinuous, is a performative utterance which shades an adjacent utterance with heightened negative, mournful, or regretful emotion. When clause initial, allas usually shades the following utterance with a negative aspect, as in Arcite’s speech in the Knight’s Tale: Allas, the deeth! Allas, myn Emelye! Allas, departynge of oure compaignye! (CT, 1.2773-2774)

The repetition of the interjection is a recognizable rhetorical device for intensifying the attitude of sadness, regret, or despair, in Arcite’s case all three. When clause final, allas shades the previous utterance more as the speaker’s evaluation of the utterance. When initial or final, the interjection effectively takes discursive control of an adjacent utterance and shades it with negative, mournful, or regretful affect or evaluation regardless of what meaning that word or phrase might have by itself. In addition to functioning as a discourse marker, interjective shading also plays a role in textual pragmatics. Allas can orient the reader to a topic or register shift or to a metacomment in the text. The range of negative emotional states or feelings of allas reflected in the MED, OED, and American Heritage Dictionary definitions (excluding pity and compassion) clearly indicate that a pragmatic or discourse marker can have a discernible cognitive and/or emotional valence precisely because it is a conventional word or vocalization in a grammatical system and therefore an expression which speakers understand with pragmatic inferencing. Sometimes, as we shall see, allas in a text can function to elicit or trigger empathy or pity from the audience for a particular character or situation. In all cases, allas can have one or more discursive functions whose effect is only inferred from the immediate textual context as constructed by the intersection of speaker intention and audience interpretation. The mournful tone or urgent accent with which Middle English allas might be pronounced in everyday speech or when read aloud is typically marked by modern editors with an exclamation point, as we have seen in the Riverside Chaucer editors’ regular punctuation “Allas!” When allas is combined with weylaway, editors of Chaucer’s writing usually punctuate

116 

The Medieval Life of L anguage

the entire formula with an exclamation point (see CT, 2.810). The editors’ contemporary punctuation choices reflect their interpretations of particular interjections and their contexts (cf. Fulton 1996). Medieval scribes, however, did not usually mark the interjection that way. In fact, scribes didn’t mark interjections much at all, although Parkes observes that the punctus exclamativus did begin to appear in manuscripts in the second half of the fourteenth century. Parkes’ examples are fourteenthcentury Italian manuscripts or books printed after 1500 (Parkes 1992: 213-227, 306). Early manuscripts of Chaucer’s writing (Hengwrt manuscript, Nat. Lib. of Wales, Peniarth MS 392D, late 14th c.; Ellesmere manuscript, San Marino, Huntington Lib., EL 26 C 9, early 15th c.) do not punctuate allas or otherwise inscribe the word differently to the surrounding words (see http://www.sd-editions.com/AnaServer?HengwrtEx+0+start.anv; Ellesmere Manuscript 1989 [facsimile]). As performative utterances, allas and many other Middle English interjections, both homophonic and nonhomophonic, exteriorize or participate in the embodied enactment of a feeling, sorrow, regret, despair, negative evaluation, depending on the implicating context. Textual context is what gives allas a particular emotional slant. When functioning as discourse markers, interjections can frame or shade syntactic sequences (phrase, clause) and are understood as textually enacting or performing a speaker’s affective stance toward someone or something. That understanding is based on the audience’s inferencing and shared knowledge of linguistic conventions and textual contexts. In this way, grammatically and pragmatically, allas and other interjections mark out subjective emotional space within the stream of spoken or written discourse. To grasp in more detail how Middle English allas functions pragmatically, I turn to some close readings of allas in Chaucer’s writing. The Middle English interjection’s conventional meanings and grammar constitute the textual parameters in the field in which characters, narrators, and author (function) generate in poetic practice conventional, different, or more specific pragmatic and narrative meanings. My readings of allas in particular contexts address several interpretive and critical questions: Does allas function more as an interjection or discourse marker in micro textual situations? What can we infer about the textual construction of speakers’ (characters, narrators) motives for using affect speech? What is the relation between textual intention and speaker eruption (recall grammarians’ erupta) with interjections? If allas is meaningful in a context, what roles does the performative utterance play in the construction of narrative context?

117

ALL AS CONTEX T

Allas occurs in Chaucer’s writing in a wide range of text types: romance, dream vision, sermon (and sermonic), exemplum, fable, saint’s life, tragedy, moral allegory, fabliau, and lai. In this chapter I focus on Troilus and Criseyde, Knight’s Tale, Miller’s Tale, Reeve’s Tale, Wife of Bath’s Prologue and Tale, Merchant’s Tale, Pardoner’s Prologue and Tale, Monk’s Tale, Nun’s Priest’s Tale, and Parson’s Tale. However, the frequency and distribution of Chaucer’s uses of allas do not fall out clearly along genre lines (contra Taavitsainen 1995; Jucker and Taavitsainen 2013: 61). Rather, allas occurs at particular moments in different narrative genres as part of particular textual voices. The frequency, distribution, and meanings of Chaucer’s uses of allas vary according to micro textual contexts. Genre is only one feature of context and not necessarily the most important. The specificity of context is what makes Chaucer’s uses of allas complex and pragmatically nuanced. To begin, consider how allas functions in two of Chaucer’s romances, the Knight’s Tale (Canterbury Tales) and Troilus and Criseyde. In the Knight’s Tale, allas is used 23 times, more than in any other of the Canterbury tales. Twenty instances occur in the speech of the lovers Palamon and Arcite and once each in Emily’s speech, the Thebans’ speech in the temple of Venus, and the narrator’s discourse (Fig. 1).6 Interestingly, Duke Theseus, who attempts to impose order on the chaos that is the narrative action, does not use allas at all. In terms of textual contexts, more than 50% of the uses of allas in the Knight’s Tale cluster in Arcite’s speech, the disappointed lover and then dying tournament victor. Half of Arcite’s uses of allas occur in the short space of five lines in his dying lament (CT, 1.2771-2775), amplifying his sadness through sheer repetition. In lines 2771-2775, Arcite’s allas in initial syntactic position enacts his sorrow that his hopes and expectations have been knocked down, although he doesn’t know it’s partly the gods’ doing. His lament also sets up one component of a significant intertextual reference between the Knight’s Tale and Miller’s Tale: Table 3.1  Allas in Chaucer’s Knight’s Tale Total tokens Palemon Arcita Emily Folk in Venus Temple Narrator

23 6 14 1 1 1

6 In this analysis, word counts and line counts for Chaucer’s individual texts are based on The Riverside Chaucer, ed. Larry Benson, 3rd ed. (1987).

118 

The Medieval Life of L anguage

“Allas, my wyf” (CT, 1.2775, 3522). Emily’s single use of the interjection expresses her distress or at least her reluctance at having to marry at all. Functionally, speakers in the tale use allas to convey general negative feelings (lament, despair) or evaluations, but each character’s use of the interjection indexes a feeling or judgment of their own situation at that moment in the narrative. A good example in the Knight’s Tale of how conventional allas is motivated by a specified context of situation is the first use of the interjection in the text: This sorweful prisoner, this Palamoun, Goth in the chambre romynge to and fro And to hymself compleynynge of his wo. That he was born, ful ofte he seyde, “allas!” (CT, 1.1070-1073)

Line 1073 combines the syntactic strategies of allas that and said allas. Palamon is lamenting his imprisonment but is just about to see Emily for the first time (which makes him suddenly cry out “A!” with surprise and complex emotional pain/desire). In this tightly woven passage, the narrator’s descriptive vocabulary of despair (“sorweful,” “compleynynge,” “wo”) works almost like a definition of the interjection and introduces the disposition and affective context for Palamon’s single spoken allas. At line 1073 allas punctuates the narrative’s indirect representation of feeling with the speech of despair. Arcite’s early uses of allas similarly lament his circumstances, but the passages are less condensed, more prolix (CT, 1.1223ff.). The conventional meaning of allas is further specified, that is, actualized, by particular subject positions in the tale. Allas spoken by the male knights does not affectively or textually signify in the way allas spoken by Emily signifies. This is pragmatic meaning in a nutshell. When we account for distribution, the two male knights/lovers are distinguished from the other speakers in the tale by their extensive use of the lamenting interjection and then differentiated from one another in the text by their uses of the interjection to perform despair in different contexts with different motivations. As the ultimate male loser in the chivalric romance narrative, Arcite may have the most reason to mourn. Accordingly, he speaks allas more than any other character. Emily’s single utterance not only suggests that her agency is more restricted than either Arcite or Palamon’s, but her sole expression in the Temple of Diana also suggests how she can be read as struggling against the marriage plot in which she is objectified. When we compare the Knight’s Tale with Troilus and Criseyde, a more elaborate romance, we discover some interesting relations in the frequency,

ALL AS CONTEX T

119

distribution, and micro contexts of allas. TC has 8240 lines (not counting incipits and explicits); KT, 2250 lines. Allas occurs 122 times in TC, 23 times in KT. TC has almost four times as many lines as the KT and almost four times as many allas occurrences. In terms of allas frequency, the two texts are similar. In TC Troilus uses allas 49 times, Criseyde 45 times, Pandarus 18, the narrator six, and other characters (Deiphebus, Antenor, Diomede) five in total. Clearly, allas belongs to the lovers in Troilus. In both the Knight’s Tale and Troilus and Criseyde, allas predominates in the last third of the narrative, although in the Knight’s Tale the uses of allas are concentrated in a specific passage. However, whereas allas in the KT is associated with the interpersonal relation between the male rivals for Emily, allas in TC pertains to the love relation between Troilus and Criseyde. Criseyde’s public and introspective speech, and not only her use of allas, is emotionally expansive, ruminating, complex, and sometimes contradictory, unlike Emily’s much more restrictive characterization. Criseyde considers and reconsiders her decisions and her fate in the romance and expresses her distress about her reputation, both in Books Two and Three with respect to her social position, the rumors, and the intrigue surrounding her relationship with Troilus, and then in Book Five with respect to her despair at her separation from Troilus and growing fear she will be publicly condemned for being false to her lover. Troilus and Criseyde’s allas uses frequently cluster around set speeches: apostrophes, aubade, complaints against Fortune, and Criseyde’s despair over her reputation. In these set speeches allas also combines with other affective speech, including exciting Os and apostrophes. Books Four and Five have more than 50% of the allas occurrences in TC. Not surprising given that the interjections index the narrative’s relentless forecasting and then direct narration of the pain and tragedy unfolding for the lovers. Also, allas in Books Four and Five often link Troilus and Criseyde’s laments for their imploding relationship with the narrator’s despair at where things are headed. The uses of allas in TC become especially complex in terms of the relation between interpersonal or character discourse and narrational discourse. For instance, in one stanza (5.708-714) the text combines spoken allas with the narrator’s depiction of Criseyde’s growing sense of loss and isolation after she has been traded to the Greeks for Antenor: Ful pale ywoxen was hire brighte face, Hire lymes lene, as she that al the day Stood, whan she dorste, and loked on the place Ther she was born, and ther she dwelt hadde ay;

120 

The Medieval Life of L anguage

And al the nyght wepynge, allas, she lay. And thus despeired, out of alle cure, She ladde hire lif, this woful creature. (TC, 5.708-714)

As in the Knight’s Tale, the TC stanza surrounds the single allas with the descriptive vocabulary of the affect it indexes: “wepynge,” “despaired,” “woful.” But the stanza blurs the distinctness of voices in the narrative discourse. The narrator’s voice depicts Criseyde in the daytime standing and looking at what she yearns for but can’t have – Troy/Troilus (5.708-711) – and then in the night weeping and mourning her situation (5.712-713). In line 712, without punctuation, allas is double voiced, spoken by Criseyde as she weeps and by the narrating voice which sympathetically evaluates Criseyde’s situation. The syntax lets the interjection drift between two narrative focalizations, the immediate diegetic and the metanarrative. In this stanza allas floats as indirect free speech, unanchored to a single speaking subject or position. The floating allas aligns Criseyde’s perception of her situation with the narrator’s. Or is it the narrator’s perception with Criseyde’s? The double voicing of allas indexes both the character’s and the narrator’s understanding that they do not have control of their own or the narrative’s outcome. The problem is, as book readers we know the alignment of these viewpoints in the text is constructed by some figure we call “Chaucer.” We know textual composition doesn’t necessarily have to go one way, so what does allas signify or textually contribute when “Chaucer” creates a text which complicates the narrator’s relationship with the characters and narrative sources and also with the imaginative power of narrative to represent events and experience? Just what is it about narrative fiction and textual historicity that makes “Chaucer” say “Allas”? The context for allas in this passage is indeterminate. Chaucer’s historical and sermonic texts use allas differently than in the romances. In this section I compare uses of interjections in the Monk’s Tale, Pardoner’s Prologue and Tale, and the prose Parson’s Tale. The Monk’s Prologue and Tale (877 lines) recounts the tragic falls of high-status men and women (based on Boccaccio’s De casibus virorum illustrium). In the prologue and tale, allas occurs fourteen times: eight in the narrator Monk’s discourse, two in the pilgrim narrator Chaucer’s speech, two in the Host’s (once directly insulting the Monk and once reporting his wife Goodelief’s speech), and four utterances by characters in the Ugolino episode. In the KT and TC allas occurs mostly in the characters’speech, but in the MkT allas is mostly spoken by the narrator Monk as he evaluates and affectively responds to the fates of the legendary and historical figures whose tragic falls he is

ALL AS CONTEX T

121

narrating. In the MkT, allas belongs to the fictional pilgrim storyteller and suggests his personal but limited involvement with why and how the tragic figures he narrates met their ends. The Ugolino episode is the only one of the Monk’s short stories in which characters use allas: Ugolino, three times and his son, once. In this passage Chaucer seems to be following his partial source, Dante’s Inferno (33.1-90), in the use of allas. Ugolino mournfully utters allas three times at the beginning and end of a single stanza as he laments his family’s imprisonment and starvation (7.2429, 2445). Ugolino’s discontinuous allas frames his lament. His young son then follows (7.2450), but the son’s allas indexes his distress when he believes his father has begun to gnaw his own arm for lack of food (“… ‘Fader, do not so, allas!’”), when in fact Ugolino is biting his arm in a conventional gesture of sorrow (7.2447-2448). While Chaucer makes several changes to Dante’s vocabulary and narrative representation, his use of allas echoes Dante’s text in which Ugolino uses the Florentine (northern) Italian interjection ahi to shade his tragic wish that he and his starving children just die quickly and then the narrator’s condemnation of the Pisan city fathers for how they have treated Ugolino (33.66, 79). Dante’s first use of ahi corresponds somewhat to English alas as an expression of Ugolino’s despair. Dante’s second use of ahi marks the narrator’s explicit negative judgment of Pisa’s citizens. Chaucer’s text compresses Dante’s version and inserts more affective and personal solo allas into the Ugolino episode. The focalization of pathos in the Ugolino episode differs from that in the Monk’s other stories. In the other stories, the narrator Monk uses allas to express his sorrowful response to the tragic stories he is relating. In the Ugolino episode, allas heightens the pathos from the characters’ point of view rather than from the narrator Monk’s. In the Monk’s Tale we can see how interjections and pragmatic markers such as allas function differently in different discursive contexts and acquire pragmatic meaning in the context of particular utterances, depending on who is speaking in the text. Genre alone cannot fully determine the meaning and function of a pragmatic marker. The Host in the Monk’s Prologue uses allas twice. First, to lament how the Monk’s sexual potential has been overtaken by his vocation – maybe, but the General Prologue suggests otherwise. Second, to represent a domestic context as a counterpoint to the ideal image of celibate male religious. Harry imagines what he thinks his wife would say if her husband ever acted like a milktoast (“‘Allas,’ she seith, ‘that evere I was shape / To wedden a milksop, or a coward ape …’”; 7.1909-1910). The Host’s speech provides a strong contrast to the register and tone of the Monk’s tragic stories and also to that in the

122 

The Medieval Life of L anguage

romances. The Host’s use of allas constructs an image of his wife’s speech which in fact supports his self-image as a man’s man. The Host’s reply to the Monk suggests there are many differences among Middle English allas contexts and the meanings they implicate, especially when we pay attention to who is speaking in a given situation. The Parson’s Tale (1016 prose lines) is an academic-style prose treatise on sin and penance which incorporates some features of the thematic sermon or moral treatise (Wenzel 1984; Patterson 1978). Chaucer’s text is based on thirteenth-century Latin treatises on sin and penance. Allas occurs eleven times in the text in emotionally heightened contexts, all in the Parson’s voice (cf. Taavitsainen 1995: 195). The prose of the Parson’s Tale reveals complex internal syntactic and discursive contexts which suggest how the discourse marker allas acquires meaning through its relation to grammar. For example, the Parson condemns anger (ira) in the following passage: And in his outrageous anger and ire – allas, allas! – ful many oon at that tyme feeleth in his herte ful wikkedly, bothe of Crist and eek of alle his halwes. / Is nat this a cursed vice? Yis, certes. Allas! (10.558-559; Riverside Chaucer editors’ punctuation.)

Chaucer uses reduplicating allas, allas just once in TC and five times in the entire CT (MilT, WBProl, MerT, FrankT, ParsT). In the Parson’s Tale passage reduplicating initial allas, allas marks especially intense emotion and the speaker’s strong negative response to the angry man in his exemplum. The third allas in this cluster then reshapes all three uses as a discontinuous evaluative frame around the Parson’s characterization of hateful anger. The syntactic position of allas, allas can still be considered as clause initial because “And in his outrageous anger and ire” functions more like a sentence adverbial or adjunct to the main clause (“ful many oon … feeleth …”). Reduplicating allas is inserted into the syntax as the writer/speaker’s negative evaluation of some people’s wickedness. Taking together the twenty-five uses of allas in the Monk’s Prologue and Tale (narrative poetry) and Parson’s Tale (sermonic prose), we find allas is used in syntactically disjunct ways, both as an interjection and as a pragmatic or discourse marker. The expression usually conveys the speakers’ sorrow or despair or negative evaluation of their own or others’ circumstances. However, not all moral texts in the Canterbury Tales use allas in the same way or as frequently. In the prose moral allegory Tale of Melibee, allas is used just twice, once by Melibee and once by the narrator, to express in fairly generic fashion the speaker’s sorrow in the immediate speech context. It seems Chaucer makes calculated

ALL AS CONTEX T

123

stylistic choices when to use allas in textual contexts, most of which do not seem to be determined by genre. The Parson’s Tale affords us interesting linguistic and pragmatic information about Chaucer’s uses of interjections and discourse markers throughout his writing. From the point of view of textual pragmatics, the relation between the Parson’s Tale and its sources suggests that Chaucer may have considered allas to be a highly effective kind of utterance or perhaps an especially relevant one for English audiences. Allas occurs eleven times in the Parson’s Tale, nine times in the third section devoted to describing and condemning the seven deadly sins. Just as Chaucer added affective allas to his reworking of Dante’s Ugolino episode, so he added allas to his sources for the Parson’s Tale. None of the allas instances has a Latin equivalent in any of the tale’s closest analogues: Raymound of Pennaforte’s (OP) Summa de paenitentia (1222) and Summa casuum paenitentiae (1229) or William Peraldus’ (OP), Summum vitiis (1236) (cf. Summum Virtutum de Remediis Anime, ed. and transl. Wenzel 1984). It’s possible Chaucer may have used one or more intermediate versions of these penance treatises and not only Peraldus. Still, Wenzel thinks that the Parson’s Tale is “… closely akin to or even dependent on similar pastoral treatises in Latin” and that those connections account for “the verbal texture of the Tale in other specific instances.” In particular, he says, “It is accurate to say, I believe, that in writing [the Parson’s Tale] Chaucer translated substantial sections from the identified sources. But it is equally accurate to point out that in doing so he not only worked selectively but also made changes and additions which reveal intelligence, purposiveness, and a fairly exact familiarity with the pastoral theological thought and language of his time” (1981-1982: 252). If Wenzel is right, then Chaucer probably added allas to particular passages as a dramatic form of sermonic discourse for Middle English audiences and as a sermonic or rhetorical rather than conversational strategy to affectively heighten without irony the Parson’s condemnation of sin. Chaucer uses allas as special affective pragmatic speech. Allas clearly marks the Parson’s negative evaluation of sins with a tone of regret and painful sorrow for humanity’s failings. For example: Allas, wel oghten they thanne have desdayn to been servauntz and thralles to synne, and soore been ashamed of hemself / that God of his endelees goodnesse hath set hem in heigh estaat, or yeven hem wit, strenghte of body, heele, beautee, prosperitee, / and boghte hem fro the deeth with his herte-blood, that they so unkyndely, agayns his gentilesse, quiten hym so vileynsly to slaughtre of hir owene soules (10.151-153).

124 

The Medieval Life of L anguage

Interestingly, this passage is followed immediately by another exhortation against moral turpitude (“stynkynge ordure of synne”; 10.156) which begins “O goode God …” Like allas, rhetorical O as an interjection elevates the speaker’s discourse register to a higher level of assumed public authority. But exciting O can also function as a discourse marker to introduce a topic shift (textual information management). In this context, O associates the following speech with negative evaluation. Allas, on the other hand, conveys the speaker’s distinctly negative or regretful orientation to the topic it is associated with. Textually, the discourse marker allas in 10.151-153 solicits the reader’s agreement and similar judgment of the sins described in the text. The Parson’s allas signals a negative evaluation of sinful behavior which the reader should also acknowledge and approve of. The pragmatic function of Middle English allas as an evaluator and as an affective expression becomes clearer when we compare allas in the Parson’s Tale with how it’s used in the Pardoner’s Prologue and Tale. As I said, it’s likely Chaucer added Middle English allas for strategic effect to his translation(s) of his Latin source(s) in the Parson’s Tale as well as his adaptation of Dante’s Ugolino episode in the Monk’s Tale. The Pardoner’s Prologue and Tale and the Parson’s Tale share the theme of sin and penance, but the Pardoner’s Tale is different discursively and rhetorically. The Parson’s Tale presents the theme of sin and penance in a decidedly written prose and academic register, with sections and subsections (distinctions). The Pardoner’s Prologue and Tale delivers an exemplum in the style of an oral thematic sermon in a more conversational register. While the texts share a general theme or topic, they deploy different generic forms and registers. Moreover, the Pardoner’s Prologue and Tale is a verse narrative which ironically complicates the dialogism of the text and narrating voice. The pragmatics of textual polysemy in the Pardoner’s Prologue and Tale needs some sorting out, and Bakhtin and Voloshinov’s theory of dialogism helps. Dialogism foregrounds how all utterances are interrelational and all meanings or signif ications are interactive, whether as cooperation or contest. Bakhtin’s theory of dialogism is founded on fundamentally pragmatic principles in both speech and writing. In interactions and in a discursive field, words do not belong only to the intending speaker or to a single or presupposed signifying context or focalization (point of view). A given utterance has a position or stance because it is embedded in a field of speech forms: official, unofficial, formal, informal, artistic, everyday, straight, parodic. Dialogism characterizes the discursive field as a complex relation of centripetal and centrifugal forces, a top-down push by ‘authoritative discourses’ and coercive speech seeking conformity,

ALL AS CONTEX T

125

control, and monolingual speech and one or more pulls of ‘internally persuasive discourse’ or counterhegemonic speech which is effectively more open, negotiating, and polylogic (see especially Voloshinov 1973; Bakhtin 1981: 259-422). Bakhtin’s theory of dialogism is linguistic and semiotic, implicitly political, and pragmatically relevant. Every utterance, every speech act is tied to a response or potential response which may or may not conclude (“finalize”) the interaction. Linguistic interactions and linguistic practice in a community inherently involve the relation between centripetal and centrifugal speech acts and pragmatic strategies. Local acts of finalization give shape and force to communication and sociability, but those acts are always contingent. A respondent’s uptake is another utterance. Depending on context or the degree of centripetal force, we can always return to or continue the interaction and dialogue, rethink it or respond again, reimpose, whether at the immediate or local level or at a more general cultural level. In Bakhtin’s account, dialogism destabilizes the authoritative word. In this respect, Bakhtin and Voloshinov’s theory of conflictual dialogic speech is also a theory of hope, sustaining the possibility that meaning is dynamic, effectively without end, because there is no necessary end to discursive interactions. Dialogism is the life of language. Dialogism in the Pardoner’s Prologue and Tale (681 lines) gives special force and depth to the text’s signification, as marked by the text’s use of interjections. Allas functions grammatically, rhetorically, and pragmatically as part of a complex textual dialogism. The sermon-like text (prologue and tale) is slightly more than half as long as the Parson’s Tale (1005 lines), but the tales have nearly the same number of allas occurrences: eleven times in the Parson’s Tale, nine times in the Pardoner’s Prologue and Tale. Allas in the Pardoner’s Prologue and Tale is spoken four times by the Pardoner as narrator, three times by the Host in his exchange with the Pardoner, and twice by the Old Man who confronts the rioters in the exemplum. In effect, allas in the prologue and tale is clustered around particular speakers and scenes. Associating allas with different speaking positions foregrounds the narrativity of the Pardoner’s tale more than its genre as an exemplum within sermonic discourse. Unlike the Parson’s penitential text, the more sermonically rhetoricized Pardoner’s prologue, sermon, and exemplum combine allas with many other kinds of interjections, expletives, oaths, curses, and light insults, all of which have what the grammarian Magister Johannes considered to be interjective force. These different kinds of expressions have different pragmatic implicatures because they are used by characters in different textual and subject positions.

126 

The Medieval Life of L anguage

In fact, with respect to interjections and discourse markers, exciting or vocative O (see 6.512, 895, 896, 897, etc.) is more prominent in the Pardoner’s discourse than allas. In the Pardoner’s discourse, the textual contexts reveal O’s function as a marker of topic shift (e.g. “O glotonye, on thee wel oghte us pleyne!”; 6.512; cf. Jucker and Taavitsainen 2013: 64-67) and also as a descriptive highlighter (“O wombe! O bely! O stynkyng cod …”; CT, 6.534). In this sense, the Pardoner’s O is more interpersonal than expressive. Continuing the Host’s interjective and cursing lead-in to the Pardoner’s Prologue (“‘harrow!’ quod he, ‘by nayles and by blood’”; 6.288), the Pardoner’s exemplum depicts rioters who constantly swear in their tavern speech with taboo interjections, curses, and blasphemous oaths (“By the blood of Crist,” “by Goddes armes,” “I make avow to Goddes digne bones!,” etc.). But significantly, none of the rioters uses allas, perhaps another indication of allas’ special status in Chaucer’s poetic discourse. Allas belongs to a higher discourse register than blasphemous curses and tavern insults, however much those interjective expressions signify negative affect and evaluation. Linguistically and pragmatically, then, allas in the Pardoner’s Prologue and Tale is more contextually restricted than the interjection is in other parts of the Canterbury Tales. Allas is used two times in the exemplum, clustered in the Old Man’s speech with the emotion implicature regret (“Ne Deeth, allas, ne wol nat han my lyf. / Thus walke I, lyk a restelees kaityf, … / Allas, whan shul my bones been at reste”; 6.727, 733). Allas occurs three times in the Pardoner’s voice: once in his initial sermonizing on the sin of gluttony (6.517) and then two times clustered as discontinuous allas in the Pardoner’s preaching on sin and the need for penance (“Allas, mankynde, how may it bitide / That to thy creator, which that the wroghte / And with his precious herte-blood thee boghte, / Thou art so fals and so unkynde, allas?”; 6.900-903; cf. Taavitsainen 1995: 196). The Old Man’s use of allas (6.727) is a rare exception to the English interjection’s disjunctive clause-initial or clause-final syntactic constraint. The Pardoner’s three uses in his own voice are positioned at the opening and the conclusion of his sermon and exemplum. In those strategic positions, allas frames the speaker’s official condemnation of sin and call to repentance. As a discourse marker, allas also marks a new topic and introduces the opening passage on gluttony (“Allas, the shorte throte, the tender mouth, / Maketh that est and west and north and south, / In erthe, in eir, in water, men to swynke / To gete a glotoun deyntee mete and drynke!”; 6.517-520), but it’s not clear to what the disposition is being directed. Nonetheless, allas in line 517 functions as a discourse marker with negative affect, perhaps simply to elicit the audience’s

ALL AS CONTEX T

127

mournful response to the fact that the human body enables ruinous excess. In the second passage (6.900-903), the discontinuous allas identifies and frames the speech with negative evaluation, following the Pardoner’s more frequent exciting O (cf. 6.895, 896, 897, etc.) (So-called exciting O was used more frequently after 1450 in spoken and written English as a topic shift marker.). Discontinuous framing allas … allas in 6.900-903 adds emotional intensity to shading and negative evaluation, suggesting the text’s more rhetorically involved style. Allas as interjection and discourse marker is part of the Pardoner’s preaching discourse on sin and penance but not part of his metadiscourse on his performance nor part of the rioters’ speech. That is, the use of allas in the Pardoner’s Tale is restricted to the Pardoner’s role as the Church’s voice of confession, repentance, and pardon. But as we read the text and observe the Pardoner’s performance unfold, we realize how much his performance undermines his presupposed institutional and moral authority even as his rhetorical skills make him attractive, or at least worth paying attention to. Allas functions pragmatically as part of the Pardoner’s “preaching” voice, but not in the Pardoner’s personal revelations nor as part of his creation of a realistic setting for the exemplum. In the larger context of the Pardoner’s performance, the judgmental voice indexed by allas and the narrative position such a voice implicates are undermined by the Pardoner’s bald on-record metacommentary, his public admission revealing his material motives and cynically calculated audience manipulation. The Pardoner’s performance implicates that all discourse, including judgmental and affective allas, is audience manipulation. As for pragmatic stance, his openness about his motives ironically puts sincerity in question. The Pardoner’s truth-telling diegetically destabilizes both his ability to manipulate the pilgrims who are his audience and textually his speaking position as a voice of the Church and moral integrity. Just as allas shades and frames individual utterances, so the Pardoner’s metacommentary shades and recontextualizes his performance. Contextually, the narrating Pardoner presents the rioters’ oaths, curses, and interjective language in quoted (direct) or reported speech, thus distancing his voice from taboo speech even as he articulates that speech as part of his narration. Narration as metalinguistic mentioning rather than using. The text’s double-voiced narrative opens the question of the Pardoner’s discursive stance, his hypocritical and ironic discourse (cf. Bakhtin 1991: 103-111). In the Pardoner’s Tale, as I noted, swearing, crude phrases, and interjections function primarily to highlight the rioters’ moral condition, but they are presented in the Pardoner’s mimicing voice. The Pardoner is constructing

128 

The Medieval Life of L anguage

both a moral and a realistic context. The taboo words the blasphemers speak do both, and while the Pardoner entones the religious discourse of sin and penance he also inadvertently or recklessly frames or reweaves them into his self-revelation of his greedy intentions. The rioters’ expressions betray their hearts. But when the Pardoner repeats their speech to characterize them as part of his moral discourse, he undermines himself. When the rioters linguistically tear Christ’s body, they reenact the crucifixion, just as they tear apart the trust that keeps society from falling over, interaction by interaction. Likewise, the Pardoner tears apart the community he aims to manipulate with his speech. The Pardoner’s narrative impersonations are performatively and pragmatically unmoored, simultaneously his speech and not his speech. As a storyteller, the Pardoner tears apart the veil of fiction and reveals the teller. In what might appear to be a self-destructive move, the Pardoner repeatedly inserts into his interpersonal discourse a self-reflexive, metanarrative exposé of his preaching motives and strategies: “For myn entente is nat but for to wynne, / And nothing for correccioun of synne … / Thus kan I preche agayn that same vyce / Which that I use, and that is avarice” (CT, 6.403-404, 427-428). The oaths, the reckless and deceiving rioters, the grandiloquent passages proclaiming the Church’s condemnation of sin, the metanarrative self-revelation, all are affective utterances, governed as medieval grammarians might say more by modus affectus than by modus conceptus. Pragmatically, they signal the Pardoner’s separateness, his pride, and his exaggerated yet engaging linguistic behavior. The way the Pardoner’s speech is contextualized in the performance narrative or the exemplum pressures and makes ambiguous, destabilizes, the pragmatic criterion of speaker sincerity as part of either medieval vox articulata or Gricean communication. The Pardoner’s Prologue and Tale show that the trope of disavowal is disseminated throughout the narrative performances in the text of the Canterbury Tales, but the Pardoner’s bald, on-record statements about his hypocritical stance constitute perhaps the most pragmatically ironized of all the narrating pilgrims’ performances (cf. Howard 1976: 339-357). As such, the Pardoner’s performance might be reimagined as agit-prop theater, the sort of action which calls other, seemingly assured behavior to account. As Donald Howard said, “There is something awesome and perversely heroic about him. We have said he is an actor: he acts the role given to him, becomes an embodiment, a charade of evil. And in doing this he reflects back to us the evil we dread in ourselves” (1976: 372). The pragmatic irony is that when he drops the ideological veil of moral authority, the Pardoner speaks sincerely to his audience.

ALL AS CONTEX T

129

Chaucer’s use of allas in fabliaux, fable, and folk tales varies quite a lot with respect to frequency and discursive contexts. The distribution and frequency of allas in his fabliaux do not show a clear correlation between the form and the genre. Allas counts are: Miller’s Tale (745 lines) fourteen times, Reeve’s Tale (403 lines) two times, Shipman’s Tale (434 lines) two times, Summoner’s Tale (585 lines) not at all, Merchant’s Tale (1205 lines) twelve times. In the Nun’s Priest’s Tale (679 lines), a fable, allas occurs twelve times. In the Friar’s Tale (399 lines), based on a folk tale, once. Rather than reflecting a specific genre context, the high number of allas uses in the Miller’s Tale, Merchant’s Tale, and Nun’s Priest’s Tale perhaps has more to do with the fact that all three tales share a thematics of mercantile (bourgeois, artisan) discourse and parodies of courtly and religious discourse. The Miller’s Tale, weaving together several discourses and two fabliaux, has a complex relation with medieval pragmatic theory and metapragmatic awareness. In Chapter 4 I discuss the tale’s narrative pragmatics more generally. Here, I focus on the tale’s textual pragmatics of allas and other interjective speech. All the tale’s uses of allas occur in the direct or reported speech of individual characters and not once in the voice of the narrating Miller. That is, allas functions as part of character development and interpersonal address more than as part of a narrational standpoint. Readers have often noted the Miller’s Tale’s complex parodic relation with the Knight’s Tale, and that connection is borne out at the micro level of allas. The lovelorn men use allas the most (Arcite and Palamon: 20 out of 23; John and Absolon: 11 out of 13), whereas Emily and Alisoun are allotted one and two allas utterances, respectively. Emily expresses allas to intensify her lament at being the marriage prize. Alisoun uses allas twice in complex discursive situations. First, as she pushes Nicholas away when he comes on to her, she tells him what she might say to raise the alarm if he persists: “‘Why, lat be!’ quod she. ‘Lat be, Nicholas, / Or I wol crie, ‘out,’ ‘harrow’ and ‘allas’!” (CT, 1.3285-3286). Her clustering of interjections gives her implied threat an urgency and idiomatic quality. In fact, she is reciting the conventional formula for raising the public alarm in an emergency. Alisoun’s speech at 1.3285-3286 is textually ironic because of the comic timing and rapid narrative juxtaposition of two situational contexts. Initially, Alisoun refuses Nicholas’ advances with negative emotional intensity and a face-threatening speech act, but five lines later she agrees to a sexual assignation with the hende Oxford clerk. Allas shades her first utterance of refusal, but her consent undercuts her initial threat. Alisoun’s later use of allas is equally insincere (1.3607).

130 

The Medieval Life of L anguage

Table 3.2  Allas in Chaucer’s Miller’s Tale Total tokens Alison Absolon John Nicholas

14 2 6 5 1

The male characters use allas in a variety of ways. Nicholas’ shading allas comes when he pretends to be overcome by his prognostication of the future flood: “… Nicholas / Gan for to sik soore, and seyde, ‘Alas! / Shal al the world be lost eftsoones now?’” (1. 3487-3489. Riverside Chaucer editors’ punctuation.). Like her first allas, Alisoun’s second, insincere allas occurs when she continues the trick to get John out of the way so she and Nicholas can “pley” (1.3607). For Nicholas and Alisoun, allas is part of an insincere, affected mode of speech. John and Absolon, on the other hand, use allas to frame or intensify their sorrow, dread, or chagrin at what they think will happen (the coming flood) or has happened (the misplaced kiss). John’s uses of allas all cluster around his fear of losing his possessions, including his wife. Absolon’s clustered allas (four times in fourteen lines) highlights his deeply negative emotional and physical reaction to his misplaced kiss and his equally deep desire for revenge (1.3739-3753). In particular, Absolson’s reduplicated allas in line 3753 intensifies his expression of regret: “‘Allas,’ quod he, ‘allas, I ne hadde ybleynt!’.” These examples of how allas is used, and many of the others discussed in this chapter, suggest how the word, in addition to being an interjection and a pragmatic marker, also functions as a pragmatic intensifier cohering not with a verb but with a speaker’s feeling. The Wife of Bath’s Prologue and Tale (1264 lines) presents a different kind of textual pragmatics involving allas and other interjections. The Prologue’s affinity with popular sermons is complicated by the polyphony and dialogism of the textual mixture of discursive modes, including “an act of deliberate self-fashioning” bound up in a contradictory rhetoric of commodification and lay literate subversion (Patterson 1991: 288-289; cf. Carruthers 1979, Amsler 1987). The Wife’s dissenting discourse – “that am nat I” (CT, 3.112) – is filled with colloquialisms and affective responses and evaluations. Most often, the Wife uses homophonic interjections, light curses (“A ha!,” “pardee!,” and the stronger “Fy!”; CT, 3.586, 712, 735), oaths (“by my fey,” “Benedicite!,” “by Goddes sweete pyne!”; 3.215, 241, 385), and other conventionally interjective speech. The Wife’s interjections and pragmatic markers constitute a conventionally masculine but less aggressive version

ALL AS CONTEX T

131

of the rioters’ speech in the Pardoner’s Tale – colloquial, affective, with interjective force – but without the blasphemy or violence. The violence in the Wife’s Prologue and Tale occurs in other ways than interjections. The Wife’s use of allas is somewhat restricted (six times) compared with her colloquial, idiomatic speech in general. She uses the interjection herself three times in the Prologue (CT, 3.474, 614) and three times in the Tale in the voice of the knight (3.1058, 1068, 1098). The Wife uses solo allas when she laments how time has taken its toll on her body (“But age, allas, that al wole envenyme, / Hath me biraft my beautee and my pith.”; CT, 3.474-475). She uses the emphatic reduplicating allas to express and intensify her regret that her personal desires and inclinations, a combination of Mars and Venus, have been at odds with the Church’s doctrine on sexuality (“Allas! Allas! That evere love was synne!”; CT, 3.614. Riverside Chaucer editors’ punctuation.). In both passages, allas shades the following clause with the speaker’s regret and negative evaluation. In the Wife’s tale, all three uses of allas are spoken by the rapist Arthurian knight. The knight uses allas in an emotional dialogue with the lady to express his deep regret that he must now fulfill his promise to marry and live with a woman he considers “foul, and oold, and poore” and “loothly” (CT, 3.1063, 1100). The knight’s repeated use of allas (3.1058, 1068, 1098) in a clause initial, shading position is combined once with the formulaic “Allas and weylawey!” when he acknowledges his promise to the lady (CT, 3.1058). Formulaic allas implies he believes his individual agency or freedom has vanished: “I woot right wel that swich was my biheste” (CT, 3.1059). As soon as he is released by the queen from his charge at Arthur’s court, the knight’s speech is filled with other nominal and adjectival descriptors which signify his sorrow and regret for having promised to marry the old lady who saved his life. The uses of strategic allas in the knight’s speech mark or punctuate the intense sorrow he feels in his situation and his emotional anguish. Allas marks his struggle against the pragmatic implicature of his speech act promise. The utterances foreground his less than chivalric attitude and perhaps his possible bad faith. Whereas in her Prologue the Wife used allas more as a commentary on her social position and self-image as a middle aged, sexually active widow, the knight’s allas signals his sorrow at what he thinks will be a humiliating match. Both the Wife and her fictional Arthurian knight believe they are being kept from what they deserve, but the Wife’s regret implicates her humanity and love for life. The knight’s speech at the end of the tale, on the other hand, is ironic in that he has fulfilled Guinevere’s challenge with the help of the lady, but he doesn’t perceive his situation to have improved. He believes that the privilege his social rank

132 

The Medieval Life of L anguage

affords is not being respected. Unlike the Wife, the knight’s allas reflects his arrogance. At this point, we can ask: When speakers use allas, are they expressing the same regret and despair in the same way? Does allas signify a universal feeling whoever is speaking? I don’t think so. Allas as a material vocalization does not ‘mean’ or indirectly signify I feel regret or I am in despair or I feel sorrow. To use the medieval grammarians’ analytic vocabulary, allas is meaningful in context as actus exercitus, not as actus significatus. The conventional Middle English word is associated with a network of descriptors, nouns, adjectives, and verbs. Recall the stanza in Troilus and Criseyde: Ful pale ywoxen was hire brighte face, Hire lymes lene, as she that al the day Stood, whan she dorste, and loked on the place Ther she was born, and ther she dwelt hadde ay; And al the nyght wepynge, allas, she lay. And thus despeired, out of alle cure, She ladde hire lif, this woful creature. (TC, 5.708-714)

In the text Criseyde’s sorrow and sense of loss is not stated directly, at least not at first, nor is it simply transmitted by allas. The force of the interjection emerges as Criseyde’s feelings emerge when the reader proceeds through the stanza and connects up the adjectives, verbals, and her represented actions. The collocation of words associated with allas in the stanza constitute a micro semantic network (context) representing Criseyde’s feeling in concrete terms, complete with category verbal and related attributes (see Aitchison 1987: 72-85). The descriptors are capped by Criseyde’s allas in free indirect discourse. Then in line 713 the text shifts from actus exercitus to actus significatus. Criseyde “despeired, out of alle cure …” The stanza concludes with the adjective “woful,” which shifts Criseyde’s allas to a more distanced representation of her state of mind. In the Wife of Bath’s Prologue and Tale, the two speakers and different narrative situations contextualize allas as pragmatic marker, intensifier, and self-revelation. Allas signifies heightened negative feeling, but just what sort of negative feeling the expression indexes in a given instance is inferred from the immediate textual context in terms of how readers interpret the speaker. The feeling or disposition allas connotes will vary. Disgust is not regret is not sorrow. Allas may conventionally be associated with negative affect, but speaking contexts and reader inferences establish the sort of ‘negativity’ the interjection indexes at a particular moment. In both the Wife and the rapist

ALL AS CONTEX T

133

knight’s speech, allas functions as a discourse marker, connoting either regret (Wife) or despair (knight). In itself the expression allas has no explicit referent or external signification. Rather, the interjection signals to audiences the speaker’s intense negative emotion (despair, regret, sorrow) and/or their negative evaluation of a situation or condition they are implicating, whether their own or others’. Interjective utterances are gestural and interpersonal. As discourse markers and features of an involved style, interjections function by shading an utterance to which they are attached and indexing the speaker’s emotional attitude to a particular referent or situation in a particular context. Allas opens an affective textual space for the speaker’s performance of sorrow or regret at what has occurred or happened to him or her, but not all allas expressions from speakers elicit empathy or pity from the audience. Textual context can ironize the utterance. No utterance occurs in isolation. To cite one example from Chaucer’s fabliaux: In the Reeve’s Tale, the clerk John complains that their horse has run off: “Allas, our wardeyn has his palfrey lorn” (CT, 1.4075). The miller’s wife, seeing what’s happened, reciprocates with the same expression, but she sarcastically accuses him of being at fault: “… Allas! youre hors goth to the fen / With wilde mares, as faste as he may go, / Unthank come on his hand that boond hym so, / And he that bettre shold han knyt the ryne!”; CT, 1.4080-4083). Diegetically, there is no need for the wife’s reply. Instead, the wife’s reply serves as a metacomment on the clerk’s attitude and lack of care for his warden’s property. In this dialogue, the clerk and the miller’s wife both use initial allas to shade the following utterance with negative feeling, but the wife’s answering allas does not signal a sympathetic bond with the clerk. She expresses some sorrow that he has lost his horse, perhaps out of conventional politeness, but then she turns her utterance in a more accusatory direction. It’s John’s own fault he lost his horse. No sympathy from her. Allas can open an affective space in a discourse and create sympathy between a character and the audience or between characters, but not automatically. It depends on who is speaking. Chaucer’s use of allas in the Merchant’s Tale, a fabliau wrapped in a problematic domestic drama, and the Nun’s Priest’s Tale, a beast fable, suggest even more complex ways interjections are pragmatically contextualized as emotional discourse and how allas functions as a pragmatic marker. Again, the text’s genre features in themselves do not adequately account for the ways allas might be used in different contexts in the text. The Merchant’s Tale suggests how speakers’ uses of interjections and pragmatic markers in micro contexts have larger narrative significance, not only genre. Allas occurs eleven times in the tale: five times in January’s speech, two times in May’s, three in the narrator’s, and once in Pluto’s. January’s

134 

The Medieval Life of L anguage

uses of allas cluster in two episodes: when he imagines sex with May and when he discovers May and Damien having sex in the pear tree. The verbal clusterings suggest some authorial intention as well as how the text affiliates allas with particular themes or activities, in this case, sex. Why allas should be associated with sex becomes one of the Merchant’s Tale’s thematic complexities. A close reading of the text reveals how particular narrative situations and speaking positions manipulate and multiply conventional significations of allas, such that the interjection no longer can be said to have a univocal pragmatic meaning. January’s first two uses of allas occur as he anticipates sex with May on their wedding night. The first time he thinks to himself: And thoughte, “Allas! O tendre creature, Now wolde God ye myghte wel endure Al my corage, it is so sharp and keene! I am agast ye shul it nat susteene …” (CT, 4.1757-1760)

He uses allas a second time when he lies in bed with May on their wedding night: And seyde thus, “Allas! I moot trespace To yow, my spouse, and yow greetly offende Er tyme come that I wil doun descende.” (CT, 4.1828-1830)

We might call January’s speech in these scenes creepy allas. In the wedding night episode, January’s interjection shades with menace his arrogant imaginings and masculine narcissism about his sexual prowess, accentuated by his fantasy of May’s innocence and vulnerability. (Later, her interior, private evaluation of January’s sexual prowess after their wedding night suggests she’s not so innocent after all.) January isn’t expressing sincere despair or lament for something lost. Whether she is or isn’t sexually naïve is not the point. Allas is connected with January’s imaginary experience and his disposition. He’s almost gloating as he apologizes for being too rough or sexually strong with May. In this context, any conventional association of allas with the speaker’s despair or lament for loss goes out the window. Rather, January’s wedding night allas pragmatically frames his sexual fantasies and his lack of empathy with creepy irony. The affect is negative, but the interjection adds a menacing threat to January’s discourse. His speech may evoke disgust or fear from the audience, but it certainly doesn’t elicit sympathy or empathy.

ALL AS CONTEX T

135

January’s second cluster of allas occurs late in the tale when he and May are at the pear tree. In this episode, January, now blind, uses the interjection more conventionally and sincerely to index his lament. First, he despairs in somewhat overwrought tones that he has no one to help get his wife some fruit from the pear tree: “Allas … that I ne had heer a knave / That koude clymbe! Allas, allas …” (4.2338-2339). Then when Pluto restores his sight and he sees May and Damien in the tree, he raises the alarm with the conventional formula: “‘Out! Help! Allas! Harrow!’ he gan to crye” (4.2366. Riverside Chaucer editors’ punctuation.). In this second episode January’s clustered, reduplicated, and formulaic allas function as urgent calls for help. Syntactically, the expressions intensify the situational affect. Thematically, they amplify January’s immediate despair and vulnerability. Contextually, because the reader is privy to the fact that May has already deceitfully arranged to meet Damien for a tree tryst, January’s affective speech around the pear tree mitigates what might otherwise be interpreted as natural justice, payback for his arrogant possessiveness and sexual presumption. Correlatively, Pluto’s allas on the sidelines also solicits sympathy on behalf of all wronged men everywhere when he calls Proserpina’s attention to the deception Damien and May are planning: “Ne se ye nat this honurable knyght, By cause, allas, that he is blynd and old, His owene man shal make hym cokewold.” (CT, 4.2254-2256)

In this second episode, January uses allas directly, intentionally, and urgently to call for help, while the text indirectly foregrounds his vulnerability. Additionally, Pluto’s allas reflects empathy with January’s state. The context for January’s second allas cluster creates a very different signification from the context of his first cluster. May’s two uses of allas, also clustered at the end of the narrative, are instances of insincere allas. In the micro textual contexts, she says it, but in the second instance she doesn’t mean it, while in the first instance it’s not clear whether “syde” refers to her feeling the baby kick in her womb or whether she’s hungry or both, or whether she’s just making up an excuse to get near the pear tree: This fresshe May, that is so bright and sheene Gan for to syke and seyde, “Allas, my syde! I moste han of the peres that I see, Or I moot dye, so soore longeth me To eten of the smal peres grene.” (CT, 4.2328-2333)

136 

The Medieval Life of L anguage

“Ye maze, maze, goode sire,” quod she; “This thank have I for I have maad yow see. Allas,” quod she, “that evere I was so kynde!” (CT, 4.2387-2389)

In the second instance and perhaps in the first, May’s use of allas is conventional but insincere as the narrative context sets her motivations against the word’s customary semantic and pragmatic meaning of sorrow or despair and the speaker’s presumed sincerity. January and May use various kinds of allas, sincere, insincere, and creepy, in the tale. The narrator of the Merchant’s Tale uses allas three times as an affective interjection or as a discourse marker. As we have seen, interjections can function as pragmatic markers to introduce a topic or topic shift in a discourse. When January suddenly goes blind, the narrator, like the Monk, uses allas to shade the event with negative affect and also to mark the turn in the narrative situation: “Allas, this noble Januarie free, / Amydde his lust and his prosperitee, / Is woxen blynd, and that al sodeynly” (4.2069-2071). The narrator’s use of allas to mark the change in January’s fortunes contrasts with January’s own earlier, creepy uses on his wedding night and foreshadows January’s more vulnerable allas later in the tale. The narrator’s metacomment on January’s change of fortune is presumably sincere, but his expression at 4.2069 conveys conventional negative affect – things just go badly sometimes – more than empathy with or despair at January’s loss of sight. The Merchant’s Tale manifests a wide range of conventional and not so conventional uses of allas: sincere, insincere, creepy, empathetic, topic shift marker. The Nun’s Priest’s Prologue and Tale is a complex, polyphonic narrative, but it manifests a more restricted use of allas. The interjection occurs twelve times: once in the tale’s linking Prologue when the Knight responds to the Monk’s Tale, three times in the narrator’s speech, two times in Pertelote’s, three times in Chauntecler’s, and three times in the fox’s. About half of the instances conform to the conventional affective performance of sorrow or despair, whether in a character’s speech (e.g., Pertelote egging Chauntecler on [7.2909]) or in reported speech when Chauntecler recounts to Pertelote stories of prophetic dreams (7.3004, 3045). The fox of course uses insincere allas when he tries to coax Chauntecler into his mouth (7.3284, 3419). The narrator’s uses of allas are primarily as discourse markers and evaluators: introducing his digression addressed to “ye lords” that they beware of flatterers at court (7.3325-3330) and then to signal narrative turns (7.3339-3340), much as the narrator in the Merchant’s Tale does. When the fox is discovered in the chicken coop, the widow uses the conventional formula to raise the alarm but without allas: “And cryden, ‘Out! Harrow and weylaway!’” (7.3380).

ALL AS CONTEX T

137

Taavitsainen (1995) is right when she argues that lexical clustering in fabliau and beast fable, both folktale-based genres, creates emotional intensity and represents frantic behavior in the narrative. January’s second cluster of allas exemplifies that narrative technique. But the lexical clustering of interjections also marks emotional intensity in Chaucer’s other poems too – romances, mock sermon, moral and sermonic texts. And not all uses of allas in Chaucer’s fabliaux function primarily to mark emotional intensity. Which emotions in which contexts for which characters or speakers are the pragmatic and interpretive questions. I’ve only discussed some tales here, but very few Canterbury tales have no occurrences of allas. That distribution implies that for Chaucer allas as an interjection and pragmatic marker was an integral, conventional part of conversational, narrative, and sermonic discourse and at the same time was an expression to be used deliberately. When we read closely the narrative micro contexts for allas in Troilus and Criseyde and several Canterbury Tales, we find that uses of allas range from flavoring a speech or consignifying a character’s intense despair, to shading a specif ic utterance with negative affect or negatively evaluating an utterance, feeling, or situation, to marking a topic shift or narrative turn. These pragmatic functions highlight how we can usually infer from the narrative situation whether speakers, characters, and narrators are using allas sincerely or insincerely. Lexical frequencies and textual distributions of particular expressions are helpful for recognizing and analyzing a discursive field, but we understand pragmatic significations and textual functions in specific contexts. Genre is just one of the contextual features of pragmatic signif ication and discourse marking. Functionally, the meaning of allas, or weylaway for that matter, changes depending on the point of view (focalization) and attitude (disposition) of the speaker, whether the speaker is a narrator or a character within a narrative or both, whether the speaker is sincerely or insincerely expressing sorrow or regret or proposing a negative judgment about a situation or outcome. Most often in Chaucer’s poetry, individual uses of allas acquire meaning because of the web of semantically and thematically related words around them. Middle English allas functions discursively and textually with other interjections, words with interjective force, and descriptors which frame or foreground mostly negative emotional intensity and affect. Allas and weylaway might be construed as equivalent in some instances, but they do not fully index the same meaning or consignify the same social context for the speaker or the topic. Allas, whether it functions as an interjection or a discourse marker or both, loads emotional intensity or negative affect

138 

The Medieval Life of L anguage

on to particular utterances or narrative situations. Weylaway combines sometimes with allas and adds a more colloquial tone. The relative frequency of Chaucer’s use of allas in his poetry and prose suggests that use of the interjection was a distinct pragmatic and discursive choice on his part and designed to achieve particular effects. In Chaucer’s writing allas functions as a discourse marker to negatively shade or frame utterances or elicit audience sympathy. In certain narrative contexts allas is used insincerely or in a creepy, menacing manner. The expression also functions to highlight the speaker’s moral negative evaluation of particular actions or situations. Recognizing that a text belongs to a text type (romance, fabliau, personal letter) only partly accounts for the pragmatic strategy writers adopt and only partly the pragmatic inferencing readers do or need to do to construct meaning from utterances. Interjections such as allas can function in local textual contexts as metatextual markers for affective response, discursive frames, or igniters of emotional appeals. Stylistically, allas is a conventionally elegant, even polite kind of negative interjection compared with other more colloquial and homophonic interjections such as insults, curses, and oaths. Middle English allas can be used in colloquial formulas, but more often it reflects a higher register and greater emotional loading. Chaucer uses allas more than almost any other Middle writer except Malory. By contrast, allas occurs very infrequently in the prose Paston letters. It seems for Chaucer allas was not only conversational but also a highly functional performative discourse marker in poetic writing. Reading closely the uses of interjections in Chaucer’s poetry gives greater depth and detail to thirteenth-century grammarians’ pragmatic theory and their analysis of interjections and other speech which signify per modum affectus.

4

Alisoun’s Giggle, or the Miller Does Pragmatics Abstract This chapter discusses Chaucer’s Miller’s Prologue and Tale as an encounter with grammatical theory and pragmatics more generally. Analyzing the tale’s use of interjections, polysemy, and equivocation, we discover how Chaucer deconstructs the notion of stable, authorial, intentional meaning and explores narrative dialogism and the pragmatics of identity and affective power for comic and satiric effect. Key words: interjection, Miller’s Tale, equivocation, polysemy, recontextualization

Is a giggle a kind of speech? Is a literary giggle? In a nutshell that’s the question I want to pursue in this chapter. In the Miller’s Tale Alisoun the carpenter’s wife plays a joke on the unsuspecting but hopeful Absolon. Then she giggles: “‘Tehee!’ quod she, and clapte the wyndow to …” (1.3738). The line is a double articulation. In semiotic terms the line is part of a poetic ‘utterance.’ In pragmatic terms Alisoun’s vocalization is a speech ‘utterance.’ What are the implications of the utterance for the tale as a narrative ‘utterance’? In his semiotics of narrative and dialogic speech Bakhtin always uses the term utterance (Rus. slovo, word or speech in use, not language in the abstract) in the sense of language in action, whether oral or written. Depending on context, an utterance can be a single word, a line of poetry, a chapter of a narrative, or the story itself. A word or sentence “belongs to nobody, and only by functioning as a whole utterance does it become the expression of the position of someone speaking individually in a concrete situation of speech communication” (Bakhtin 1994: 84). Context and dialogic interaction are everything with respect to pragmatics and communication. For Bakhtin, “The boundaries of each concrete utterance as a unit of speech communication are determined by a change of speaking

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch04

140 

The Medieval Life of L anguage

subjects, that is, a change of speakers” (Bakhtin 1994: 81; cf. 1994: 62-72, 81-87). An utterance presupposes and entails both a prior utterance and an answering utterance, an addressee and a situation: “Utterances are not indifferent to one another, and are not self-sufficient” (1994: 85). As part of conversational ‘work,’ speakers need to discover what the other means to say and respond in a way the speaker thinks is needed or appropriate, which does not necessarily mean in the way the other expects or wants or wants to say. Bakhtin’s theory of the utterance and dialogism has important implications for pragmatics, both medieval and modern. Speakers and listeners construct meaning by relying on context, inference, and implicature to augment or complete their interpretations of linguistic form and communicative action. As we have seen, both medieval grammar and contemporary pragmatics expand, in different ways, the horizon of grammar and meaning to include not only propositions and formally explicit utterances but also incomplete utterances, inferences and implicatures, nonlinguistic vocalizations (moans, grunts, laughs), gestures, and pragmatic and metapragmatic acts. Word and word-like utterances such as Ugh, OMG, papae, and giggles become meaningful and communicative within particular speech contexts, whether or not they are uttered with deliberate intention. Conventional and natural interjections, laughing, and other vocalizations manifest a speaker’s attitudes, feelings, thoughts, and point of view on topics or information as much as words can or do. And such vocalizations solicit or trigger responses and replies from listeners. Bacon and other medieval grammarians did not always regard cries of pain, interjections, and other vocalizations as words per se (dictio), but they did understand them as meaningful and affective utterances in particular contexts. Moreover, Magister Johannes and others argued that much elliptical speech is perfectio and therefore meaningful with respect to the listener’s understanding and the speaker’s intention (intellectus), even if such speech violates grammatical form (sensus). I think the Miller’s Prologue and Tale has a lot to tell us about pragmatics and dialogism. The tale is regarded as one of Chaucer’s best comic narratives from a structural and stylistic point of view and one of the densest narratives of any period (e.g., Donaldson 1970: 13-29; Patterson 1991: 244-279; Bryan 2016). The tale’s verbal texture is complex and resonant. Readers have identified several discursive threads embedded and parodied in the tale: courtly, exegetical, theatrical, fabliau parody. Donaldson (1970), Neuss (1974), and others have drawn attention to the way the text repeats words or phrases in

Alisoun’s Giggle, or the Miller Does Pr agmatics

141

different contexts to evoke puns or suggest multiple meanings, subversions, and misunderstood intentions. The intertextuality of the Miller’s Tale unfolds a complex exploration of pragmatics, and that’s where Alisoun’s giggle comes in. Thematically, the tale constitutes a metapragmatic rumination on utterance, meaning, and context. The narrative also enacts a face-threatening act by deploying folkloristic tale-telling as a form of indirect insult (Lindahl 1989: 129-134). Structurally complex, the tale’s interlocking verbal constellations derive from metapragmatic awareness and a sophisticated critique of intentionalist pragmatics and monologic linguistic order. The text achieves this by juxtaposing everyday, learned, and elite discourses in a dialogic narrative action. The tale explores the logical and pragmatic aspects of equivocatio, not as a fallacy but as a polylogue of communications and miscommunications, with troubling but comic effect. As we know from the Hous of Fame and the Nun’s Priest’s Tale, Chaucer was conversant with fairly sophisticated ideas in medieval grammar and rhetoric (cf. Irvine 1985). I would add the Miller’s Prologue and Tale to that shortlist. First, I’ll discuss the prologue and tale’s metapragmatic orientation toward utterances, especially interjections and gestures, and the way the text activates equivocal or polylogic language and (im)polite and (non) cooperative speech. Then I’ll focus on how the text’s relentless narrative re-contextualization of words and speech acts questions textual stability and univocality in terms of pragmatic relevance and performativity. The narrative recontextualizations produce multiple foci of situated knowing, not all of them legitimated or authorized in the text. The tale’s critique of intention, agency, sincerity, and event structure intersects with medieval and contemporary theorists’ interest in the formal and pragmatic grounds for meaningful as opposed to grammatical speech and for narrative construction around subject position (deixis) and point of view, what Genette calls “focalization” (Genette 1980: 189-194). Unlike ‘point of view,’ focalization doesn’t presume a visual context. Rather, the critical concept foregrounds linguistic action and dialogism. Focalization also helps us grasp what counts as a narrative unit, what determines “the difference in function and, particularly, the difference in information” from one narrative sentence or part of a sentence to the next (Genette 1980: 194; cf. Bakhtin 1994: 82). Focalization supplements Bakhtin’s theory of dialogism and reaccentuation. Pragmatic utterances and speech acts are embedded within culturally constructed speech genres. Bakhtin argues that without a “repertoire of oral (and written) speech genres,” “speech communication would be

142 

The Medieval Life of L anguage

almost impossible …” (1994: 82, 84). Bakhtin usually formulates his theory of dialogism in terms of individual speakers: for example, “The speaker’s speech will [something like ‘speech plan’] is manifested primarily in the choice of a particular speech genre” (1994: 83). But speech genres and speaking positions are not entirely individual or independent of social formations. They derive from the speech community’s shared intentions: “… we cast our speech in definite generic forms … We are given these speech genres in almost the same way that we are given our native language …” (1994: 83). Institutional and social conventions for utterance establish conditions for subjectivity and the field as defined by hegemony and antagonism, conflict, disruption, and possibilities. The conventions and the expectations those conventions give rise to intersect with a speaker’s goals, intent, and (re) voicing or reaccenting of the other’s speech (pereaktsentuatsiya; Bakhtin 1986: 87, 91). Reaccenting is a pragmatic strategy for intervening in speech and asserting speaker agency. Bakhtin’s theory of heteroglossia and dialogism connect with medieval pragmatic theories of equivocatio as both problem and potential. Bakhtin emphasizes how individual utterances belong to no one until they are contextualized, and then every context inflects the meaning and relevance of an utterance in one or more ways. Like Bacon’s re-impositio, reaccenting is potentially never ending: “‘What joy!’ in the context of a particular utterance can assume an ironic or bitterly sarcastic tone …” (1994: 85). In a speech community, dialogism’s structure is the dynamic between centripetal and centrifugal forces within language as utterance (Bakhtin 1994: 74-75). Centripetal language forces emphasize utterances which participate in monologic discourse and seek to draw together textual elements, points of view, and meanings around a central, dominant ideomeme. Centripetal discourse establishes an authorized or “authorial” focalization or viewpoint and promotes or enforces subordinated sources and references, including different languages or dialects, in order to produce stable types of speech, meaning, and social order. Centrifugal language forces, on the other hand, produce polylogism and potentially equivocation. They push textual connections and meanings away from a single central ideomeme. Centrifugal discursive force, aligned with carnivalesque discourse, enables multiple meanings and standpoints to come into play in a given text or textual environment. Multiple speaking agents are part of centrifugal discourse. When different discourses or standpoints are juxtaposed, speech is pluralized and the assumed consensus, shared presuppositions, or community intentions (meanings) are disassembled and disseminated. For Bakhtin, dialogic context is unfinalizable.

Alisoun’s Giggle, or the Miller Does Pr agmatics

143

Does a Giggle Mean? Small words and paralinguistic vocalizations can make a big difference pragmatically, and that’s certainly the case with the Miller’s Prologue and Tale. What do the text’s interjections tell us about the tale’s pragmatic orientation? As we have seen, interjections and other vocalizations signal a speaker’s mental state (cognitive), feelings and attitudes (emotive), and/or stance (social) in a given situation. They also convey the speaker’s relation with and attitude toward the relevant situation the speaker is involved in at the moment of utterance. Focalization produces a restricted field. Most often, the vocalization or natural or conventional interjection is part of the emotional response, a languaged and embodied act, rather than a representation of the disposition (heu vs doleo). Readers often point to certain scenes or utterances in the Miller’s Tale as turning points, narrative plot twists or climaxes which initiate or resolve action. These turning points or climaxes are interpreted as signs of Chaucer’s literary art. One such turning point in the tale, but one not usually discussed in pragmatic terms, comes when the priest Absolon, thinking he has a chance with Alisoun, woos her one night beneath her “shot-wyndowe” (1.3358). Mimicing vocabulary from the Song of Songs, he kneels and pleads with Alisoun, weaving a love speech and craving a kiss: “Lemman, thy grace, and sweete bryd, thyn oore!” (1.3726). The MED glosses ore as mercy or any opening, not just the mouth. In the dark, Alisoun sticks her butt or genitals (it’s not clear which orifice is in play) out the privy window, and Absolon eagerly kisses what he thinks is Alisoun’s mouth. He is surprised: He felte a thyng al rough and long yherd And seyde, “Fy! allas! what have I do?” “Tehee!” quod she, and clapte the wyndow to, And Absolon gooth forth a sory pas. (1.3738-3741)

The scene is loaded with rude affect and comic surprise. In the space of two lines (1.3739-3740) we have two interjections, a giggle, a question, and a rude gesture breaking off the exchange. The scene is a pragmatic combination of speech and body moves. The unsuspecting Absolon verbally reacts to kissing Alisoun’s “queynte” with disgust, indignation, and regret, marked by the juxtaposition of the two interjections “Fy” and “allas.” Alisoun enjoys her prank and giggles, “Tehee!” before closing the shutter in his face. I suspect more readers over the centuries have been inclined to sympathize with Alisoun’s response than with Absolon’s, whether or not the text has been politely proscribed from student editions, as it was in my high school.

144 

The Medieval Life of L anguage

How is Absolon’s humiliation and disgust the source of comic pleasure for Alisoun and many readers? This climactic narrative moment – the tale has many such – encapsulates the tale’s pragmatic and even richer metapragmatic play. The scene works in three steps. First, Absolon’s indignation and regret are encoded in “Fy! allas! what have I do?” Absolon’s intentions are thwarted and radically recontextualized by Alisoun and her kiss trick. The MED glosses “fy” as an interjection (“An exclamation expressing contempt, disapproval, or indignation”) and cites English instances from the Wycliffite English Bible translations of Latin vah and Vulgate/Aramaic rac(h)a. We have encountered some of these interjections before in medieval grammatical discourse. Many medievals thought racha (Matthew 5:22) was Hebrew and so untranslatable, but apparently not the Wycliffite translators. Latin vah is ambiguous as it can signify ‘surprise,’ ‘anger,’ or as Isidore and grammarians report, ‘joy’ (Etymologiae, 1.14). Joy? Context matters. “Fy” is a word-like element, but it is not semantically referential. The expression is comprised of a fricative and a vowel and when articulated, performs a bodily act, namely, spitting or breathing out from the mouth and/ or nose something strongly not wanted, something undesirable. Pragmatic utterances expressing physical or symbolic disgust are found throughout Indo-European languages (e.g., Polish fe and tfe; Russian fu; German pfui; Yiddish feh; English phew, phooy, pooh). Most are monosyllabic and combine an initial bilabial or labiodental fricative ([p, f]) with a vowel and sometimes a final velar ([g, k]) (cf. Wierzbicka 2003: 302-317 for further examples). Earlier English forms of pragmatic utterances similar to fy stress the initial bilabial or labiodental consonant and are phonologically CV, whereas some more recent English versions delete the initial consonant, stress the vowel, and add a final consonant (VC), e.g., ugh, aargh. English yuck seems to be an exception to this phonological structure, combining an initial glide with a final velar. Absolon’s “fy” is intentional but not deliberate in medieval pragmatic terms. Recall that Bacon described an expression like “fy” as a natural interjection emanating from the rational soul but signifying naturally in modus affectus. Absolon’s sudden, explosive word-like utterance is performative. It simultaneously expresses his attitude (disgust and indignation) toward what he has just experienced and is part of his physical response (cf. Wierzbicka 2003: 285-292, 302-315). The expression fy performatively and symbolically enacts the bodily gesture of spitting something out, a general “repulsion olfactive” or even a “universal gesture of revulsion” (Karcevski 1969: 198; Haiman 1989: 157). By the late sixteenth century, “fy/fie” had lost much of its physical performative force and become a common, generalized expletive

Alisoun’s Giggle, or the Miller Does Pr agmatics

145

which speakers used when angrily rejecting something or someone (e.g., Shakespeare’s Troilus and Cressida [1601/1602], 4.5.66-69). In the Miller’s Tale, “fy” is fully part of an affective gesture. Absolon’s vocalization explicitly relates utterance and gesture, linking an explosive articulation (erupta) with a physical act of oral revulsion. The expression is fully somatic. The radically performative signification of “fy” is reenforced when, away from the window and seething with resentment, Absolon literally tries to get Alisoun out of his mouth: Who rubbeth now, who froteth now his lippes With dust, with sond, with straw, with clooth, with chippes. (1.3747-3748)

“Fy” is a strong verbal gesture, most likely to be pronounced with consonantal stress, but in step two of the scene the pragmatics of this brief utterance get more complicated when Absolon immediately follows his aggressive “Fy” with his regretful “Allas, what have I do?” (Recall the discussion of ‘flavor allas’ in Chapter 3.) Absolon’s second interjection introduces a separate complete clause, a WH information-seeking question which implicates his emerging new understanding of the situation and connotes his self-reflexive feeling. Textually, the juxtaposition of the two interjections, one gestural ( fy), the other a more conventional English expression (allas), signifies Absolon’s two almost simultaneous but different feelings, indignation at the act and the perpetrator, remorse for his own responsibility for getting himself into the situation in the first place and allowing himself to be duped. Then, in the third step in this exchange, Absolon’s surprise, disgust, and regret receive a counterpunch and new context when Alisoun giggles at his expense: “Tehee!” quod she, and clapte the wyndow to, And Absolon gooth forth a sory pas. (1.3738-3741)

Whereas allas is a conventional interjection and fy a somatic-verbal gesture, Alisoun’s capping “Tehee” is another natural interjective sign, punctuated by her rude gesture of closing the shutter in Absolon’s face. How do you write a giggle? Perhaps Chaucer’s inscription of Alisoun’s giggle follows a fourteenth-century convention for rendering so-called natural utterances in reported direct speech in a text, but the MED only cites Chaucer’s text. The OED cites two written English examples: an English proverb (c. 1300) and Chaucer’s Miller’s Tale. But I don’t think Chaucer invented the written English word to represent a giggle.

146 

The Medieval Life of L anguage

In terms of focalization, “Tehee” solicits and aligns the reader/audience’s pleasure with Alisoun’s at this point in the narrative. We are in on her joke. Medieval grammarians understood vocalizations such as giggles as utterances with ambiguous grammatical status signifying mental feeling (passiones animi, affectum commiti animi). A giggle functions like an interjection. It has interjective force. It arises quickly, if not explosively, without deliberate intention and signifies according to modus affectus rather than modus conceptus. Priscian had claimed that vocalizations such as laughing, hissing at someone (sibilus), and sighing (gemitus) are understandable but are not usually writable and therefore not part of grammatica, vox litterata (Institutiones 1.2-3; KGL 2: 5-6). But the Latin grammarians’ sometimes rigid alignment of “grammatical” with “literate” rendered their descriptions incoherent on this point, as any number of classical and medieval texts (Plautus, Terence, Apuleius, Miller’s Tale) make clear. Latin grammarians claimed such vocalizations are “illiteratae” even as they were perfectly content to inscribe both conventional interjections (e.g., heu, vah, papae) and natural, word-like interjections such as a laugh (ha ha ha) or expression of disgust (phy) as examples of vox (e.g., Priscian, Institutiones, 15.41; KGL 3:91. Cf. “laetitiam animi significamus, ut aaha” in Charisius, Ars grammatica, 2.16; KGL 1:238). Interjections keep escaping the Latin grammarians’ classifications. There are different kinds of laughs. As speakers acquire a language, they learn to understand the cultural significations of different kinds of laughing in different contexts. A belly laugh connotes a different situation and meaning than does snickering or giggling. Ælfric (c. 955-c. 1010), following Priscian in his grammar, says that “haha and hehe getacnjað hlehter on leden and on englisc, forðan ðe hi beoð hlichende geclypode” [haha and hehe signify laughter in Latin and in English because they are called ‘laughing’] (2001: 279; cf. Hobson 2018). Some medieval English texts represent laughter directly as “Ha,” but more often indirectly as a noun (OE hleahtor, ME laughter) or in reported speech with a referential active verb (e.g., “And whan a wight is from hire [Fortune’s] whiel ythrowe, / Than laugheth she and maketh hym the mowe” [TC, 4.7]; cf. Sir Gawain and the Green Knight, line 69). In Beowulf, both noun and verb forms are used (lines 611a, 730b). Some texts index laughter more indirectly with nouns such as OE dream or ME joie. In Middle English different laughs or intentions were sometimes marked with adjuncts or qualifiers: e.g., laugh ascorn (derisive laughter). A giggle such as “Tehee” is subtle, a high-pitched light laugh rather than a belly laugh. The Peanuts cartoons regularly contrast Lucy’s boisterous “HAHAHA” with the sly Woodstock’s quieter, subtler or mocking “heehee,” and both forms are attested in ancient Greek written representations of

Alisoun’s Giggle, or the Miller Does Pr agmatics

147

laughter. Woodstock’s subtle laugh invites the reader to sympathize or agree with him. Lucy’s does not. Chaucer’s use of tehee as a direct performative utterance is unusual and more restricted than other Middle English interjective expressions and oaths. His assignment of the giggle to Alisoun also correlates with the mostly female register of tehee and related forms (teehee, tee and hee, etc.) as recorded in the OED. Chaucer’s inscription gives the pragmatic gesture a certain grammatical status as a conventional written sign and narrative salience. Affectively, tehee performs Alisoun’s mischievous delight and connotes her trickster agency in the narrative. We could paraphrase the giggle in reported speech – “She giggled.” – but the narrative represents her delight pragmatically and gesturally in direct speech, as modus affectus rather than modus conceptus (cp. heu vs doleo). Pragmatically, tehee conveys her lack of sympathy for Absolon’s disgust and humiliation and her pleasure in using her body and desirability as weapons to deny Absolon what he most desires, to prank him where it most hurts. Pulling the shutter closed in his face further connotes Alisoun’s agency as she controls Absolon’s access to her body. The gesture functions as a pragmatic utterance and a physical full stop to her performance. It also functions as a narrative discourse marker, signalling the transition to the next episode (scene change), which turns out to be a multiplication of pragmatic contexts, Nicholas’ undoing, and Absolon’s revenge. The three interjections discussed so far are part of the tale’s broader thematizing of the pragmatics of utterance, including polysemy, gestures, elliptical sentences, and a radical recontextualization of normative semantics. These thematic nodes were also addressed by medieval grammarians’ theories of equivocatio, constructio, and perfectio ad intellectu. All texts have some pragmatic structure, functions, connotations, and effects. But the Miller’s Tale thematizes or foregrounds pragmatic intents, significations, and effects from the narrative positions of different characters/actants. The narrative puts focalization into dialogical spin. The tale’s dissemination of focalization disrupts the notion that an utterance has a stable meaning and narrative authority from the standpoint of only one speaker or participant. Pragmatically and dialogically, the Miller’s Tale decenters authorial and speaker authority.

Impoliteness, Hedging, and Textual Pragmatics The pragmatic critique of meaning and communication informs the Miller’s Prologue and Tale in other ways besides interjections. The tale also foregrounds narrative focalization, equivocation, (im)politeness, and

148 

The Medieval Life of L anguage

face (threats) as strategies for pragmatic and dialogic reaccentuation. These pragmatic strategies and metapragmatic awareness are interconnected, as we shall see, with the prologue and tale’s reflexive narrativity. Brown and Levinson’s (1987) theory of pragmatics and politeness extends Goffman’s notion of “facework” (1967: 5-45) and proposes an alternate version of Grice’s Cooperative Principle of “rational” discourse. Goffman defined “face” as “the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact” (1967: 5). Brown and Levinson argue that maintaining one’s face and avoiding or mitigating “face-threatening acts” (FTAs) are the basis for successful oral and some written communicative interactions and the ground for rational and cooperative interaction. They begin their description of politeness strategies with speech acts, such as making a request or issuing a command by using a question form instead of an imperative. Adapting Grice, Brown and Levinson argue that a “rational” speaker (we can add writers in formal, informal, and online situations) seeks to avoid threatening another’s face and to cooperate in order to achieve common goals. Positive politeness entails using discourse strategies which minimize threatening or damaging the listener’s positive face (sense of self) and which support the listener’s self-esteem (e.g., compliments, solidarity, hedges, avoiding conflict). Negative politeness entails using strategies which avoid imposing on the listener and thereby threatening their autonomy or independence. Brown and Levinson also distinguish between “bald on-record” and “off-record” utterances. Bald on-record speech most risks threatening the listener’s face; speaking that way without much risk usually presupposes a close relationship between speakers (e.g., family, longtime friends). Off-record speech, on the other hand, relies on minimizing face threats and impositions on the listener, uses indirect utterances, and relies on the listener to infer the speaker’s intent (1987: 61-65, 94-95). The pragmatic concept of impoliteness evolves from Brown and Levinson’s Politeness Theory but shifts the orientation in significant ways (e.g., Culpepper 2011). Doubtless, polite and impolite discourse have been around as long as humans have been speaking. It’s Pragmatics the discipline whose orientation has shifted. Some theorists of impoliteness call Politeness Theory to account by noting the theory’s “inability to adequately account for impoliteness by the same concepts that explain politeness” (Eelen 2001: 245). This critique implies that politeness and impoliteness are symmetrically ordered with respect to speech and pragmatics. However, as Leech (2014: 81-86) notes, indirect utterances are not necessarily strategies of negative politeness, especially when the utterance conveys impolite or rude

Alisoun’s Giggle, or the Miller Does Pr agmatics

149

information or beliefs. The indirect expression ‘Could you turn off your phone?’ with articulatory stress on the verb particle (off) could indicate the speaker’s irritation rather than any face-saving effort. Treating impoliteness simply as the opposite of politeness doesn’t really account for the complexity or agency of hostile interactions, conflict talk, and how we process bald on-record utterances. Nor does Politeness Theory sufficiently take account of how speakers can and often must negotiate unequal power relations in order to participate in particular exchanges. (In Chapters 5 and 6 I’ll discuss how Margery Kempe and William Thorpe deal with face threats and the pragmatics of (im)politeness.) Critics have analyzed historically the language and narrative strategies of the Miller’s Tale as being drawn from the fourteenth-century’s elite and less elite social and poetic registers, parodying the aristocratic language of the Knight’s Tale, embodying the “most vigorous oppositional force within [Chaucer’s] society – the rural world of peasant culture,” or presenting a “mixed style” debate about class consciousness from a bourgeois point of view (e.g., Muscatine 1957: 223-230; Muscatine 1976; Donaldson 1970: 13-29; Patterson 1991: 254-258). These readings describe the fictional discourse and characters’ speech as reflecting or staging the clash between different medieval estates codes and genres. However, from a pragmatic perspective, talk as represented in the Miller’s Tale presents a complex and dialogic textscape of (im)politeness and disruptive polysemy (equivocatio) in direct and reported speech. The text depicts discursive power as flowing in several directions, discursive force moving like a fluid constellation through the social world rather than fixing in one place or with one speaker. The Miller’s Tale verbally enacts the creative force and the life of language, which Patterson characterizes as the tale’s “narrative staging of the vitality and resourcefulness of the natural world” (1991: 258). The tale smudges the boundary between the natural and the conventional by giving narrative concreteness and physical embodiment to the constitutive power of reaccentuation and pragmatic dialogism. Impoliteness reigns in the Miller’s frame narrative and tale as narrators and characters hurl insults and swear at one another. Some affirm their certainty or sincerity, some aggressively attack other characters and threaten their social status or face, several just talk tough to assert their social status and masculinity. With respect to politeness, the Miller’s Prologue and Tale are weighted toward oaths, swearing, and insults rather than to “positive face.” As impolite and threatening speech, some insults are grammatical utterances, some are interjections, all are hostile. Once we start to think about it, many of the Canterbury Tales are full of insulting talk, and the

150 

The Medieval Life of L anguage

tales themselves often function as indirect insults within the pilgrimage tale-telling fiction. The Miller’s Prologue and Tale is just one among many. In the Miller’s Prologue, the voice of narrative report, which I’ll provisionally name the supernarrator “Chaucer,” contrasts spoken “curteisie” with the Miller’s loud “Pilates voys” and the speech act verb “swore.” This nexus of insults and aggression foregrounds the Miller’s interrupting speech and explicitly tries to excuse his rude challenges to those who say he is drunk and disorderly: The Millere, that for dronken was al pale, So that unnethe upon his hors he sat, He nolde avalen neither hood ne hat, Ne abyde no man for his curteisie, But in Pilates voys he gan to crie, And swoor, “By armes, and by blood and bones, I kan a noble tale for the nones, With which I wol now quite the Knyghtes tale.” (1.3121-3127)

In just the first twenty-five lines of the Miller’s Prologue, the narrating, sweary, cantankerous Miller is not alone, as other pilgrims also swear casually “By [Christ’s] armes, and by blood and bones” (1.3125), “Goddes soule” (1.3132), “a devel wey” (1.3134), etc. The verbal soundscape is aggressively conversational and full of affective speech, most of it threatening others’ face. The Miller’s drunken voice is replicated in the voices he gives to the characters in his tale. The Miller’s blasphemous oath backs up his promise to “quite” the Knight’s tale, and in the tale John the carpenter similarly swears by Jesus’ body and blood (1.3464, 3508, etc.) and numerous saints. Swearing for emphasis, attack, or to guarantee a hostile promise (‘By God, I’ll …’) are associated with rural and suburban tradesmen lacking indirect courteous habits. In terms of social status, the narrative ethos initiating the Miller’s Prologue and Tale is not mercantile and certainly not “curteis.” Rather, it’s impolite and fundamentally exclamatory, aggressive, rude, and rough, which in medieval social codes are counted as affects and speech acts associated with non-courtly and lower-class persons. The Miller’s refusal to abide by the norms of positive and negative face “curteisie” is matched later in the tale by Nicholas and Alisoun’s subversive parody of the proprieties of flirting. Some readers have interpreted the two as acting like incompetent courtly types, but I read their speech and actions to be intentionally parodic within the narrative diegesis. That is, it’s the game they play. They are deliberately, metapragmatically, playing

Alisoun’s Giggle, or the Miller Does Pr agmatics

151

around with the presupposed, polite conventions of flirting and seduction. We might understand them as going through the motions with a knowing wink or laugh as part of the foreplay before the pley itself. When Nicholas suddenly grabs Alisoun by her “queynte” (1.3276) and holding her tight against his thigh declares “For deerne love of thee, lemman, I spille” (1.3278), Alisoun appeals to the higher class-marked code of deference and courtesy, especially with regard to women’s status: “‘Why, lat be!’ quod she. ‘Lat be, Nicholas … / Do wey youre handes, for youre curteisye!’” (1.3285, 3287). In the very next line, Alisoun also threatens to raise the public alarm by using the conventional interjection and cry “out, harrow” (1.3286; cf. Nun’s Priest’s Tale, 8.3380-3382). In fifteenth-century French depositions, the idiom for calling for help when a crime was in progress was “Haro! Haro!” (Jager 2014: 75 and n.30). Alisoun’s sudden appeal to social courtesy and physical distance clashes with the Miller and supernarrator’s earlier naturalizing description of Alisoun with flora and fauna images. However, when in the next line Nicholas persists and pleads in courtly fashion for her “mercy” (1.3288), she quickly consents, and like a savvy bridge partner follows his lead as they plot their liaison. In this brief exchange, Alisoun’s first speech act of subjective agency is a quick two-step. First, she refuses Nicholas’ advances, then she consents. The episode is parodic within the narrative diegesis, as Nicholas’ words and behaviors are exaggerated and Alisoun’s sexual refusal is explicit but shortlived. Both of them seem to be explicitly, self-consciously performative. After the supernarrator’s descriptions of John, Nicholas, and Alisoun, which read slowly as if in cinematic stop motion, the narrative starts to pick up speed and intensifies the repetition and polylingual play of loaded words, including “queynte” (clever, female genitals), “derne,” “spille,” “pley,” “hende,” “ore,” and so forth. Donaldson and Neuss have teased out many of the punning implied meanings and polylingual semantics of these words, especially pley (Donaldson 1970: 13-29; Neuss 1974). Nicholas and Alisoun speak the vocabulary and syntax of romantic and pseudo-courtly politeness, but their goal is physical and emotional pleasure with each other. They use the words but with a different semantic and pragmatic valence, reaccenting them for fun. Alisoun seems to play with the socially-approved norm of female “curteisye” when what she really wants to do, according to the narrator, is “rage and pleye” (1.3273), though she never says so herself directly. Moreover, her light refusal in the early scene contrasts sharply with her more aggressive pley later on when she uses her body to trick Absolon, another extramarital suitor seeking her “mercy” with romantic, courtly language which he can’t really inhabit.

152 

The Medieval Life of L anguage

Hedging speech can mitigate the face-threatening force of an utterance or a Gricean maxim (Mey 2001: 151). In pragmatic terms, the most explicit deferential or hedging speech in the prologue or tale occurs early on when the Miller condescends to placate the Reeve, who has taken offense at the Miller’s promise to tell a tale of domestic betrayal between a carpenter and his wife: This dronke Millere spak ful soone ageyn And seyde, “Leve brother Osewold, Who hath no wyf, he is no cokewold. But I sey nat therfore that thou art oon;” (1.3150-3153)

Having brushed aside the Host and the Monk, the Miller tries to create solidarity with his fellow artisan by calling him “leve brother.” Refusing to confirm the Reeve’s inference that he has baldly threatened the Reeve’s face, the Miller hedges by proposing that the story he will tell reflects on whoever fits the storyline, but that that is something the listener must decide for themself. The Miller seems to be juggling explicit and implicit modes of address with respect to his storytelling. At first, the Miller baldly proposed his tale in order to “quite” the Knight’s, but by the end of his Prologue the tale he teases his audience with seems more relevant to the Reeve, and the Miller tries to distance himself from the Reeve’s inferred face threat. Conversations are performed as ordered sequences, as CA has shown, but the dynamics of conversational interaction, the push and pull of speech as social action, can shift a speaker’s discursive footing from one position or supposition to another, depending on the flow of power or the stance of the speaker in the exchange. Goffman defines footing as “the alignment[s] we take up to ourselves and the others present as expressed in the way we manage the production or reception of an utterance” (Goffman 1981: 128). During the time of a talk, contexts, presuppositions, and implied or explicit meanings can shift, emerge, be reinforced, or fade away. Just before the Miller launches into his tale, he withdraws his speech from any direct connection with the previous tale or with any individual pilgrim (l.3153). Structurally, as readers have noted, the Miller’s Tale dramatically interrupts what would have been a different, more hierarchical opening sequence of tales (General Prologue, Knight, Monk). We know that the opening section of the Canterbury Tales was one Chaucer had carefully worked over before his death. The Miller’s Prologue brutally shifts the footing of the tale-telling order as conceived by the Host and initiates a deliberate, intertextual interrogation of narrating voice, style, genre, and social status. But after interrupting the

Alisoun’s Giggle, or the Miller Does Pr agmatics

153

Host’s ordering of tales, the Miller moderates his conversational stance and hedges on the larger interpersonal context for tale-telling he had opened up. Hedging and distance prevail. Thinking of a text as a polyphonic conversation or Bakhtinian dialogue, we read the authorial order of the text (Canterbury Tales, Part One) as representing a destabilizing of the initial textual form or ‘idea’ by the interpersonal and narrative pragmatics of the narration. The Miller’s Prologue then mitigates that interruption when the narrating Miller withdraws from direct interpersonal face-threatening. Hedging is a kind of pragmatic management strategy, a staged withdrawal, as a speaker attempts to mitigate an utterance or act which they think might provoke or has provoked a strong response to a face threat. The problem with hedging language – but really, it’s a problem with any utterance – is that once something is uttered or written and circulates within a speech context, no matter how restricted or broad, I can’t take it back or erase it entirely. I can only take back or hedge on what I have said by saying something more. Or I can say something and hedge it at the same time. Writing may give us the opportunity to revise what we have drafted, but not when I release the text into public discursive space. I can add something or change something in the text, but I can’t revise the public text itself. Hedging is addition, not erasure. As a discursive strategy hedging produces another utterance inside of a version of the utterance, one less intense, less certain, or less direct. Even good-natured ribbing or ritual insulting still leaves the face threat out there in public as a potential attack if the context or attitude changes. When someone goes too far and crosses a line, when the talk threat becomes too direct or too specific or is taken as genuinely face-threatening by the listener, as when the Reeve thinks the Miller is telling a story about his family, then shared civility, solidarity, politeness, and social cohesion can dissolve into alienation, hatred, or violence. If Nicholas and Alisoun’s flirting and foreplay self-consciously parody the courtly talk they imagine their social superiors using, the Miller and Reeve’s interaction represents the extent to which hedging and social tolerance are also maintained only by the willing agreement of those involved. Judging by the Reeve’s Tale as a dialogic response to the Miller’s utterance, the Reeve can’t be understood to have accepted the Miller’s hedge. We can rethink the idea of the Canterbury Tales not as roadside or horseback drama but as a pragmatic interrogation of our assumptions and imagined outcomes for how talk and interaction create, maintain, or disestablish community relations, how associations are made and threatened, how relevance and context are continually constructed and reconstructed. Such a pragmatic reimagining of the text helps us discover in the comic structure of the

154 

The Medieval Life of L anguage

Miller’s Prologue and Tale a sophisticated metapragmatic representation and micro critique of accepted or received meanings in social interactions and the possibilities for alternate hegemonies. The interplay in the Canterbury Tales between text and talk as part of individual tales’ intertextuality adds a further level of textual pragmatics to the semiotic dialogism and linguistic field (cf. Amsler 2012a). The Miller’s initial hedging around face threats in the prologue to his tale becomes textually contagious and forecasts the tale’s metapragmatics. After the Miller strategically hedges his tale’s relation to any immediate pilgrim, the supernarrator – that interpretive construction referred to as “Chaucer” – intervenes and takes hedging and mitigation further. The well-known passage slyly interweaves different narrative levels within the tale telling situation: M’athynketh that I shal reherce it heere. And therfore every gentil wight I preye, For Goddes love, demeth nat that I seye Of yvel entente, but for I moot reherce Hir tales alle, be they bettre or werse, Or elles falsen som of my mateere. And therfore, whoso list it nat yheere, For he shal fynde ynowe, grete and smale, Of storial thyng that toucheth gentillesse, And eek moralitee and hoolynesse. Blameth nat me if that ye chese amys. Millere is a cherl; ye knowe wel this. So was the Reve eek and othere mo, And harlotrie they tolden bothe two. Avyseth yow, and put me out of blame; And eek men shal nat maken ernest of game. (1.3170-3186)

The supernarrator’s casual speech (“M’athynketh”) is filled with hedges and discursive intimacy or involvement with an audience: “every gentil wight I preye, / For Goddes love,” “Blameth nat me,” “ye knowe wel this,” “Avyseth yow.” That’s what they say. Don’t take it too seriously. I’m gonna do this now. It’s all good. This passage elaborates the dialectic of ernest and game in the Canterbury Tales and the tension between aural and written reception with the simultaneous emergence of the authorial voice within the fictional text. But there’s another, no less significant aspect of the passage more relevant to our theme. The supernarrator’s discursive maneuver is a complex, subtle

Alisoun’s Giggle, or the Miller Does Pr agmatics

155

bit of pragmatic hedging and metapragmatic textuality. In the midst of the text’s fiction of oral tale telling, perhaps first rendered by Chaucer the poet-man for a small group reading aloud session in a London neighborhood, the narrator’s metanarrative digression directly addresses – but how? are they in the room with books? who’s reading? or is this an imagined readership? – a book-reading individual or group. The speaker hedges on what the supernarrator wants the reading audience to do with the story he teases them/us with. In effect, the supernarrator says: “You (all), ‘every gentil wight,’ have access to a book. You know how to use it. I’m just letting you know what to expect in the next part. It’s not my fault if you don’t like what you read. Be active, discontinuous readers. Make your own decisions. And take responsibility for your choices because I can’t be held responsible for what you do.” The implicature is not unlike that of the “Stay Alert” advice from many governments during the 2020 Covid-19 pandemic: “You have to be responsible for your own health. We can’t really be held responsible for your behavior.” Perhaps that’s just another way of saying, “You’re on your own. It’s your fault if something bad or unwanted happens.” The supernarrator’s language shifts not only agency and autonomy but also responsibility on to the readers addressed in the text. The narrative prolepsis here embeds genre expectations, social stereotypes, speaker sincerity, and good faith in a winking performance of literate hedging, which adds to the pilgrimage narrative presented thus far rather than erasing any part of it. As the Miller’s Tale proper is about to get underway, the supernarrator invokes the empowering medieval tactic of discontinuous reading, advocated in devotional and scholarly reading, as a textual hedge against an imagined audience’s negative response to having been made to listen to or read a story which they might find socially or personally objectionable or not to their ‘taste’ (on discontinuous reading, see Amsler 2011: 116-117, 207). The supernarrator’s rhetoric of sincerity and gesture of good faith alert the audience to disturbing content and style ahead, but of course that good faith speech may also serve as a narrative tease. The passage transforms an oral recitation context into a book reading one controlled not entirely by the teller but also by other readers, readers at a distance. The Miller’s hedge is also a tease, and some (many?) readers might actually be enticed rather than put off by this strategy. Like Alisoun, the supernarrator is engaging in hedging as seductive foreplay. The distance textuality makes doesn’t alter the immediate discursive situation of the Miller’s Prologue, but individual readers can respond and participate in different reading formations. The passage sets a pragmatic practice for the rest of the unfinished Canterbury Tales, a practice which I think has been

156 

The Medieval Life of L anguage

too often overlooked or avoided in Chaucer criticism. At this moment, the multi-addressed text of sincere speech opens several spaces for different kinds of audience involvement, with speech, book, and reading gestures, or perhaps a single participant with multiple imagined reception practices (cf. Chafe 1982). The supernarrator’s discursive hedging pushes out toward centrifugal textuality and disseminates interpretive authority. Is Alisoun giggling yet?

Polysemy, Bullseyes, Misfires, or How Narrative Escapes Intention The discursive gesture of sincerity and good faith at the end of the Miller’s Prologue imaginatively presupposes there is a some-one, the “authorial” supernarrator, who addresses the reader or listener directly and can demonstrate good faith and control meaning. However, the complex ethical tone, action, and speech registers of the Miller’s Tale are established by a more complicated genre code. Fabliaux, like Greek New Comedy and trickster narratives, are constructed around a series of structural oppositions, between the devious and the gullible, the plotters and the dupes, the real and the charade, the truth-exposer and the hypocrite, youth and age, the lovers and the obstructers. Youthful sexual energy and skillful language pley trump jealousy, old age, greed, and literal or singular referentiality. Fabliaux also target with undercutting laughter many characters who presume moral superiority. One strategy narrative tricksters use to achieve their aims is good theater, the creation of illusion. Nicholas’ scheme is to get into bed with Alisoun. To do so, he in effect creates an elaborate stage act, complete with props, images, script, set design, choreography, narrative plants, and Alisoun’s consenting support role. Nicholas and Alisoun stage the elaborate astrological ruse to trick John the husband/carpenter into believing a Second Flood is coming (l. 3487-3600). The premise enables them to convince John to use his skills to build an elaborate set in-house with tubs in the rafters filled with victuals to survive the foretold catastrophe. Guided by Nicholas as director, principal actor, script writer, and set designer, John effectively transforms his two-level house into a stage. The action and physical layout suggest a relation between the tale and the cycle plays’ physical performance context of pageant wagons. The tale/theater relation is underlined with references to dramatic performance (e.g., the Miller’s “Pilates voys” [1.3124], Absolon playing Herod [l.3383-3385]), and Noah’s Flood as part of the ruse itself. Even the word pley(nge) acquires quadruple meaning in different

Alisoun’s Giggle, or the Miller Does Pr agmatics

157

contexts: having fun, having sex, playing music, acting in a play (see 1.3660, 3273, 3333, 3384, etc.). The young couple’s diversionary theater foregrounds their performativity, especially Nicholas’ performative language, with its exaggerated speech of prophecy and apocalypse, which controls the plot, at least until Nicholas takes the script too far. Meanwhile, Absolon’s performance is presented as ridiculous. He’s the target of Alisoun’s rude trick, a foil in someone else’s script. In the second half of the tale the medieval pragmatic concepts of equivocatio, perfectio ad sensum, and perfectio ad intellectum are subject to vigorous narrative critique. Individual characters act on their intentions and desires for semantic and social control, but their acts and speech are continually undermined through narrative turnarounds, verbal equivocation, and others’ radical recontextualizations and reaccentuations. The tale subverts the monologic of the authoritative intentional act or speech act with pragmatic reimpositions and dialogic equivocation. In Chapter 1 we saw how the logical and pragmatic topic of equivocatio (polysemy, verbal ambiguity, analogy, reimposition, figurative language) was important and productive in medieval pragmatic theory. Olivi investigated the semantic potential of words beyond immediate explicit intention in terms of virtus sermonis: Does the text or speaker really mean all that? (Quaestiones Logicales, q.7; ed. Brown 1986: 353). The question also applies to the potential semantics of poetic language. In a given situation individual utterances may be understood differently and ‘correctly’ by different participants according to different semantic contexts and presuppositions and their situated knowing in different interpretive or epistemological positions (deixis). According to Austin, Grice, and many medieval logicians, disambiguating propositions according to speaker intent and audience cooperation is the principal criterion for confirming the meaning of an utterance and pragmatic competence. Medieval pragmatists, however, perhaps with scriptural interpretation in mind, frequently argued that grasping the changing circumstances of an utterance and the differences between literal and non-literal or figurative meanings were key to understanding the linguistic potential of multiple meaning and markers of pragmatic sophistication, even though the same argument was turned around to justify the clergy’s privilege to interpret scripture correctly in the face of increasing lay involvement with religious texts. Medieval pragmatic theory afforded relevance and interpretive authority to both speaker and listener’s interpretations of utterances, but not always equally or easily. It’s not hard to recognize how the power of the pragmatic concept of equivocatio permeates the Miller’s Tale. We have already pointed to several

158 

The Medieval Life of L anguage

passages where ambiguous or outright lying language initiates plot turns and subverts stable, monologic interpretive contexts among the characters. In the tale the tacit community is continually fragmented by insincere and bad faith utterances. The notion of a relatively stable set of communi intentiones is continually subverted. Moreover, control of interpretive and semiotic context shifts from scene to scene, as characters (actants, speaking subjects) remake relevance and context from different deictic positions. The narrative evades a consistent focalization. Certain words and phrases become salient in the Miller’s Tale precisely because different speaking subjects (actants) and the “authorial” supernarrator understand or use them differently in different contexts. The salience of key words and phrases is an implicature of their referential instability and deictic dialogism, as speakers and narration itself retext words and phrases from the Knight’s Tale, medieval drama, scripture, and everyday speech. These words become salient in the tale as puns or polysemes with different but simultaneous or accumulated relevance contexts for different listeners or readers: e.g., “queynte” (‟As clerkes ben ful subtile and ful queynte; / And prively he caughte hire by the queynte” (1.3275-3276); ‟deerne” (1.3278, 3297); ‟berd” (1.3737, 3242); ‟pryvetee” (1.3164, 3454, 3493, 3558, 3603, 3623; cf. 1.3276); ‟sencer/ynge” (1.3340-3341); ‟oore” (1.3726); ‟pley” (l.3306 etc.) (cf. Neuse 1974). Interestingly, “hende” is equally polysemic (courteous, nice, gentle, handy, quick with his hands), but the epithet is used exclusively to characterize Nicholas. Some of the semantic slippages are worked out explicitly in the tale; others remain implicit and rely on readers’ and audiences’ semantic and narrative competence. That a word can mean something doesn’t mean it does mean something in a given context, as Olivi acknowledges. That is the pragmatic lesson. One character or set of characters understands salient words to intentionally mean one thing, while another character or set of characters understands them to intentionally mean something else. Polysemy and trickery are part of how fabliau narratives work, as many medieval French fabliaux exemplify. In most fabliau narratives, characters’ understanding of particular utterances is motivated by their self-interests. What the text and the characters collectively lack is any stable, overarching context for interpreting the various speeches in the narrative. There is no univocal meaning. The diegesis itself becomes a centrifugal force in language. If there is such a stable semantic and referential position, it resides outside the text in the reader’s consciousness over the time of narration. Mostly, however, readers are more aware of the polysemy and therefore as readers occupy more than one subject or deictic reading position.

Alisoun’s Giggle, or the Miller Does Pr agmatics

159

The thematization of equivocatio and perfectio ad intellectum in the Miller’s Tale reaches a climax when Nicholas screams, ‟Water! Water!” (l.3815) after Absolon pushes a hot poker onto his butt. This narrative highwater mark brings together the two plot lines in exquisite metapragmatic play. Intertextually, Nicholas’ cry appropriates the grammarians’ example of elliptical speech (‟Aqua. Aqua.”), but contextually his expression pushes toward an indeterminate interpretation, or at least polysemy. Initially, Nicholas’ elliptical utterance functions as an interjective cry and urgent request for help.7 But the Miller’s Tale exploits semantic equivocation by multiplying meanings and pragmatic contexts such that all the characters understand the single utterance at the same moment, but each understands it according to his or her individual interpretive (deictic) position and expectation. From the narrational and aesthetic point of view, these multiple significations are knotted together to bring the tale to a dramatic and artful climax, almost. The syntactically elliptical ‟Water! Water!” ignites a narrative explosion and simultaneously functions as something like what Lacan calls a point de capiton, an anchoring point, however illusory, in a dialogic discourse. A point de capiton draws the elements of a discourse or utterance or subjectivity together to implicate, however contingently, a coherent “whole” or “unified subject.” The utterance anchors the intertextual network such that the “signified and signifier are knotted together” within the tale’s language and semiotic play. The point de capiton, to intertwine Lacan’s theory of meaning and slippage here with the tale, “stops the otherwise endless movement of the signification” of the tale and produces a moment of fixed meaning among the characters, a fixity which turns out to be only temporary (Lacan 1977: 303; 1993: 268). In fact, the tale takes contingency further and instantiates multiple points at the same diegetic moment, further unmooring fixed meaning. The text motivates both centrifugal and centripetal forces in language. In pragmatic terms, Nicholas’ cry knots the different narrative threads together and initiates different implicatures and uptakes. All the characters’ responses to Nicholas’ cry are in fact physical actions, pragmatic 7 According to Stephen Toulmin, Wittgenstein used the same example in his Cambridge University lectures around 1946-1947. From his lecture notes and memory, Toulmin reconstructs a version of the example in Wittgenstein’s voice: “Suppose a young child who has been playing outdoors runs into the house and grasps the kitchen tap, calling out as he does, ‘Water, water’ – this being a word he heard used for the first time only yesterday. And suppose someone now raises the question, ‘Is the child telling us something, or showing that he has learned the meaning of this word, or asking for a drink?’ What are we then to do? Need there be any way of answering that question?” (Allan Janik and Stephen Toulmin 1973:228). Wittgenstein’s reading of the expression as indeterminate has more in common with Chaucer’s text than with Magister Johannes’.

160 

The Medieval Life of L anguage

acts. Each is ‘correct’ with respect to the presupposition and context each character deploys to interpret and respond to the utterance. Retrospectively, we realize how the ordinary word water has become more and more salient, more relevant, as the narrative proceeds, indexing the Flood plot and then Nicholas’ cry for pain relief. The word also intertextually underlines the relevance of the two plots to one another while pushing the tale into the metapragmatic zone of textual instability. The closest known analogue (Beidler [2015] thinks it’s the source) for the Miller’s Tale is the Middle Dutch Heile van Beersele (14th c.), the story of a prostitute and her male clients, including a priest. After the priest has sex with Heile, he is persuaded to play the misplaced kiss trick on Hughe the blacksmith, another of Heile’s clients who has arrived at her house. After Hughe kisses the priest’s butt, he returns enraged with a hot poker, which he lays on the priest’s bottom. The priest cries out in pain, ‟Water, water, ic ben doet” [Water, water. I am dead] (Brussels, Royal Library, MS II.1171, l. 149). Chaucer’s narrative significantly embellishes the plot in the Middle Dutch text and switches various characters’ roles, but the English tale repeats the colloquial and elliptical cry of pain with interjective force. The elliptical utterance performs an urgent commannd or plea for help. The reduplication of the noun without a congruent verb connotes urgency (cp. allas, allas). However, Chaucer’s vernacular text intersects with medieval pragmatic theory in a deeper, more complex way than just citing or reworking a Dutch source. The Hous of Fame alone testifies to Chaucer’s familiarity with and intellectual interest in aspects of grammatical theory. The Miller’s narrative reaccents medieval grammatical and pragmatic theory in a critical and performative way. When Magister Johannes and Pseudo-Kilwardby discuss the communicative potential of interjections and elliptical (imperfecta) utterances such as ‟Aqua. Aqua.” shouted in front of a burning house, they, like contemporary pragmatists, argue that the form of the utterance doesn’t have to map directly or literally onto the speaker’s intended meaning for the expression to be communicative or meaningful. Emotionally charged utterances such as “Bene” and ‟Aqua. Aqua.,” says Magister Johannes, are perfecta when understood by the listener. The construction performs a speech act, command, with the affect ‘urgency’ precisely because it is elliptical. At stake at this moment in the Miller’s Tale is nothing less than the status of grammatica in everyday speech (usus) and the role of context for establishing linguistic meaning. As we noted earlier, pragmatically-oriented medieval grammarians took many of their examples from everyday speech. ‟Aqua. Aqua.” is just the sort of cry they might hear in a medieval city or village when fire breaks out.

Alisoun’s Giggle, or the Miller Does Pr agmatics

161

Like the Heile van Beersele, the Miller’s Tale retexts, reaccents the grammatical example of an interjective utterance. The narrative embeds the utterance in multiple, more complex narrative contexts which simultaneously activate the potential meanings of the utterance in an explosive, centrifugal narrative moment. Nicholas instinctively cries for water to soothe his pain. John thinks the Second Flood has arrived and cuts the ropes holding his tub in the rafters. Yes, the words mean all that at once, collectively in the text, beyond any speaker’s intention or agency, beyond even what Olivi might have imagined. Here, we might reimagine the text itself as a “nonhuman actor” which helps modify states of affairs, shape outcomes, and influence actions in a renovating pragmatics of interpretation and reaccenting discourse (cf. Felski 2011: 583).

Centrifugal Narrative Contracts I said Nicholas’ scream knots together the plot threads and brings the tale to a dramatic and artful climax, almost. In fact, Nicholas’ cry is not the last plot turn. The Miller’s Tale has another narrative turn to make, one which further complicates the text’s pragmatic meaning and problematizes the ethics of pragmatics in the shadow of equivocatio. Just as the text’s narrative energy seems to have reached its most centrifugal point, the text turns and asserts a more normative, stable centripetal mode through genre codes and social class domination. The result is socially condescending. The f inal episode of the Miller’s Tale questions the foundations of a rational and cooperative pragmatics based on sincerity and truth-telling, while simultaneously and disturbingly reproducing the power of dominant hegemonic discourse. Alisoun has had her giggle. Nicholas has tried to pile on, but he’s been scalded by Absolon’s misdirected revenge on the object of his desire. “Water! Water!” has alerted John to what he imagines is the impending flood, and he has cut the ropes, sending the washing tub into the house, breaking through the upper floor, and dumping everyone on the ground in the middle of the night. Amid this wild scene, John, dazed and confused with a broken arm, tries to explain what he believes has happened to the townspeople who have come to “gauren … kiken and … cape” (1.3827, 3841). He was protecting himself and his household, he tells them, from “Nowelis flood” (1.3834). Chaucer’s seeing and motion verbs (gauren, kiken, cape) are pragmatically and semantically precise for rendering the Oxford townspeople’s attitudes toward and understanding of what they are looking at. The verbs all connote

162 

The Medieval Life of L anguage

the observer’s negative attitude toward the seen object. As John tries to tell “his reson” (1.3844), he is made into a nasty spectacle, framed and ridiculed by Nicholas and Alisoun and laughed at by the townspeople, especially the clerks (“For every clerk anonright heeld with oother”; 1.3847). Reaccenting the Miller’s more hedging language (“leve”) from the Prologue, the Oxford clerks are described as performing and affirming their class solidarity in the street: “They seyde, ‘The man is wood, my leeve brother’” (1.3848; cf. 1.3846). Collectively, they treat the carpenter’s speech as “fantasye” (1.3840) thanks to glosses and commentary from Alisoun and Nicholas, himself an Oxford clerk. At this moment, student talk and street talk merge to condemn the carpenter. In the shouting we can hear how some of the Oxford clerks’ negative evaluation words were used earlier in different contexts in the Miller’s Prologue and Tale. The clerks’ narrative position establishes a particular context for reading John’s “reson,” a context which continues Nicholas and Alisoun’s domestic theater in the street with more participants and different motivations and speaker focalization. John speaks sincerely what he believes is the truth, but his words lack the interpretive authority already constructed for those who listen against him: “no man his reson herde” (l.3844). Nicholas, Alisoun, John, and the townspeople presume they know the truth, however different those truths might be, but in fact no one knows the whole story. Except maybe readers and the tale’s imagined audience with access to the “authorial” supernarrator’s comprehensive narrative context. But again, that presupposes that a narrative can be fully, exhaustively understood. Although lying and pretending on behalf of sexual desire have dominated the speech of three of the tale’s characters, the language of the tale’s conclusion is caustic and othering toward just one: the married, older, religiously sely artisan. John is depicted by the principal lovers and the local community (“neighebores, bothe smale and grete”; 1.3826) as crazy, ignorant, and complicit in his own harm, someone worthy of contemptible laughter and derisive humor. The Miller, the assigned narrator, confirms such responses. Earlier in the tale, the narrator had criticized John’s “affeccioun” and killing “ymaginacioun” (1.3611, 3612). John’s lack of reading the auctoritas Cato, the narrator implied, has resulted in his unexamined devotion to his young wife and his naive, sely willingness to believe Nicholas’ fabricated prognostications contrary to scriptural truth. Now at the end of the tale the narrator adopts the standpoint, focalization, of the deceivers and the ready-to-believe-them townspeople for an additional negative evaluation of John the carpenter. Readers have pointed out how the tale wraps up in the final thirty lines (1.3824-3854) by declaring John a fool, appropriately punishing Absolon and

Alisoun’s Giggle, or the Miller Does Pr agmatics

163

Nicholas, and letting all the townspeople and clerks have a good laugh at the carpenter’s expense: “And [they] turned al his harm unto a jape. / For what so that this carpenter answerde, / It was for noght; no man his reson herde” (1.3842-3844). At this point, Alisoun’s giggle has become a wider, crueler social laugh. As far as narrative construction is concerned, however, the tale could have been concluded with a brief coda some twenty lines earlier, more or less at line 1.3823, “Upon the floor, and ther aswowne he lay.” The narrative’s metapragmatic critique helps us understand why the text continues with a new scene and speaking subjects for another thirty lines. However, I think there’s something excessive about the last thirty lines of the text. Just when everything seems to be over and the plot threads have been brought to a rambunctious close with an explosion of equivocating and centrifugal verbal meanings, the narrative withdraws from its centrifugal play of pragmatic relevances and multiple significations. The last thirty lines reassert a dominant code of fabliau ideology: wit, verbal manipulation, and scheming are successful; hypocrisy and being uncool are punished; older people are duped by younger, more savvy, and sometimes better educated people; educated clerks maintain hegemonic discursive power. Using the chaos after the crash to get their version of events out on the street first, Nicholas and Alisoun seize the discursive floor, shouting “‘Out’ and ‘Harrow’ in the strete” (1.3825). Once again, theater and interjections mark these characters’ efforts to gain pragmatic advantage and discursive control. When Nicholas and Alisoun raise the public alarm (as Alisoun coyly implied she would earlier), they preemptively overdetermine the speech context within which their and John’s subsequent utterances will be understood. Genre code and social literate power are deployed by the dominant group, to which Alisoun now belongs, while Nicholas’ scheming subverts any relevant ethos of truth telling, cooperation, or sincerity. John is left hanging in the out-group. The final Oxford street episode and brief narrative coda pose another pragmatic question: What are pragmatic ethics? In his theory of face and facework, Goffman notes that “… while concern for face focuses the attention of the person on the current activity, he must, to maintain face in this activity, take into consideration his place in the social world beyond it” (1967: 7). The characters in the Miller’s Tale are all concerned with maintaining face, especially John, Absolon, Nicholas, and the Oxford clerks. Some do so more successfully than others. The young fabliau couple’s deceit and insincere speech, fully intended and accomplished, render John a spectacle of derisive laughter among an audience ready to believe the story Nicholas spins for them. After the tale’s radical and equivocal critique of monologic

164 

The Medieval Life of L anguage

meaning and univocal relevance, the last thirty lines move centripetally and return the tale to a more stable but hypocritical social norm. In this last episode, Nicholas and Alisoun subvert any claim the story might make to be revealing truth or bringing about justice, poetic or otherwise. The tale critiques the notion of justice, poetic or otherwise. The last episode in effect hedges on the tale’s radical re-contextualization and resemanticization, which seemed to culminate with “Water! Water!” The street scene presupposes an audience for whom the witty lovers are heroes or at least successful deceivers. Recontextualizing the genre and speech register expectations of the Knight’s Tale, the Miller’s Tale adopts a complex metapragmatic orientation by insisting not only that meaning is situated but also that genre codes can work to sustain dominant hegemony and social order. As part of pragmatic framing, a genre code produces powerful relevant interpretive contexts which give meaning to individual utterances. A genre code as point de capiton can only assert a frame temporarily or contingently. Alisoun’s giggle is transformed into condescending laughter in the Oxford street and the imagined reader’s laugh at the narrative joke, mostly at John and Absolon’s expense. The reasserted genre code of fabliau as it is enacted in the streets of Oxford produces a discursive hedge which seeks to contain the narrative’s centrifugal potential and radical pragmatics. However, the text as a pragmatic act eludes that frame, and pragmatic play continues. The text overruns the code.

5

How Heretics Talk, According to Bernard Gui and William Thorpe Abstract This chapter juxtaposes Bernard Gui’s accounts of heretic speech in his Inquisitor’s Handbook with William Thorpe’s personal narrative of his interview with Thomas Arundel, Archbishop of Canterbury. Gui’s thirteenth-century text is a surprisingly rich source for medieval linguistic and pragmatic ideas. His accounts of dialogues with suspected heretics reveal how pragmatic ideas and analysis informed the official Church’s view of heretics’ uses of equivocation, hedging, and recontextualization as strategies for evading examiners’ questions and establishing speaker agency. Thorpe’s early fifteenth-century narrative describes his use of similar discursive and conversational strategies to evade Arundel’s repeated calls that he affirm or disavow his heretical views. Keywords: heretic speech, equivocation, echoic speech, pragmatics, Bernard Gui, William Thorpe, conversation analysis

In 2011 Barry Bonds, US Major League Baseball’s all-time homerun slugger, was convicted of “providing evasive testimony to a federal grand jury” in 2003 at the height of baseball’s steroid era. Before he testified in court, Bonds promised to tell the truth, the whole truth, and nothing but the truth. Prosecutors also argued, unsuccessfully, that Bonds should receive a fifteen-month prison sentence for “his pervasive efforts to testify falsely, to mislead the grand jury, to dodge questions …” (New York Times, 16 December 2011). As a result of his conviction, Bonds’ batting records now come with an asterisk. Cooperation, truth telling, and sincerity are values prized in not only western law and Gricean pragmatics but also by medieval inquisitors investigating people’s orthodoxy and defending the faith. Bernard Gui and Olivi, for example, would likely agree with Grice’s Maxim of Quality:

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch05

166 

The Medieval Life of L anguage

“Try to make your contribution one that is true.” subdivided as a) “Do not say what you believe to be false.” and b) “Do not say that for which you lack adequate evidence” (Grice 1975). And like the prosecutors in the Bonds case, medieval theologians, theorists, lawyers, and inquisitors took a dim view of evasive testimony and deceptive speech as well as heterodox and nonconformist behaviors. For them, evasive and deceptive speech violated the commonly presumed or institutionally imposed communi intentiones of the European Christian faithful. As we seek to grasp the scope of medieval pragmatic thinking, the status of vague, ambiguous, or uncooperative speech opens an intriguing and complex space for reading how pragmatic ideas and metapragmatic thinking informed both orthodox and heterodox discourses. Medieval accounts of heresy, whether from the point of view of Church officials or nonconformists themselves, manifest the struggles over what sinceritas means for evaluating different speakers’ intentions and implicatures. We might be surprised to discover that some records of heresy inquiries offer explicit and detailed pragmatic accounts of participants’ speech and interpretive interactions. In addition to medieval grammatica and narrative, heresy texts thicken the social contexts of pragmatic and metapragmatic thinking as they were applied to highstakes contexts, life and death. Questions about sincere and deceptive speech and the possibilities for resemanticizing (re-impositio) given terms and conditions also posed complex problems and solutions for establishing an ethics of communication and for determining ‘successful’ interaction and social stability. As we shall see, sinceritas was an important pragmatic value for both investigators and nonconformists or dissenters alike. Moreover, sinceritas highlights how a pragmatic ethics addresses the role of authority and agency as they contribute to the moral compass in speech interactions, especially those in asymmetric power relations. Medieval inquisitors and other investigators were especially keen to ascertain intent when examining suspected heretics. In this chapter, I argue the investigators’ inquiries sometimes took a surprising pragmatic and metapragmatic turn, as shown in official accounts of Waldensian and Lollard heresy inquiries in France and England in the thirteenth, fourteenth, and fifteenth centuries. Conversation analysis (CA) and critical discourse analysis (CDA) are useful for unpacking the pragmatic strategies and implications of the recorded or represented interactions of examiners and suspected heretics. Records of asymmetric talk regarding suspected heretics’ beliefs reveal some complex and sophisticated pragmatic performances and metapragmatic thinking on the part of both investigators and their

How Here tics Talk, According to Bernard Gui and William Thorpe

167

suspects. For the history of linguistics and pragmatics, these accounts are an important source for reconstructing medieval pragmatic thinking and practice, despite coming from hostile examiners. The pragmatic strategies described (footing, stance, hedging, discourse markers, etc.) are part of everyone’s speech, but different speakers with different motives and metapragmatic awareness do not always or necessarily cooperate nor do their strategies and presuppositions always align. Cooperation cannot be assumed to be everyone’s goal.

Pragmatic Talk, Pragmatic Action Many examinations of suspected Waldensians or Lollards reveal pragmatic analysis and metapragmatic awareness focused on the relations between words, actions, and intent with respect to the doctrines of priestly legitimacy, mass rituals, and transubstantiation. For example, in 1498, Thomas Boughton, a shoemaker and woolwinder from Hungerford (Southeast England), was examined by the Salisbury diocese on his orthodoxy, especially regarding the doctrine of transubstantiation. According to the official record of the inquiry, Boughton gave testimony that “I have every yere receyved the said holy sacrament, not for that I had any stedfast byleve therin, but that I shuld not be noted and knowen of the people. And, beyng in the church or ellyswher whan the said holy sacrament was present, I feyned with myn hondys to honoure it as Cristen men use to doo, but my mynd and entent was nothyng therto but to God almyghty above in heven, thinkyng that he was not ther present in the blessyd sacrament” (Blythe register, fol. 74; quoted in Hudson 1985: 118). The diocesan record uses Boughton’s own words to expose the difference between what the shoemaker says and does during the Mass and what he thinks or understands (“my mynd and entent”) about the signification and meaning of his religious practices. The recorded testimony does not constitute an explicit confession. Rather, the examiners use the text to infer that Boughton’s action and speech are deviant, insincere, and therefore heterodox. Boughton’s testimony (surely spoken in the vernacular) states his desire to conceal his heterodox beliefs from his immediate community and religious authorities. But he also asserts that his mental and spiritual intention was directed to God in heaven because he believed God/Christ was there rather than in the host. Boughton describes how he deceived (“feyned “) his community with outward gestures in order to be faithful and sincere inwardly to a higher principle. Boughton’s words as recorded and interpreted cut

168 

The Medieval Life of L anguage

in two directions: they assert his true intent and nonorthodox belief, and they expose, at least to the examiners, his duplicity and lack of doctrinal conformity. Boughton feigns with his body, and he believes with his mind and will or soul. More than a century earlier, in 1389, the self-identified ‘priest’ and accused Lollard William Ramsbury preached across the northwestern portion of the Salisbury diocese from Swanage to Malmesbury. According to his examination recorded in Bishop John Waltham’s register (3 July 1389), Ramsbury celebrated Mass wearing vestments and closely followed the form of the orthodox service, but he omitted specific prayers or words (“nichil dicendo”) assigned to the celebrant at different parts of the Mass. Ramsbury was also reported to have made the sign of the cross over the host and chalice on the altar, “nichil dicendo set labia movendo ac si diceret” [saying nothing but moving his lips as if he were speaking] (Salisbury register, fol. 222v; quoted in Hudson 1985: 118, 122. Translations are mine unless noted otherwise.). Whereas Boughton claimed to do one thing with his body but think another with his mind, Ramsbury was accused of two, somewhat different transgressions: falsely acting as a licensed priest and making a mess of the Mass. Regarding the latter, he was said to have deleted words at key moments in the liturgy in order to deceive others as to his actual belief regarding transubstantiation as official doctrine. Interestingly, Ramsbury’s reported linguistic behavior implies he believes that a priest’s speech at the altar is in fact efficacious. Otherwise, why not just pronounce the words of consecration just to keep the act going? The efficacy of ritual language when pronounced by the priest was the basis for the Donatists’ belief, declared heretical by the Church in 409 CE (cf. Rosier 2004: 263-352). Ramsbury’s speech and actions imply that he accepts the efficiacy of priestly language and wishes to speak sincerely at the altar according to his own religious beliefs. Without the appropriate intent and speech, Boughton and Ramsbury’s religious actions were interpreted by orthodox authorities as inauthentic, deceptive, and hypocritical. Wyclif’s claim that “sacramentum eucharistiae est in figura corpus Christi et sanguis” [the sacrament of the eucharist is the body and blood of Christ in figurative form] (Fasciculi Zizaniorum, 1858: 186) might have satisfied Olivi’s criterion for pragmatic sincerity, but it wouldn’t have satisfied the heresy investigators in Salisbury. The relation between intent and performance as the basis for meaning and truth made the pragmatics of speech from both the speaker’s and the listener’s point of view a critical question in heresy inquiries in the later Middle Ages. Sincerity in thought and word was a criterion for orthodoxy and belief as well as for communal norms for cooperation. In the Church’s official view,

How Here tics Talk, According to Bernard Gui and William Thorpe

169

heretics were by definition insincere or misleading and their speech was regularly critiqued and negatively evaluated in pragmatic ways. The Benedictine chronicler Hugo of Poitiers (d. 1167) wrote an account of suspected Deonarian or Publican heretics arrested in Vézelay in 1167, probably based on his involvement with their examination: “et adducti in quaestionem per ambages et circuitus verborum tentabant velare foedissimam sectam haeresis suae” [brought up for examination, they sought to conceal the utterly loathsome tenets of their heresy by evasions (ambages) and circumlocutions] (Historia Vizeliacensis monasterii; PL 194: 1681C; Heresies, 1: 248). The examiners evaluate the Publicans’ speech with the presupposition that sincere and honest people will answer truthfully their interrogators’ questions and cooperate with their examiners by using words with their conventional and straightforward meanings in order to communicate truthfully and without evasion. Since the examiners were already suspicious of certain individuals’ behavior, hence the inquiry or trial, a speaker’s perceived unwillingness to cooperate with the examiners’ questions or to answer questions directly only made that person more suspect. ‘If you have nothing to hide …’ We are familiar with this sort of interaction across centuries in asymmetric adversarial, especially legal situations. In theory, such inquiries may presuppose or require cooperation from all participants, but because the participants’ power relation is asymmetrical, investigators’ speech is situationally noncooperative. ‘Work with us’ is not the same as ‘Let’s work together.’ Since medieval trials or inquiries for heresy were inherently and structurally adversarial, examiners were alert for what they regarded as vague or evasive language, insincere responses, and manipulations of speech which falsely implied cooperation or sincerity. When they thought suspected individuals were speaking in those ways, the examiners were quick to believe the individuals were lying, dissembling, or possibly harboring heretical beliefs. Given such potentially face-threatening circumstances, we can begin to understand why many of those suspected of heterodoxy were not just trying to maintain pragmatic ‘face’ but were trying to preserve their freedom and perhaps their lives by using evasive strategies rather than directly confronting their accusers. Those coerced to confess usually made direct but tightly scripted statements about their heretical beliefs and behaviors. (For examples of formulaic records of judgments and outcomes in heresy trials, see Tanner 1977.) Before that, however, some of those suspected did speak directly, that is, baldly on-record, challenging their questioners and trying to gain discursive control of the hostile situations. Their speech and life narratives provide us with more complex evidence for medieval pragmatic awareness and metapragmatic understanding.

170 

The Medieval Life of L anguage

Bernard Gui’s Conversation Analysis and Institutional Discourse I turn now to two extensive accounts of heresy examinations: Bernard Gui’s Practica inquisitionis and William Thorpe’s Testimony. These and other contemporary accounts suggest that a key part of official practice was to expose the discursive strategies of suspected heretics through pragmatic and metapragmatic analysis. I compare Gui’s construction of heretic subjectivity and pragmatic analysis with the English Testimony of William Thorpe, composed eighty years after the Practica. Thorpe’s text is one of the few accounts of heresy inquiries written substantially from the subject position of the suspect. These texts present complex dialogic narratives of hostile talk and pragmatic interaction. Together, they present two sides of the dialogic pragmatic coin in asymmetric power contexts. Both foreground equivocatio, polysemy, intent, sincerity, and discursive power, but from very different subject positions. The inquisitor Bernard Gui (OP, 1261/62-1331) does not readily come to mind as a theorist or analyst of pragmatics. He was educated at several Dominican studia in southern France, published several historical chronicles, and compiled a life of Thomas Aquinas on behalf of the latter’s canonization. His handbook, Practica inquisitionis heretice pravitatis (The Inquisitor’s Handbook), completed in 1323 or early 1324 (Practica, ed. Mollat 1964: 1.xixii; Inquisitor’s Guide, ed. Shirley 2006: 15) turns out to be a rich resource for medieval pragmatic thinking. Gui uses both his own experience and accounts by other chroniclers to develop a strong pragmatic analysis of heterodox speech among southern French nonconformists, especially Waldensians, Beguines, and Cathars. From 1307-1323 the Dominican administrator Bernard Gui was the chief papal inquisitor in the Languedoc region. At that time, the area was still riven with heresy, despite the Albigensian Crusade against Catharism a century earlier. He composed his Practica inquisitionis heretice pravitatis, he said, to assist his successor, Pierre Brun, as chief inquisitor of Toulouse and also to help others discern and know (“discerni valeant et cognosci”) what heretics and suspected heretics are really up to (Practica, 5.2.4). Gui describes the beliefs and practices of various heretical sects, including Cathars, Beguines, and Waldensians, along with Jews and sorcerers. Gui’s text relies on his own observations and experience in various trials and inquiries and the reports of others involved with heterodox believers, especially Stephen of Bourbon (OP, c. 1180-1261) and David of Augsburg (OF, d. 1272). The text did not have a wide readership. However, Gui’s handbook illustrates how official religious culture constructed an authoritative view of nonconformists’

How Here tics Talk, According to Bernard Gui and William Thorpe

171

speech behaviors. Somewhat unintentionally, the Practica is a formidable, detailed account of late medieval nonconformist groups and the responses of the institutional Church to the religious and lay Other in their midst. For the history of medieval pragmatics, Gui’s Practica constitutes an unlikely but in-depth analysis of medieval dialogic discourse, primarily from the point of view of institutional power, but not entirely. Unlike some other antiheresy writers, Gui outlined a distinct pragmatic perspective on heretical, face-threatening, and evasive speech. The result constitutes, I believe, an implicit theory of pragmatic discourse. In Practica, Part Five, Gui discusses pragmatic strategies of deception and agency and constructs a ‘proper’ institutional listener and a paradigmatic heretical subject, both of whom are pragmatically adept and dialogically linked in a hostile interaction. Gui says his Practica should give an investigator the knowledge to be a skeptical, detail-oriented linguistically sophisticated interlocutor, alert to the speaker’s evasions and deviations from conventional and communal norms of meaning and implicatures. Discussing suspected heretics in both extra- and intra-community contexts, Gui represents ‘how heretics talk’ as a guide for official investigators. Knowing how the Other speaks and thinks is key to controlling the Other. Gui constructs heretic subjectivity not only in terms of the conflict between orthodox and heterodox beliefs but also in terms of different pragmatic strategies. He describes evasive speech, vague language, and hedged or modal syntactic forms as inherently insincere, but his vignettes and narratives also depict suspected heretics as pragmatically sophisticated and therefore doubly dangerous. Gui’s careful account of how heretics talk reveals his complex awareness of linguistic pragmatics as enacted by different participants in hostile situations. Like Bacon and Pseudo-Kilwardby and to some extent Olivi, Gui’s discussions of the vocabulary and grammar of spoken discourse emphasize the speaker’s intentio, voluntas, and concept formation. He highlights interior speech/concepts and affectus and contrasts acceptable and non-conforming performativities as criteria for examiners’ interpretations of meaning, judgments of acceptability and grammaticality, and discernment of sincerity, orthodoxy, and therefore cooperation with the institutional Church. Gui calibrates heretics’ speech with questions of speaker intention and responsibility, mostly in relation to the Waldensians living, worshipping, and preaching in southern France and northern Italy (on the Waldensians, see Cameron 2000). The Waldensian sect was initiated by the merchant Peter of Lyons around 1170. Followers adopted a lifestyle of neo-apostolic piety, preaching, and scriptural reading. Members held especially antisacerdotal views, advocated lay vernacular preaching, and refused to

172 

The Medieval Life of L anguage

recognize ecclesiastical courts as having jurisdiction over them in matters of religion and faith. In his account (partly based on Stephen of Bourbon and David of Augsburg), Gui outlines the declared beliefs and practices of the Waldensians and their deviance from the official faith. Along the way, he analyzes how Waldensians intentionally use evasive or ambiguous speech, words, and syntax. Gui claims they indirectly reveal rather than hide their heretical beliefs: “Et hoc est unum signum evidens per quod cognosci possunt esse heretici in duplicitatibus respondendo” [Responding with duplicity is one clear sign by which they are able to be recognized as heretics] (Practica, 5.2.8, ed. Mollat 1964: 1.72; Heresies, 1991: 400. Italicized words are borrowed from Stephen of Bourbon. I have slightly modified some of Wakefield and Evans’ [1991] translations of Gui’s handbook.) Gui says the Pseudo-Apostles also use evasive, insincere, and outright lying language (Practica, 5.3.5; ed. Mollat 1964: 1.96), but he doesn’t analyze their speech in much detail. Throughout his handbook, Gui refers to duplicitous and insincere speech as sophismata, duplicitas, perjuria, or equivocation. He uses some standard logical and grammatical metalanguage, especially fallacia, to describe heretics’ speech and performances in terms of truth and falsity. Like logicians, he uses the logical and semantic topic equivocatio in the sense of a sign of logical error requiring disambiguation and as a sign of pragmatic insincerity (cf. Peter of Spain, Summulae logicales, ed. de Rijk, 1972: 98-105). Gui’s use of equivocatio therefore differs from the way Bacon, Olivi, and other grammarians used the concept, but it does accord with how other interrogators of heterodoxy responded to equivocation, as William Thorpe’s narrative of his examination also suggests. Clearly, Gui regards equivocatio and insincere public speech as the Cathars’ and Waldensians’ characteristic strategy to disguise their dissent from orthodoxy and official ecclesiastical power. Gui amplifies the logical fallacy as a sign of spiritual error and heterodoxy. Interestingly, some inquisitors did not regard equivocatio and duplicitous speech as in themselves necessarily wrong, immoral, or heretical. The Catalan inquisitor Nicholas Eymerich (c. 1316-1399), for example, in his influential Directorium inquisitorum claimed that an inquisitor was justified using equivocal, ambiguous, or even lying speech with an accused heretic if the inquisitor’s goals were to get the accused to reveal the truth, bring the sinner to penance or conversion, and secure the community’s spiritual health. According to Eymerich, the moral rightness of speech fully depends on the intentions of the speaker. Aquinas mostly supported this institutionally sanctioned view of deceitful (dolus) or equivocal language on the part of examiners and those entrusted with the spiritual care of the community.

How Here tics Talk, According to Bernard Gui and William Thorpe

173

(Sullivan [2011: 169-196] summarizes Eymerich’s career and writing.) Medieval pragmatics, meaning, and sincerity are political. The efficacy of authoritative speech and the social construction of the ‘heretic’ derive from structures of institutional power. Heretic talk as it was perceived was constructed in pragmatic dialogism. When the Church appropriated logicians’ notions of equivocatio and ambiguous speech for its own purposes, the suspects’ use of similar strategies for their own, different ends became even more potent but also more suspect. Despite Gui’s overt hostility to religious nonconformists, we can read in his account of Waldensian discourse a grudging respect for heretics’ or suspected heretics’ pragmatic sophistication. It’s hard to tell just how self-aware Gui might have been with respect to his own ambivalent position when portraying how heretics talk and think. Waldensian equivocation and evasion, he and his sources claim, are part of the sect’s discursive pragmatics and interpolating discourse. More generally, he suggests, equivocation and deception are just how heretics talk in threatening, hostile, or out-group situations: Notandum est autem quod Valdenses sunt valde difficiles ad examinandum et inquirendum et ad habendum veritatem ab eis de erroribus suis propter fallacias et dupplicitates verborum quibus se contegunt in responsionibus suis ne deprehendantur; et ideo de eorum dupplicitatibus et fallaciis aliqua perstringenda sunt in hoc loco. [It should be noted, moreover, that it is very difficult to examine and question Waldenses and to get the truth from them about their errors. This is due to the tricks and double meanings of the words they use in their testimony, behind which they hide in order to prevent being entrapped. Therefore, some facts should here be set forth briefly to show their wiles and their deceptive use of words] (Practica, 5.2.7; ed. Mollat 1964: 1.64; Heresies 1991: 397).

Here, Gui implies that at least some lay Waldensians, whom the Church and Gui frequently disparaged as illiterati, simpliciores, and unsophisticated false believers, were in fact quite adept at using polysemy, vague language, and other pragmatic strategies to resist pressure from authorities to confess or reveal their heterodoxy. Both ‘unlearned’ and literate Waldensians, says Gui, try to conceal their true beliefs from religious inquisitors. But in doing so, they reveal themselves to be not only heretics, because they are speaking insincerely, but also master manipulators who exploit linguistic polysemy and ambiguity to frustrate their more learned and supposedly

174 

The Medieval Life of L anguage

more rhetorically skilled examiners. Gui’s account often portrays heretics as more than competent challengers to the Church’s investigators; hence the need to alert examiners about the evasive and other pragmatic strategies suspected heretics use. Suspect speakers whose utterances were perceived to be vague were understood to be acting uncooperatively or insincerely. Their evasive utterances and strategies, while sophisticated, are evidence of their guilt as much as their stated beliefs, because they challenge the dominant cooperative principle in medieval pragmatic hegemony, that shared goals, quality replies, and relevant information are the default norms of sincere conversation exchange. But as we have seen repeatedly, trust or good faith is both something which participants may or may not presume in an interaction and a relational condition which is ongoing, something constructed, maintained, or unravelled in discursive dialogism. What speakers intend or imply isn’t necessarily what they say and not necessarily evasive either. It depends. Nor is implied meaning necessarily what listeners receive or want to hear or think they hear. In many speech exchanges, trust, like truth, is always already on a knife edge. Gui described certain kinds of speech formulas and styles as markers of heretical discourse, in particular within the Beguine and Waldensian communities. The Beguines (sometimes called ‘Poor Brethren’) were a group of pious priests, laymen, and laywomen which emerged around 1300 in southern France and in the Low Countries. Gui’s account refers specifically to the Beguines in Provence, Narbonne, and Toulouse. He claims they adopted a devotion to apostolic poverty, guided by the writings of Olivi and the third order of St. Francis (the reformist Spirituals) (Practica, 5.4.1-2; ed. Mollat 1964: 1.108-114). Gui describes the Beguines’ patterns of speech and actions, which distinguished them from other people, as intentional and antagonistic because they wished to form a nonconformist subgroup and avoid public and official scrutiny. Both Gui and the Beguines themselves acknowledge their alternative discursive register as a practice which actualizes their subgroup identity, but for Gui that usage marks their deviant practice and dissent from Church authority. For example, Gui claims the Beguines and the Waldensians use certain kinds of address as passwords or code to recognize one another and initiate interaction (conversation openings). When a Beguine approaches or enters a house, Gui claims, or meets another Beguine on a journey or on the street, they say, “Benedictus sit Jhesus Christus” [Blessed be Jesus Christ] or “Benedictum sit nomen Domini Jhesus Christi” [Blessed be the name of Lord Jesus Christ] (Practica, 5.4.4; ed. Mollat 1964: 1.118. My translation.). It’s not clear why Gui thinks such greetings are problematic. The conversation opening

How Here tics Talk, According to Bernard Gui and William Thorpe

175

echoes Jesus’ instructions to the apostles to preach and when entering a house to say, “Pax huic domui” (Matthew 10:12; Luke 10:5). The expression became the standard Franciscan blessing when entering a home. The ingroup practice would fit with Beguine and Waldensians’ neo-apostolic piety and scriptural modeling. Gui’s Latin presumably translates the vernacular speech he is referring to. Perhaps what he objects to is how lay people have usurped the clergy’s right to bestow blessings or perhaps how they have appropriated the blessing as a way of signifying their otherness rather than solidarity with the official Church. Gui also describes how Beguines use distinctive body movements and recite certain Latin prayers at meals and worship to maintain in-group identity. He widens his description of nonconforming subgroups by similarly claiming that the older sect of Waldensians, also devoted to neo-apostolic poverty, greet one another with coded language and gestures and adopt certain gestures and prayers centered around the ritual meal (Practica, 5.2.5; ed. Mollat 1964: 1.48-58). Beguines’ greetings and Waldensians’ prayers, he says, are heterodox not because of their syntax or vocabulary – probably in the vernacular but reported by Gui in Latin – but because of who is speaking them, that is, nonconforming or suspect lay people. For Gui, Waldensians and Beguines are exemplars of the ‘heretic’ type and identifiable by their ways of speaking, that is, their unauthorized religious discourse, reading and reciting scripture in the vernacular and using equivocatio, evasiveness, ambiguity, deceptive syntax, and dissenting gestural pragmatics. Gui’s handbook engages in what Ian Hacking (2006) calls the practice of “making people up.” Hacking describes this “style of reasoning” as “dynamic nominalism” (2002). Our notions of types of people are social constructions based on the dissemination of knowledge/power, often with the aim to control dominant hegemony and social order. Gui’s point is that those who identify with Beguine or Waldensian beliefs and communities will use evasive syntax and exploit polysemic, vague, or ambiguous meanings to hide in public and avoid directly and sincerely answering their accusers. At the same time, their use of coded greetings and key terms gives implicit or private notice to likeminded others. They perform their otherness in speech strategies as part of their subgroup identity. In order to effectively discern heresy, Gui implies, the clergy need to understand how specific speech and gestures create communities of others, not just the community of the one true Church. In the thirteenth and fourteenth centuries, secrecy and community-building are hardly surprising behaviors for neo-apostolic reformist groups hunted and persecuted by the official Church. But Gui’s examples suggest how subculture pragmatics

176 

The Medieval Life of L anguage

were already a category of knowledge in authoritative, official discourse. Nonconforming people were called out and marked; their speech, gestures, even their dress were interpreted as pragmatic signifiers of dissent and disorder. Waldensians were sometimes referred to as “sandaliates” or “sabotati” (“sabbatati”) because of their footware in imitation of the first apostles (Practica, 5.2.4; ed. Mollat 1964: 1.38; cf. Amsler 2012: 235-236; Hoose 2016). Margery Kempe was criticized by clerics for wearing white clothing, a practice which was supposed to be reserved for “chast womyn.” In the eyes of the official Church, such behaviors were insincere “imitationes apostolorum.” But others might understand that speech and behavior differently, as a better spirituality. Gui’s manual for discerning heresy provides other materials which reflect various speech practices as well as some medieval official attitudes toward so-called deviant or unorthodox speech practices. The text includes reported speech to characterize heretical groups and purports to reproduce the words, syntax, and conventions which suspected heretics actually use when questioned by Church authorities. He gives special attention to equivocation, ambiguity, and evasive speech. Like most everyone, from Bacon, Aquinas, and Olivi to Kempe, merchants, and poets, Gui recognized that utterances can encode more than one meaning or express meaning implicitly as well as directly. For Gui, the true meaning of heretics’ speech, not just what they superficially say, depends on their intentions as inferred by listeners’ interpretations and presuppositions. Thus, in-group listeners are likely to interpret a suspected heretic’s utterances differently from how an inquisitor would. Gui flexibly notes this as another form of equivocatio and deceptive speech. Accordingly, he explicitly interprets a suspected heretic’s reply to an interrogator’s question on doctrine as deceptive because the speaker’s superficial agreement, he claims, is undermined by the speaker’s hidden and contrary warrant for the utterance. The problem is not what they say but what they mean: “Respondent bene et firmiter: Credo, et intelligunt et exponunt, id est: Bonam et firmam credentiam habeo de pertinentibus ad sectam meam” [They reply well and firmly, “I so believe,” but they understand and mean [something else], that is, “I have a true and firm belief in such of these doctrines as are held by my sect”] (Practica, 5.2.8; ed. Mollat 1964: 1.72; Heresies 1991: 401). For Gui, the central questions are: How are accused heretics using ambiguous and equivocal language and deceptive performances? What are they pretending to say, and what do they really think? How should people in the wider (orthodox) speech community respond? Although Gui’s narratives of reported speech afford heretics or suspected heretics a great deal of pragmatic sophistication, he still finds it disturbing

How Here tics Talk, According to Bernard Gui and William Thorpe

177

that suspects can successfully achieve some agency and hide their true beliefs while reveling in their linguist cleverness and evasive performance. He detests their palliationes verborum [verbal screens; lit., ‘cloaks of words’] and duplicitates verborum [verbal duplicities] (Practica, 5.2.7, 5.2.8; ed. Mollat 1964: 1.66, 72). I suspect Gui found it particularly irksome for a suspect to be brought before him and “Interrogatus si scit quare sit captus, respondet valde mansuete et subridendo: ‘Domine, libenter discerem a vobis causam’” [When asked if he knows why he has been arrested, he replies quite calmly with a smile, “Sir, I should be glad to learn the reason from you”] (Practica, 5.2.7; ed. Mollat 1964: 1.65; Heresies 1991: 397. Italicized words are borrowed from David of Augsburg.) Gui’s narration adds the smirk on the accused’s face to suggest the person is proud of their discursive skills. Gui is also frustrated by heretics claiming they are too simple and unlearned to understand his questions: “… simplex homo sum et illiteratus, nolite me capere verbis” [“I am a simple man, without learning; pray do not trap me in my words”] (Practica, 5.2.7; ed. Mollat 1964: 1.66; Heresies 1991: 398.). Reading Gui’s account within a pragmatics and metapragmatic awareness frame, we see he clearly doesn’t think most accused heretics are illiterate or crude language users at all. Rather, he explicitly acknowledges how suspected heretics and dissenters pose special challenges for heresy examiners precisely because of their linguistic and pragmatic skills. Beguines and Waldensians, he claims, are intentionally evasive and strategic in their speech when confronted by the interrogators. Elsewhere, Gui cites their use of indirect assertions and then evaluates such speech strategies as dangerous ambiguity and a sign of heretical belief: “unum signum evidens per quod cognosci possunt esse heretici … per verborum equivocationem” [one plain sign by which heretics can be recognized [is] … equivocation] (Practica, 5.2.8; ed. Mollat 1964: 1.72; Heresies 1991: 400). Gui’s pragmatic analysis links logic with moral judgment; his syntax frequently connects equivocatio with duplicitas and sophismata (e.g., “respondent in equivocato et in duplicitate”; Practica, 5.2.8). Gui’s point of view and that of most Church examiners is that heretical views are an implicature of equivocal and duplicitous speech. By exploiting words’ and constructions’ polysemy and equivocal potential, suspected heretics reveal not only their nonconformity, as Gui and others claimed, but also the inherent instability of linguistic representation. Combining his own, Stephen of Bourbon’s, and David of Augsburg’s observations, Gui describes Waldensian equivocation and semantic slipperiness as part of the sect’s discursive pragmatics. His narrative vignettes frequently focus on heretics’ strategic uses of vague language, lexical and syntactic. For

178 

The Medieval Life of L anguage

instance, Waldensians resist questioning by offloading their assent to what the interrogator asserts: Interrogatus de fide quam tenet et credit, respondet: ‟Omnia credo qie debet credere bonus christianus.” – Examinatus vero quem reputat bonum christianum, respondet: ‟Illum qui credit sicut sancta Ecclesia docet credere et tenere.” – Interrogatus qua vocat sanctam Ecclesiam, respondet: ‟Domine, quam dicitis et creditis vos esse sanctam Ecclesiam.” Si dicitur ei: ‟Ego credo sanctam Ecclesiam esse Romanam ecclesiam, cui presidet dominus papa et alii prelati sub eo,” et ipse respondet: ‟Et ego credo,” intelligendo quod ipse credit me ita credere. [Asked about his faith and belief, he [suspected heretic] responds, ‟I believe everything a good Christian believes.” Asked what he means by “good Christian,” he replies, ‟Someone who believes what Holy Church teaches and believes.” When asked what he calls “Holy Church,” he replies, ‟Sir, what you say and believe to be the Holy Church.” If he is told, ‟I believe the Holy Church to be the Roman Church, over which presides our lord pope and other prelates subordinate to him,” then he responds, ‟That I do believe,” meaning that he believes that I believe this …] (Practica, 5.2.7; ed. Mollat 1964: 1.64, Heresies, 1991: 397-98. Italicized words are borrowed from Stephen of Bourbon. Reported speech in 5.2.7-8 are mostly Gui’s own examples.)

Gui analyzes this and similar exchanges as examples of ‘pronoun trouble’ and how heretics evade direct answers by manipulating pronominal reference and context. Gui, grafting Stephen of Bourbon’s text onto his own, analyzes how Waldensians evade their investigators’ questions by using conditional clauses or phrases to recontextualize their speech and enable them to intentionally assert their own beliefs while not affirming those of the official Church. For example, says Gui, “Item est aliud genus sophisticandi adjectionem, ut, cum queritur: ‘Credis hoc vel illud?’, respondent: ‘Si Deo placet, ego bene credo hoc vel illud,’ intelligentes quod Deo non placet quod ipsi credant.” [This is another kind of sophisticating addition [to the discourse]: so when it is asked, “Do you believe this or that?” they reply, “If it please God, I truly believe this or that,” knowing that it does not please God that they believe it themselves.] (Practica, 5.2.8; ed. Mollat 1964: 1.72,74. My translation. Italicized words are borrowed from Stephen of Bourbon.) Gui’s typology of heretics’ pragmatic strategies for evasion and deception also includes answering a question with a question: “cum queritur ab aliquo si ita credat, respondet cum admiratione,

How Here tics Talk, According to Bernard Gui and William Thorpe

179

quasi iratus: ‘Quid crederem ego aliud, numquid debeo ego ita credere?’” [Upon being asked if he believes so and so, the suspect replies in surprise, as if indignant, “What else should I believe? Is this not what I ought to believe?”] (Practica, 5.2.8; ed. Mollat 1964: 1.74; Heresies 1991: 401. Italicized words are borrowed from Stephen of Bourbon.). Gui’s account represents heretics as boldly uncooperative and untruthful because in conversation, despite the hostile situation, they seize the floor and coopt the inquisitor’s dominant role as questioner seeking confirming answers from the suspect. Returning questions to questions is a strategy of evasion, misdirection, discursive struggle, and, says Gui, a sure sign of heresy. Elsewhere, Gui describes how Waldensians resist cooperating with their interrogators by confounding the protocols for intentional meaningful speech. For instance: Tunc ille contremiscens et quais qui nesciat eadem verba formare, cespitabit in eis ut vel ipse vel alius interloquatur et verba aliqua interponat ne fiat directa forma jurandi, set quedam locutio non juratoria, ut tamen ab aliis jurasse putetur. [Trembling [when taking an oath] and acting as if he cannot pronounce the words, he will falter repeatedly in saying them, so that either he or someone else will interrupt and interpose some words, with the result that a straightforward form of oath is not taken but rather a certain jumble of words, which is not actually an oath ( juratoria; lit., juratory) but which gives others the impression that he has taken an oath] (Practica, 5.2.7; ed. Mollat 1964: 1.68; Heresies, 1991: 399. Italicized words are borrowed from David of Augsburg.)

By interrupting or manipulating the “felicity conditions” for oath taking, says Gui, the Waldensian’s performance effectively evades the investigator’s questions. Because canon law presumes that an oath sincerely taken is binding, the Church regards the suspected heretic as not bound by the words of the oath he or she has falsely expressed. Waldensians’ other performative pragmatic strategies of deception include displaying surprise, returning the questioner’s assertion as an embedded utterance (e.g., “I believe that you believe that – ”), switching topic, and pleading illiteracy (illiteratus, simplex), that is, cognitive incompetence (Practica, 5.2.8; ed. Mollat 1964: 1.74; Heresies 1991: 400-402). Gui regards heretics’ speech as a discursive performance in the worst way possible: deceptive, manipulative, insincere, arrogant. Karen Sullivan also reads Gui’s narrative as a critique of heretics’ deceptive performances.

180 

The Medieval Life of L anguage

She says Gui believes the most important goal of any heresy examination is to maintain the power and position of the papal office of the inquisition (2011: 124-145), not to uncover the truth or correct the suspect’s beliefs. Maintaining the Church’s authority is an important aspect of Gui’s text, especially in Practica, Part Five, but I also think Gui’s, Stephen’s, and David’s analyses of interaction narratives stand out for their explicit attention to the details of linguistic and pragmatic interaction. Repeatedly, it is the actual talk which draws Gui’s attention: “Multos autem modos alios decipiendi habent quod magis docet usus quam ars” [They have a great number of other methods for deceiving [us], which one learns better from practice (usus) than from theory (ars)] (Practica, 5.2.8; ed. Mollat 1964: 1.74; Heresies 1991: 401). His observations lead him to criticize heretics’ pragmatic strategies for topic shift, misdirection, and syntactic embedding because they reaccent examiners’ questions, control the conversation floor, and potentially subvert the interrogation (Practica, 5.2.8). That is, he criticizes the suspected heretics for threatening the Church’s ‘face.’ Gui’s narratives of Waldensian and Cathar evasion and equivocation are filled with metapragmatic awareness and detailed pragmatic analyses of discourse in adversarial circumstances. When he adds details to David and Stephen’s accounts, he highlights the suspects’ performative practices (“respondet valde mansuete et subridendo”) and emphasizes the accuseds’ insincerity and deception. His examples consistently support the claim that Waldensian speech in conversations with Church authorities asking questions is all an act, a pragmatic theater of evasion, equivocation, and deceit. Heresy. From a critical pragmatics perspective, we can say Gui’s accounts of the exchanges between examiner and suspect deploy institutional power to describe and analyze the studied resistance of nonconformists. Because he interprets heretics as sophisticated manipulators of discourse, he encourages investigators to engage in a performance of their own, one which sustains the Church’s authority and the office of the inquisition rather than primarily seeking to discover sin and impose penance for the salvation of souls. Thus, Gui reveals within the official Church’s own discourse the politics of medieval pragmatics and the conflict over whose discursive performance is acceptable and whose isn’t. The irony is that by rehearsing the heretics’ speech, Gui ends up acknowledging their pragmatic sophistication and metapragmatic awareness. The dialogues presented in Practica, especially those in 5.2.7-5.2.8, portray how some Waldensians intentionally exploit polysemy and equivocation and manipulate syntax and modality in their exchanges with investigators seemingly in order to cooperate with their interlocutors and produce conversation

How Here tics Talk, According to Bernard Gui and William Thorpe

181

cohesion and coherent clauses. But, Gui says, their speech often refers not to doctrine (external substantive context) or beliefs but to other interlocutors’ speech (external discursive context). It is evasive speech. Heretics, he implies, are sophisticated clause embedders and pronoun manipulators. They recontextualize investigators’ language and say intentionally only what they mean to say, hoping the investigator will think otherwise. Such dialogic strategies were not restricted to southern France or the thirteenth century. In fifteenth-century England, William Thorpe, by his own account, also repeatedly qualifies his terms or syntactic constructions with semantic recontextualizations so as to make his utterances ‘truthful’ in his own eyes, while he hedges or avoids direct agreement with Archbishop Arundel. Margery Kempe uses echo responses, syntactic embedding, and strategic vague language to similar effect. Gui’s narratives and pragmatic analyses render the heretic speaker a pragmatically effective, dangerous interlocutor whose time is up because, thanks to Gui’s handbook, inquisitors are now on to their linguistic strategies and tricks. Building on practice, Gui mounts a theory of pragmatic competence, however disruptive or misused, and a theory of linguistic (in)sincerity. Both motivate Gui’s account of heretic subjectivity, but he acknowledges that different speakers will read certain pragmatic strategies positively or negatively depending on the standpoint of the examiner or suspect in different contexts. I have mentioned how Gui’s analyses of heretic discourse draw on some of the same philosophical and semiotic concepts as do the pragmatic theorizing of Bacon, Olivi, Pseudo-Kilwardby, and the anonymous commentators on interjections we discussed in Chapter 2. However, Gui implies intentionality is more a problem than a criterion for meaning, unlike the thirteenth-century intentionalist grammarians. Intention and sincere utterance establish the relation between heart and mouth, a key trope in medieval religious writing, courtly romance, and lyric. The trope was also prominent in discussions of oath-taking and swearing, two contested kinds of speech acts in suspected heresy encounters. (As utterances and speech acts, swearing as insult and blasphemy is different from swearing as promising or pledging, with different pragmatic implications.) Gui’s pragmatic analyses contribute to his broader construction in the Practica of the religious and social Other as a dissenting subject. His accounts of Beguine and Waldensian pragmatics, evasive speech, and syntactic manipulation become a kind of pragmatics theory, if not quite a critical pragmatics, in that he foregrounds their discourse strategies as conventions of dissent, that is, as part of an alternative discursive practice and ideology.

182 

The Medieval Life of L anguage

In the twelfth century, Walter Map (c.1140-c.1209) had ridiculed Waldensians’ intellectual and discursive incompetence (De nugis curialium, 1.31). A century later, Gui represents heretic speech as a counter practice (alt-usus) or alternate hegemony. Despite his hostility, he acknowledges heretics’ subtle use of discursive strategies to manipulate or pragmatically gain control of their interactions with Church officials while also maintaining their belief integrity and avoiding complying or cooperating with examiners’ questions. As an official account, Gui’s Practica indirectly manufactures Waldensians and other heretics as auctores, textual authorities, by reproducing their discourse and discursive practices as Latin text, even if his primary goal in writing the text was to equip investigators and confessors with the metacognitive and metapragmatic awareness they need to infer or discern what heretics are up to and what they really mean when they respond to authorities’ inquiries. The Practica is a polysemous archival text. It fixed and disseminated prototypical signifiers of heterodox discourse from the Church’s point of view and located heterodox speech strategies within a broad set of pragmatic tropes, conventions, and rhetorical maneuvers which assessed those speech strategies and performances on a continuum of good/bad, cooperative/disruptive, orthodox/heterodox, sincere/insincere. Gui’s construction of heretic subjectivity situates vague, evasive equivocal speech, polysemy, embedding syntax, and empty gestures as indirect signifiers, even idioms, of heretical beliefs. He and his sources claim that the pragmatic strategies they identify are the sort of linguistic behaviors used by those with something to hide from Church authorities rather than as general discourse strategies. If we know anything about pragmatics and ordered talk, Gui was wrong to restrict his analyses only to suspected heretics, but of course that was his motive for the analysis in the first place. He regarded the sophisticated pragmatic and metapragmatic strategies of heretics as exceptional precisely because he considered them to be inauthentic and insincere speakers. In Practica, Part Five, Gui goes on to imply that many religious dissenters were dangerous to the faith and the community not only because of their professed beliefs but also because they were armed with sophisticated linguistic survival strategies which prepared them to engage effectively in semi-public disputations with authorities. Church authorities need to be ready with their own metapragmatic awareness, critical skills, and conversation strategies. Gui’s narratives and analysis represent suspected heretics not as staging a contest between institutional norms and individual conscience but as an alternate hegemony comprised of religious dissenters who refuse to cooperate with institutional authority and abide by the conventional meanings

How Here tics Talk, According to Bernard Gui and William Thorpe

183

and presuppositions of the communi intentiones. In the later Middle Ages, investigators such as Gui were keen to systematize and expose the pragmatic strategies nonconformists and disruptive reformers used to support their beliefs and challenge Church authority. The dialogism of Gui’s text reveals the network of subject formation in heresy discourse. Heretic subjectivities were constructed by the imbrication of structuring structures established by the Church and micro maneuvers of pragmatic and metapragmatic understanding. As Gui’s Practica makes clear, ecclesiastical authorities used pragmatic analysis to intentionally make more visible the social type heretic as a certain kind of dissenting or nonconforming subject, while those religious Others were also fashioning their own resistant subjectivities.

William Thorpe’s Relationship Pragmatics Medieval understanding of how heretics talk, like any understanding grounded in metapragmatics, emerges from relationships. In Bernard Gui’s case, understanding the discourse practices of the religious Other meant observing and imagining what it was to speak from the heretic’s subject position and then to critically account for the context, presuppositions, and beliefs manifested or implied in those language strategies. As Clare Monagle succinctly formulates this institutional dialogism, “The demarcation of heresy was necessary to the making of Christian orthodoxy. The heretic in fact enabled the definition of the faithful” (2017: 40). The dialogism of discourse is reproduced at the level of social subjectivity. From a cultural semiotics perspective, dialogism also complicates Moore’s (1987) claim that dissenters and heretics never constituted an ‘order’ in opposition to the official Church (except maybe the Cathars) and that the persecution of heretics, lepers, and Jews was a deliberate effort by the clergy to reinforce their authority. True, lay people gathering together for private worship and Bible study and supporting one another in their beliefs and practices apart from the official Church were almost never recognized as legitimate religious groups. Heresy examiners used the term “religio privata” to identify such practices. By 1300 the official Church was fully mobilized against what were regarded as dangerous, deviant subgroups which needed to be erased or separated from society at large. The heretic subject is at once a constructed Other and a social agent. When dissenters positioned themselves as following apostolic faith or the ‘truth of scripture’ rather than orthodox doctrine and legitimate Christian priests, their alternate hegemony motivated the Church and its agents to adopt new strategies to address dissent. One

184 

The Medieval Life of L anguage

strategy was to attack how heretics speak and to advocate for ‘orthodox’ speech. Discursive pragmatics were part of that strategy. Subjectivities are relational, not individual or strictly autonomous. In their critiques and antagonisms, heretics and dissenters proposed alternate hegemonies to that of the established Church. Most accounts of dissenting or heretical speech and action come from the official Church’s point of view. But not all of them. The English text known as The Testimony of William Thorpe (Oxford, Bodleian Library, MS Rawlinson C.208, ff. 51v-57r) presents us with an extraordinary narrative account of how heretics talk from the point of view of a dissenting, Oxford-educated Wycliffite priest. Thorpe’s first-person narrative describes his ‘interview’ with Thomas Arundel, Archbishop of Canterbury, in 1407. The text, actually a letter probably smuggled out when Thorpe was in prison, displays a keen understanding of conversation pragmatics and metapragmatic thinking in the context of face-threatening, potentially life-threatening, talk. Like Gui’s narratives, heresy trial documents sometimes record the speech of the accused beyond the formulas for abjuration or confession. But Thorpe’s Testimony gives us a richer contextualized account, a highly involved narrative and metanarrative of what the men said (according to Thorpe) and, just as importantly, how they said it. In this section I read Thorpe’s text as both a narrative account and as a rehearsal for dissent, a kind of discursive handbook for English dissenters. The text presents the sorts of cognitive arguments and pragmatic maneuvers which Thorpe thinks Lollard dissenters need to be familiar with in order to preach their faith and avoid punishment or death at the hands of the official Church. In short, the Testimony reverses the motives which prompted Gui’s Practica. For all the debate about the text’s authenticity as an account of interrogation, Thorpe’s first-person narrative is thoroughly focused on his discursive position in a hostile dialogic relation with the Archbishop. Full of angry language and threats of physical violence, the text recounts a difficult interaction in which the institutionally ‘superior’ authority more than meets his match in the ‘inferior’ dissenter. That Thorpe is a priest and a Lollard doesn’t change the asymmetrical power relation much. Pragmatically, however, the narrative action is more complicated than that. Thorpe is a shadowy figure. There is little information about him and his examination beyond the Testimony. According to the text, Thorpe was questioned by Archbishop Arundel on the afternoon of 7 August 1407 at Saltwood Castle (Kent). Thorpe probably composed his narrative letter in English while still in prison, perhaps in 1407 or early 1408. Somehow, the text became available outside Saltwood. In addition to the English

How Here tics Talk, According to Bernard Gui and William Thorpe

185

version, two Latin manuscript versions (Prague, Metropolitan Chapter Library MS O.29 [c. 1430]); Vienna, Österreichische Nationalbibliothek MS 3936 [c. 1420]) and an early printed text (STC 24045, Antwerp 1529-1530) survive. The authenticity of Thorpe’s text and even Thorpe’s identity have sometimes been challenged. The Archbishop’s register does not record that an examination ever took place. However, the scholarly consensus, based on other ecclesiastical records, is that the exchange between Thorpe and Arundel did take place and that Thorpe did write the account (Copeland 2001: 191-219; Jurowski 2002). Thorpe’s explanation for why he composed the letter lends some support to the idea the text constitutes a kind of informal handbook for dissenting discourse. Thorpe’s narrative reenacts rather than simply transcribes his interaction and dialogue with Arundel. His account reflects dialogic understanding in that it presents both Thorpe’s and Arundel’s speech within a general, narrativized frame of contest and struggle. Thorpe’s narrative doesn’t present a search for communion or agreement between interlocutors. Rather, it strongly connects with what Bakhtin describes as the “dialogic idea,” a text “inter-individual and inter-subjective,” “a live event, played out at the point of dialogic consciousnesses,” which “wants to be heard, understood, and ‘answered’ by other voices from other positions” (1994: 98). Thorpe’s dialogic narrative presents the encounter from more than one point of view or subject position, however self-interested the narrative might be. Like Gui, Thorpe represents his and Arundel’s standpoints with detailed conversation exchanges and his own narratorial position which inflects and reaccents both participants’ speech. The text concretely portrays not only doctrine but also both speakers’ pragmatic, affective, and cognitive strategies and metapragmatic understandings. The text demonstrates the pragmatic strategies dissenters and reformers can and do use to defend their positions when involved in difficult conflict talk with Church authorities. The Testimony is organized in two parts: a prologue and a narrative of the 7 August 1407 afternoon exchange between Thorpe and Arundel. In the prologue (ll.1-164. All citations to Thorpe’s Testimony are from Two Wycliffite Texts, edited by Anne Hudson [1993]), Thorpe explains he wrote his narrative at the request of friends who visited him in prison before his examination. They asked that he “bisie me wiþ alle my wittis to go as nyȝ þe sentence and þe wordis as I can, boþe þat weren þere spoken to me and þat I spak, enaunter [perchance] þis my writynge come ony tyme bifore [þe] Erchebischop” (ll. 36-39). They wanted him to write an account of his examination “as nyȝ þe sentence and þe wordis as I can” so fellow religious dissenters in similar conflict situations might benefit from his answers to

186 

The Medieval Life of L anguage

the “opposynge.” The request is for an accurate and detailed account. The phrase “as nyȝ þe sentence and þe wordis as I can” also resonates with the description of the translation strategy of the Wycliffite English Bible. The text, says Thorpe, will help others be more informed as to Lollards’ idea of Christian “truþe,” give them the capacity to better maintain and declare their faith in the face of hostile questioning, and “wiþouten feynyng oblischen hemsilf wilfulli and gladli aftir her kunnyng and her powere to suen Crist pacientli” (ll. 110-112). Discerning sincerity and “feynyng” are central to Thorpe’s narrative, as they were to Gui’s. The narrative section of the text then divides roughly into two parts: an account of Thorpe’s declaration of his beliefs and the conflict over oathtaking (ll. 165-616) and a longer account of Arundel’s questioning of Thorpe on five charges lodged by the Shrewsbury elders alleging Thorpe’s heretical preaching at St. Chad’s church in April 1407 (ll. 617-2255). The narrative section concludes with the Archbishop rendering no judgment and Thorpe being returned to Saltwood prison, where he sits or stands alone in his cell, thanking God for keeping him safe “boþe amonge þe flateryngis specialli, also amonge þe manassingis of myn adversaries [þat] wiþouten hevenynesse and agrigginge of my conscience I passid awei fro hem” (ll. 2242-2245). The text ends with Thorpe encouraging his indeterminate readers and listeners “to ben oonyd in trewe feiþ, in stidfast hope and in parfiȝt charite. Amen, amen, amen” (ll. 2254-2255). Thorpe’s narrative is exemplary on several levels. First, he explicitly declares his intention to better inform Lollards as to how to explain and justify their beliefs in situations where Lollards might preach in public or have to answer authorities’ questions about their faith and behavior. Also, we can read Thorpe’s narrative as an allegory of the conflict between the institutional Church, represented by Arundel, and the followers of reformist or apostolic piety, represented by Thorpe. Historically, the text reflects the crisis, especially among Lollards, caused by Arundel’s renewed efforts to hunt down heretics and suppress lay religious individualism. Having played a key role in Henry IV’s usurpation of the throne from Richard II (1399), Arundel was made Archbishop of Canterbury and thus head of the English Church in 1399 and then reinstalled as Lord Chancellor in 1407. In 1407 Arundel convened a synod at Oxford, which passed regulations for legitimate preaching, prohibited lay people from preaching or possessing even a translation or copy of the Bible, and stipulated what theology could be taught in the schools (cf. Watson 1995). Although the full force of the Constitutions didn’t really come into effect until after 1408, Arundel at the time of Thorpe’s examination had already successfully engineered the recantations of several

How Here tics Talk, According to Bernard Gui and William Thorpe

187

Lollard leaders, including John Purvey, Nicholas Hereford, Robert Bowland, and Philip Repyngdon. Repyngdon was rewarded by being made Bishop of Lincoln in 1405. These men had been associated with Wyclif and Oxford reformers in the 1380s, and the first two were definitely associated with the Wycliffite English Bible translation project. But for one reason or another, after 1400 they publicly abandoned their earlier nonconformist views. As a result, as Copeland argues, Thorpe’s text positions him as a remainder, an historical figure of Wycliffism’s foundational moment still standing in the persecuted present (Copeland 1991: 191-219). Apparently, Thorpe had been on Arundel’s nonconformity radar for twenty years, wilfully, according to the archbishop, poisoning northlanders with “fals doctryne” (ll. 180-185). Existing anti-Lollard regulations about preaching and Arundel’s renewed attacks on Lollards and other nonconformists probably made Thorpe’s activities as a Lollard preacher even more conspicuous among the Shrewsbury elders. In 1407, after he had preached a controversial sermon at St. Chad’s, the town elders forwarded their charges against Thorpe to Arundel (Jurkowski 2002). Unlike medieval trial records, Thorpe’s dialogic text, with its embedded reported and direct speech and external and internal narration, concretely and extensively represents the to-ing and fro-ing of official questioning and answering in a hostile, face-threatening situation. In doing so, the text reveals the participants’ use of several pragmatic strategies for asserting agency and for negotiating, if not cooperating, in talk. The text also implicates attitudes toward various pragmatic assumptions and actions in discursive interaction. Thorpe’s description of his examination as “apposynge and answerynge” draws from medieval grammar school pedagogy and university disputation. By many accounts, the oral exchanges in schools were inherently hostile, sometimes violent, as the authoritative teacher questioned the student about Latin grammar based on a read text. These classroom interactions were defined by asymmetric power relations between the questioner with the power to punish and the student who needed to come up with the correct conjugation or face the consequences. The narrative situation in Thorpe’s text is similarly laced with coercive power and assertive authority. However, Thorpe as a qualified respondent adopts a stronger agentive stance and discourse style. The first-person narrative establishes the autobiographical split subject, the narrating I and the narrated I. Both Thorpe’s narrating and narrated voices seem to embody self-conscious metapragmatic awareness of speech and discourse strategies. The narrated Thorpe adopts a distinct Other-subject position and responds to Arundel’s coercion and assertive questioning with articulations of an alternate set of beliefs and reasoning delivered with steadfast

188 

The Medieval Life of L anguage

resistance and fortified by various pragmatic and conversation strategies for controlling the discourse, gaining the floor, and determining topics. Thorpe’s self-representation transcends Gui’s ‘heretic’ type and enacts a more positive combination of assertive and performative utterances, where reason, style, and belief are equal parts of his argument (see Bell 1984 for a general theory of discourse style and performance). The narrating Thorpe sometimes emerges explicitly in the text with evaluations or affective responses. Like Arundel, some modern readers of the Testimony describe Thorpe as ‘evasive’ and ‘equivocating’ in his responses to Arundel’s questions, and it’s not always clear whether those readers think that’s a good thing or a bad thing (e.g., Somerset 2002). Thorpe is evasive, but not quite in the way some readers might mean. It’s more adequate to the narrative and discourse pragmatics to think of Thorpe and Arundel as struggling for discursive footing and frame. Goffman (1981) argues that conversations are structured around footing and frame – stance, presupposition, and context – which provide participants in specific situations with normative expectations and strategies for conducting their interactions in socially acceptable ways. Frames make up the organizational principles and interaction expectations for establishing and sustaining our situated experiences with others. But frames and interaction expectations also give direction and shape to alternative discourses or resistance to dominant assumptions about how talk is supposed to go and the power relations which support those assumptions. Frames therefore posit the norms and expectations for appropriate interactive talk and establish what counts as violations of those expectations and norms. Norms and violations are defined in relation to one another. Often, but not always, discursive frames are set by those empowered to do so. In Thorpe’s narrative, the genre frame turns out to be a Russian doll, a nested series of letter > narrative > informal examination > two-handed dialogue. More to the point, Thorpe and Arundel have very different ways of interacting within that complex set of literate and institutional frames. From a pragmatic and CA perspective, we can describe Thorpe and Arundel as conversation participants marked by their power differential who struggle for footing in a conflict dialogue. As we saw in the Miller’s Prologue and Tale and will see later in Margery Kempe’s encounter with the Archbishop of York, discursive footings constitute the presuppositions, maneuvers, and implicatures which participants adopt or use to act and interact in a talk event (Goffmann 1981: 128). Each participant in a speech interaction adopts a stance or position ( footing) or, because talk is dynamic, different footings at different times during the interaction. Footing can

How Here tics Talk, According to Bernard Gui and William Thorpe

189

change slowly or frequently in a given talk, and such changes mark the extent to which power, control, agency, and community are at stake in our interactions. Individual utterances are produced with certain intents and meanings and are received within certain understandings (also intentions), expectations, semantic contexts, and implicatures. Alignment or misalignment between participants’ understandings of what has been said or meant by an utterance reflect the degree to which participants can be said to be adopting the same conversation frame, sharing the same presupposition or context for what the talk is accomplishing, and meaning the same thing by what is being said – or not. Frames and footings are closely related. A frame presupposes what roles or positions, footings, participants can or are expected to adopt in their interactions. But in action a participant’s stance or footing will largely determine the experiential rather than assumed frame she understands herself to be acting and participating in. Assumptions or representations of power and cognitive understanding are inherent to the relations of frames and footings. Saltwood Castle, the Archbishop’s residence and prison, embodies the established Church’s hierarchical power. There, in “a privy closet” with “a phisician þat is clepid Malverne” as an observer, Thorpe struggles for footing in his overtly hostile exchange with Arundel (ll. 173, 176). Thorpe combines direct assertion and claims to speak sincerely and with reasoned argument with numerous pragmatic strategies for exploiting verbal polysemy (equivocatio, re-impositio) and conditional or hedged syntax. He uses these discursive strategies to resist Arundel’s efforts to get him to swear explicitly and sincerely that he will conform his religious beliefs and public utterances to the doctrine of official “holi chirche.” For his part, Arundel tries to bring Thorpe around with both reasoned argument and explicitly hostile speech acts (face threats) such as demand and threat, backed up by his institutional authority to imprison or even execute Thorpe. The interaction between Thorpe and Arundel is anything but cooperative. Thorpe regularly shifts his discursive footing when required to submit and obey, while Arundel vacillates between solicitation, forbearance, and condemnation. One of the most prevalent discursive struggles in the text is the way Thorpe and Arundel repeatedly clash over the definition and authority of “holi chirche” and other key words, especially those related to the doctrine of transubstantiation. Resemanticization (re-impositio) is a powerful pragmatic strategy for resistance. Throughout the dialogic narrative Thorpe elaborates and defines “holi chirche” in ways aimed to hem in Arundel’s claim to speak with the authority of the official Church and therefore his authority to

190 

The Medieval Life of L anguage

require Thorpe’s obedience. Beginning with a declaration of faith, a mostly orthodox version of the Creed, Thorpe says he will humbly and obediently submit himself to “holi chirche” as he, not Arundel, defines the term: I bileve þat þe feiþful gederinge togidre of þis peple, lyvynge now here in þis liif, is þe holi chirch of God, fiȝtinge here in erþe aȝens þe fend and þe prosperite of þis world and her fleischli lustis. And forþi þat þe hool gederinge togidre of þis forseid chirche and every part þerof neiþer coveitiþ, ne wilneþ, ne loveþ, ne sechiþ ony þing, no but to eschewe þe offence of God, and to do his plesyng wille, mekeli, wilfulli and gladli of al myn herte I submitte me to þis holi chirche of Crist, to be [ever] buxum and obedient to þe ordinaunce and þe governaunce of it and of every membre þerof aftir my kunnynge and my power by þe help of God (ll. 285-295. My emphasis.).

Thorpe introduces pronoun trouble as a pragmatic strategy of dissent. His heavy use of the deictic and anaphoric pronouns þis and it foregrounds his definition of “holi chirche” and his condition for his obedience. The deictic pronoun þis and adverbs of time and place now and here also establish his stance and his antagonism with Arundel’s. They make clear through pragmatic orientation he is speaking on behalf of people in the lived and immediate present and not about the abstract ‘official’ Church. Thorpe’s referent for “holi chirche” is near to him and something he himself is part of (“þis holi chirche of Crist”), but outside the official Church as Arundel represents it. For Thorpe, the true church is the present community of the truly faithful as determined by Lollard belief. Thorpe’s speech takes possession of the ideational referents and directly challenges Arundel’s authority. Thorpe surrounds his definition with the language of faithfulness, obedience, and compliance, just not to the institution Arundel asserts is the proper referent for the name “holi chirche” (cf. Schirmer 2009). Later, after an especially difficult and angry discussion about Thorpe’s preaching, Arundel again asks Thorpe what he means by “holi chirche.” Thorpe answers with a slightly peevish tone: “And I seide, ‘Ser, I toold to ȝou bifore what was holi chirche. But siþ ȝe axen me ȝit þis demaunde, I clepe Crist and his seintis holi chirche’” (ll. 898-901). When Arundel pushes him to be more explicit, that is, less equivocal, Thorpe replies that “holi chirche” is comprised of “hevenli pilgrimes” who abhor “ony doctrine discordinge from holi writt” (ll. 911-913). Thorpe’s discourse resemanticizes “holi chirche” to mean not the established Church and its officials nor its material power, but the people who follow scripture alone in neo-apostolic faith and “treweli”

How Here tics Talk, According to Bernard Gui and William Thorpe

191

reject worldly desires and vanities. Thorpe’s definition in his talk aligns with Wyclif’s notion of the “congregatio praedestinatorum” (e.g., English Wycliffite Sermons, 1987: 84, 1990: 233; Hudson 1988: 314-327). Thorpe consistently resists the Archbishop’s effort to coerce a ‘proper’ response by refusing to accept the presuppositions and footings offered in Arundel’s questions. Thorpe is as uncooperative as he can be with respect to Arundel’s intentions, which he construes as implications. In these instances, Thorpe’s pragmatic strategy is not to distract or evade but to seize control of the debate by redefining terms and conditions away from how his accusers use the terms. For example, he distinguishes literal or fleshly referents from spiritual ones and true from false actors. He justifies his disapproval of religious music by interpreting David harping before the Lord in a “goostly” manner rather than “after þe letter” (ll. 1254-1255). In an extended discourse on the meaning of “trewe pilgrimes,” Thorpe the narrated I defines true pilgrim as anyone “travelynge toward þe blis of hevene whiche, in þe staat, degree or ordre þat God clepiþ hem to, bisien feiþfulli for to occupie alle her wittis, bodili and goostli, to knowe treweli and to kepe feiþfulli þe heestis of God …” (ll. 1238-1241). The adverbs “feiþfulli” and “treweli” sharpen his definitions and his footing. Thorpe concludes with a direct assertion that his criteria for “trewe pilgrimes” are nonnegotiable: “Þese hevenli condiciouns and suche oþer have þo pilgrimes eiþer þei bisien hem to have, whos pilgrimage God acceptiþ” (ll. 1277-1278). Referring to literal pilgrims, Thorpe explicitly judges contemporary religious practices as failing to meet his criteria: “… þe moost parte of hem, boþe men and wymmen, þat gon now on pilgrimage have not þese forseide condiciouns” (ll. 1279-1281). In the debate between Thorpe and Arundel, literal and figurative interpretations of scripture and doctrine turn out to be not in a fixed interpretive relationship but are fluid discursive strategies for establishing footing and semantic contexts and deployed contingently depending on the stance taken by individual participants at particular moments. Often in the narrative, Thorpe’s shift between literal and figurative interpretation seems to be determined in opposition to whatever interpretive stance Arundel has chosen at that moment. I’m not suggesting Thorpe’s beliefs are ill considered or arbitrarily adopted. I’m making the pragmatic point that Thorpe’s assertions and replies to Arundel’s questioning constitute his unrelenting resistance to the Church’s institutional power and the imposition of a discursive frame not aligned with the institution. Equivocatio, reaccenting, and redefinition are key discursive strategies for Thorpe’s resistant noncooperation and struggle for footing. Another strategy Thorpe exploits to gain footing and control over Arundel’s terms and conditions is adverb and adjective qualification. He

192 

The Medieval Life of L anguage

repeatedly attaches a strategic modifier to certain important institutional or doctrinal nouns or verbs in order to recontextualize the meaning of the term and thus retext (unravel, reweave) Arundel’s assertions. Thorpe resignifies the assumptions and implicatures of the terms and institutional authority which initially warrant them. Lollard discourse frequently used this strategy to distinguish, for example, the “false” Church and pilgrims from the “trewe” Church and pilgrims. Thorpe’s maneuver is double-voiced. His speech can seem at least superficially to use the same language (terms) as Arundel or official Church discourse, but he is intending those terms ‘slant,’ as Emily Dickinson might say, in order to adhere in his own mind with Lollard beliefs and principles for what is proper, scripturally authorized Christianity. For Thorpe, true Christian faith is founded on a neo-apostolic presupposition of “symple,” “wilful” poverty, “þe fredom of Cristis gospel,” and the responsibility of all Christians to “preche bisili, frely, and treuli þe word of God” (ll. 713-714), the “truþe.” Similarly, when Arundel challenges Thorpe to interpret Chrysostom’s sentence “bene jurare peccatum est” [it is a sin to swear well], Thorpe engages in a complex sophisma analysis which resituates the passage in a different discursive context. He embeds Chrysostom’s sentence in a matrix clause signifying a mental act, “þei knowen …” or “þei gessen and seien” (ll. 1713-1714). Then Thorpe concludes people are wrong to say so and “… þis sentence of Crisostom mai be aleggid skilfulli aȝens alle sich swerers, witnessinge þat alle þese synnen grevousli, þouȝ þei d[em]e hemsilf to sweren in þis forseide wyse wele. For it is yvel don and gret synne for to swere truþe, whan in ony manere a man may excuse him wiþouten ooþ” (ll. 1726-1730; cf. Somerset 1998: 184-191). Here, Thorpe’s interpretive strategy is similar to the kind of evasive or echoic sentence embedding Gui attributes to Waldensians during heresy examinations. Arundel is suspicious of all Lollard speech, assertive and nonassertive, and he doesn’t usually accept Thorpe’s redefinitions and contingencies. However, in the case of Chrysostom’s proposition, he admits “þat Crissostom myȝte be þus undirstonde” (ll. 1731-1732). Thorpe’s use of academic and logical argument shows he can walk a fine line between addressing Arundel’s questions with some directness while not to his mind betraying his religious beliefs. The section on the doctrine of transubstantiation is another example of how the Testimony foregrounds Thorpe’s discursive strategies in an academic mode. Readers have noted how in the sacring bell episode Thorpe and Arundel engage in academic-style debate regarding the definition of key terms and conditions which shape the doctrine of transubstantiation. Somerset (1998: 180-215), for example, discusses how Thorpe’s use of academic

How Here tics Talk, According to Bernard Gui and William Thorpe

193

modes of argument in religious contexts outside the schools illustrates the way intellectual discourse was disseminated and deployed as part of Lollard communication practice and vernacular theology (cf. Thomson 1989). I suggest the dialogue between Thorpe and Arundel about the Eucharist also shows how intellectual modes might be incorporated or used differently by participants in medieval discursive pragmatics. Arundel informs Thorpe of the Shrewsbury elders’ charge that in his sermon at St. Chad’s church he had “opinli” (clearly, explicitly) preached “þat þe sacrament of þe auter was material breed after þe consecracioun” (ll. 931-933). Thorpe of course denies the charge, but not outright. He argues that the words he used in the sermon implied a different meaning with a different intent. He shifts the frame and context away from what he said in his sermon to how when people in the church heard the sacring bell and rushed away from him to see the sacrament, he urged them to stay and “heere Goddis word” and fortify “þe bileve þereof þat ȝe owen to have in ȝoure soulis” rather than simply submit themselves to “þe outward siȝt þerof” (ll. 934-943). Thorpe responds to the elders’ charge by setting up a different interpretive context, not explicitly denying transubstantiation but affirming the liturgy of the Word as equally nourishing to the soul. At this moment, Arundel unsurprisingly says he doesn’t believe Thorpe is telling the truth (“I trowe þee not, whatevere þou seist’”). But then according to the text Arundel immediately and for unclear reasons offers Thorpe the floor to explain his view on transubstantiation, specifically the knotted question about “material breed.” Would Arundel really have offered Thorpe the chance to explain his doctrinal position just after he says he doesn’t believe a word the man says? Unlikely. But the question was standard for Lollard examinations (Hudson 1985). The pragmatic and discursive gap may be textual, or it may be an unexplained disjunction in the two men’s interaction, a kind of narrative jump cut in the text. The passage raises questions about the literal accuracy of Thorpe’s textual narrative. However, the text skips past that disjunction, and Thorpe immediately replies to Arundel that “I knowe nowhere in holi writ where þis term ‘material breed’ is writun” (ll. 947-952). In medieval and modern theories of discursive pragmatics and conversation strategies, topic and context are crucial markers of discursive cohesion. In the Testimony Thorpe and Arundel’s ways of talking about a topic give us some striking examples of metapragmatic awareness. When he replies to Arundel, Thorpe shifts the footing for his speaking about transubstantiation and implies that all doctrine, not just transubstantiation, depends on scriptural authority. With their different dialogic strategies, the speakers’ struggle to control the floor and enframe

194 

The Medieval Life of L anguage

doctrinal information in the way each wants to. Frame, footings, and topic (doctrinal material) establish the controversial relationship between the two men as representatives of the established Church and Lollard dissenters. Thorpe explains the doctrine of transubstantiation by giving priority to “holi writt” and his vernacular definition of forme in the context of the phrase “verri Cristis fleisch and his blood in forme of breed and wyne” (ll. 968-969). Thorpe is exploiting the potential semantic ambiguity of English forme. According to Arundel, Lollards say that “forme of breed” is synonymous with “substance of breed” or “material breed,” whereas official Church doctrine declares that after the priest’s consecration the host becomes the true body of Christ. Rather than accept Arundel’s terms, Thorpe spends the entire episode dodging Arundel’s demand for explicit agreement or an explicit statement to the contrary: “Telle out playnli þi bileve þereof’” (l. 957). The Archbishop wants Thorpe to issue a bald on-record statement which Arundel can then use to charge Thorpe with teaching heresy or a different statement he can interpret to mean that Thorpe agrees with the doctrine of holy Church. Thorpe expresses neither. Instead, he evasively shifts the immediate referent of their talk to a more general context and asserts the authority of scripture for all religious practices and doctrines. Thorpe then cites other texts which support the meaning of English forme as kynde instead of material or substance (ll. 951-952, 974-985). The more demanding Arundel becomes for Thorpe to answer Yes or No to a direct question, the more Thorpe avoids directly answering the Archbishop at all. Thorpe’s strategy for handling Arundel’s demands for explicit answers is to introduce a topic related to but different from the one Arundel proposes or to redefine a key word in Arundel’s utterance. Bacon might have referred positively to Thorpe’s strategy as transfer, analogy, or equivocatio. At one point, Thorpe even reverses the discursive roles and asks Arundel, “And also, sir, seiþ not þe churche in þe houres of þe moost blessid Virgyne … wheþer þe ‘forme of our bodi’ be not clepid here þe kynde of oure bodi?” (ll. 981-984). Sensing the ploy, Arundel himself refuses Thorpe’s request for an answer by posing another question at a metapragmatic level: “Woldist þou make me to declare þese tixtis to þi purpos?” (ll. 986-987). Arundel responds to Thorpe’s maneuver by declaring his institutional power: I’m the one to ask the questions. Arundel knows pragmatically as well as Thorpe that there’s more than one way to evade a request for a direct answer. Talk is structured, ordered, but shifting topics so as to allow the talk to continue or bringing a conversation to a satisfactory close are complicated matters. Often, such changes require participants to engage in some cooperative maneuvering, but just as often there can be a struggle for footing

How Here tics Talk, According to Bernard Gui and William Thorpe

195

as participants perceive their stance needs to shift or terms need to be redefined or they don’t want to agree to close the topic or conversation (cf. Schegloff and Sacks 1973). Conversation analysts assume that an interaction is ‘ordered’ according to conventions of talk in a given culture or speech community (cf. Hymes 1962). In Thorpe’s narrative the conversation topics are sequenced according to the order of the five charges Arundel asserts against Thorpe’s preaching. However, to shift topics, the two use more complex discursive strategies. As I said, Thorpe, not Arundel, is the one who initiates closing the topic of transubstantiation when he evades Arundel’s direct question as to whether he believes in the sacrament as official Church doctrine. Once again, Thorpe’s reply shifts the context to a more general level by questioning the meaning and status of belief in different contexts and whose belief is at issue: “Ser, whatever prelatis han ordeyned in þe chirhe anentis þis mater and oþer, oure bileve stondiþ evere hool. And I have not lerned þat þe ordinaunce of men bineþe þe bileve schal be putt into bileve” (ll. 991-994. My emphasis.). Thorpe embeds a bald on-record assertion of collective belief (“oure”) within other clauses and footings. His first-person plural pronoun sets up an opposition between “oure bileve” and whatever the Church has “ordeyned.” Thorpe’s plural pronoun creates a community around him. In the context of the examination, Thorpe and Arundel frequently use bileve as both Noun and Verb, but they do not always imply the same semantic presuppositions or contexts. Which preposition governs the noun “bileve” is implicitly a continuing topic in the Testimony. The meaning of bileve is determined by the kind of syntactic construction it is embedded in. Earlier, Thorpe had challenged Arundel by putting words in his mouth, saying: “Sir, siþ ȝe demen me an eretike out of bileve, wole ȝe ȝeve me audience to telle to ȝou here my bileve?” (ll. 195-196. My emphasis.). In Thorpe’s speech at lines 991-994, he draws a crucial distinction between “bineþe þe bileve” and “putt into bileve” (cf. Two Wycliffite Texts, 1993: 118, note on Testimony, ll. 992-998). For Thorpe, “bineþe þe bileve” refers to what Church leaders and authorities declare to be doxa; the preposition bineþe implies the subordination of Christian lay people. “putt into bileve,” on the other hand, refers to what scripture alone authorizes for all Christians. Thorpe of course adopts the second meaning, Arundel the first. Then Thorpe, not Arundel, concludes the questioning on the sacrament by shifting topic again to an even wider procedural context. He dismisses both the terms and the academic mode of reasoning which he and Arundel have just been engaged in (ll. 1026-1038). In effect, Thorpe dismisses both his and Arundel’s earlier discussion by saying it doesn’t really matter what is the proper belief regarding transubstantiation because he

196 

The Medieval Life of L anguage

can’t commit one way or the other to the academic terms “no substaunce of breed” and “accident wiþouten soget.” Having just demonstrated he can hold his own in academic debate with the Archbishop, Thorpe then privileges the literal text of scripture and rejects academic debate as a mode of argument for theology along with Aquinas’ support of the doctrine of transubstantiation because, says Thorpe, the words “an accident wiþouten soget” do not appear in scripture (ll. 1047-1050). In the Testimony Thorpe often privileges the text and literal or historical sense of scripture over logic or theology, but not always (as with his strategic interpretation of the Chrysostom passage). With his evasive strategies in this section of the narrative Thorpe exploits potential equivocatio in order to contextually redefine or qualify key words, while at the same time asserting the textually encoded authority of scriptural language in order to counter the ‘word’ of the official Church and Arundel’s claim to doctrinal authority. In this way Thorpe gains some footing in the interaction or at least mitigates Arundel’s control of the hostile discourse while also continuing to assert or imply his own dissenting beliefs. In the remainder of the text Arundel questions Thorpe on other charges regarding his beliefs and statements on pilgrimages, images, and tithes, all standard Lollard examination topics. Accordingly, Thorpe deals with Arundel’s questions by using the same evasive discursive strategies of polysemy, equivocatio, footing conditions, and topic shifting. Clearly, pragmatic factors such as sincerity, context, and presupposition are in dispute in Thorpe’s text. Olivi had made sincere intention and metapragmatic context crucial for his theory of sincere communication. Gui presumed sincerity is inherent in orthodox belief and therefore that evasive but recalcitrant heretics speak insincerely when interacting with official examiners. In Thorpe’s Testimony, speaker intent and truth-telling are from Arundel’s point of view the principal criteria for cooperation. But, as we know, the text represents Arundel as skeptical of practically everything Thorpe says. He repeatedly declares metatextually he doesn’t believe what Thorpe says when answering his questions or demands: “I trowe þee not, whatever þou seist …” (ll. 947-948); “Wherto tariest þou me wiþ sich fablis?” (ll. 611-612). Both Gui and Arundel were experienced examiners of presumed heretics and nonconformists, so we can imagine Arundel hearing Thorpe’s speech in Saltwood through a filter or frame of ‘heretic’ insincerity and deception, based on his presuppositions about suspected heretics and earlier reports of Thorpe’s behavior and speech. As Arundel says at the start of the interview, Thorpe’s reputation precedes him: “William, I knowe wel þat þou hast þis twenti wyntir and more traueilid aboute bisili

How Here tics Talk, According to Bernard Gui and William Thorpe

197

in þe norþ lond and in oþir diuerse contrees of Ynglond, sowynge aboute fals doctryne …” (ll. 180-183). More than once Arundel tells Thorpe to make sure “þi conversacioun be suche þat þin herte and þi mouþ acorden trewli in oon …” (ll. 962-963). Arundel’s admonition has the flavor of a Gricean maxim. He constructs Thorpe as a speaking subject by insisting, but not assuming, that Thorpe speak sincerely and truthfully. Pragmatically, the Archbishop as examiner begins from an accusatory and skeptical position, in effect demanding that Thorpe ‘prove’ himself by giving Arundel the responses he wants. That would mean Thorpe would be recontextualizing his own sense of his speaking position on the fly. Thorpe for his part consistently evades directly answering many of Arundel’s questions according to the terms and conditions the Archbishop proposes. Pragmatically, Thorpe recontextualizes his speaking position, but not in the way Arundel wants. The hostility inherent in the two men’s encounter is indexed by Arundel’s insulting speech and angry responses to Thorpe’s utterances, but their selfrepresentations and forms of address are quite different. In the text Thorpe verbally respects Arundel’s social and institutional position. He represents himself as vulnerable, sometimes fearful for his life, and as someone who conducts himself with deliberation in good conscience. He repeatedly addresses Arundel as “Ser” (Sir), even when being assertive in response to Arundel’s questions; e.g., “And I seide, ‘Ser, I telle ȝou at oo word, I dar not for þe drede of God submitte me to ȝou aftir þe tenour of þe sentence þat ȝe haue aboue rehersid to me’” (l. 614). Thorpe is surprised and chagrined to discover that a visitor to him in prison had betrayed him by reporting back to the Archbishop (ll. 1830-1832) and saddened by reports that fellow Lollards have recanted. In several places in the text Thorpe, the narrating I, tells the reader his interior thoughts and affective reactions to what was said and done around him (e.g., ll. 365-368). At one point (ll. 408-436), when Arundel threatens to burn him at Smithfield, Thorpe becomes silent. The text represents Thorpe’s interior deliberations as to how best to respond to Arundel’s overt face threats and life threats. The passage’s interior narration explicitly reveals Thorpe’s presuppositions and reasoning in the midst of the conversation, how he interpreted Arundel’s motives and why he responds to Arundel as he does at that moment. He infers (“And I lokinge biheeld inwardli þe Archebischop”) that Arundel has no remorse for having burned William Sawtry as a heretic and now is thirsting for “more innocent blood.” (On Sawtry’s trial and subsequent burning as a heretic in 1401, see Strohm 1995, 2006: 32-62; Forrest 2005: 34, 105, 202.) Thorpe responds to this intimidation by using another topic shifter. He engages in a detailed personal narrative which effectively recontextualizes

198 

The Medieval Life of L anguage

Arundel’s stage setting for the examination. He establishes a different relevance frame, one based on personal testimony rather than institutional doctrine and obedience. He tells how he reluctantly became a priest, how he formed his reformist Christian beliefs based on scripture, his conscience, and the teachings of Wyclif, how he fears promising to do what Arundel what wants him to do because he won’t in good conscience be able to deliver on what he promises and therefore would be speaking falsely, not to mention relapsing as a heretic. Thorpe concludes with why his actions and speech are more honest and sincere than those of the Lollard recanters, Purvey, Hereford, Bowland, and Repingtoun (ll. 427-516). Recontextualization and relevance framing are Thorpe’s principal metapragmatic strategies at this moment in the dialogue. Arundel, on the other hand, is represented in the text as an angry, insulting bully who commands obedience (e.g., “Submitte þee þan now here …”; l. 2063). After initially addressing Thorpe respectfully as “William” (l.180) and being rebuffed, he quickly and repeatedly refers to Thorpe as “lewid losel” (ll. 406, etc.). Clanvowe [1341-1391) in his The Two Ways used the word in his description of insults thrown at religious nonconformists: “swiche folke þe world scoorneth and hooldeþ hem lolleris and loselis, foolis and schameful wrecches”; 1975: 70.) At several points the Archbishop threatens Thorpe with harsher punishment or death if he does not “consente to myn ordynaunce and submytte þee to stonde to myn decre” (ll. 407-408). The text represents Arundel’s anger and bullying tactics as both explicit and implicit. Sometimes the narrating Thorpe describes Arundel’s mood or tone with affect terms (“as if he hadde ben wrooþ,” “seide to me al angrili”; ll. 617, 1387) – modus conceptus. Other times he represents the Archbishop’s hostility implicitly, most famously when Arundel slams his fist against a cupboard in anger at Thorpe’s evasions and “spake to me wiþ a grete spirit …” (ll. 2070-2071). Dialogism depends on the subtlety of linguistic features. One of the most interesting linguistic features in Thorpe’s Testimony from an historical pragmatics perspective is the speakers’ use of personal pronouns to indicate social status. Arundel addresses Thorpe with the superior pronoun þou. Thorpe returns inferior ȝe/ȝou to the Archbishop (ll. 1540, 1546, etc.), but he uses second person pronoun address less frequently in the text than does Arundel. Given that the Testimony is dated to 1407/8, the text provides evidence for early fifteenth-century English conventions and presuppositions about politeness deixis as presented in written discourse which records or purports to record or reenact spoken interaction. Many histories of the English language assign the regular use of the second-person pronouns þou/ þee and ȝe/ȝou to mark speakers’ social status to the late sixteenth century.

How Here tics Talk, According to Bernard Gui and William Thorpe

199

But Arundel and Thorpe’s regular use of the þou/þee-ȝe/ȝou, as presented by Thorpe in the text, suggests a much earlier date for the T/V structure, at least in written English (cf. Bergs 2004). After command, the most salient speech act in the Testimony is swearing (promising, pledging), but that speech act is discussed (modus conceptus) more than performed (modus exercitus). At one point, after Thorpe declares his beliefs and the Archbishop begins to question him on the charges brought against him, Thorpe addresses Arundel directly and asks, “But, ser, I praie ȝou þat ȝe wole telle me if aftir ȝoure biddyng I schal ley myn hond upon þe book in entent to swere þerbi?” (ll. 332-334). Arundel replies, “Ʒhe! wherto ellis?” (l. 335), as if to say, ‘Sure, why else are you here?’ (Arundel uses the interjection ȝhe as roughly equivalent to Modern English ‘Sure.’) The two then engage in a complex discussion of swearing as a speech act in relation to scriptural authority and the material and hermeneutic status of the book as an authoritative ritual object. In this case, swearing signifies promising, as opposed to swearing which signifies speaking blasphemously. Implicitly, Thorpe and Arundel are discussing the pragmatics of speech and writing. Thorpe launches several critiques of swearing as a ritual speech act in the way Arundel seems to understand the speech act. One critique is material, another is based on the pragmatic notion of intention, and a third on the authority of scripture. In effect, Thorpe destabilizes the presuppositions which support Arundel’s commands for obedience via the speech act swearing on the Bible. He also provokes the Archbishop with his own questions, thereby seizing the discursive floor and directing the talk on his own terms, at least for a time. Materially, Thorpe challenges the speech act swearing’s felicity condition. “to swere bi a book,” he says, “is to swere bi dyverse craturis,” which is prohibited by the laws of God and humans (ll. 337-339). Thorpe works out a critical strategy of material componential reduction which undermines the religious and cultural resignification of books as more than material objects. The material production of manuscript books from sheepskin and leather, he says, means that books are just compilations of animal parts. Wyclif in De veritate sacrae scripturae had made a similar argument regarding the transcendental rather than material nature of ‘the book’ (1905: 114-115). In his De Contrarietate Duorum Dominorum Wyclif argued that the material form of linguistic expression was not relevant to the intentional meaning expressed by a text, and he dismissed criticisms of the burning of manuscripts of “God’s law”: “Lingua enim, sive hebrea, sive greca, sive latina, sive anglica est quasi habitus legis domini” [for language, whether Hebrew, Greek, Latin, or English, is nothing but the clothing of the Lord’s law] (1883,

200 

The Medieval Life of L anguage

2: 700; cf. Hudson 1986). Thorpe follows Wyclif by refusing to adopt either the conventional presuppositions of book production as the production of authoritative language or the ritual fetishization of the material Bible as a divine text. For Wyclif, the material form of a text is a disposable wrapper. We wonder what Thorpe might have thought about the semio-cultural significance of his own act of writing a letter to his community. Later in the dialogue, Thorpe returns to the topic of the text and material book when one of the clerks in the Archbishop’s household instructs him to swear on the Bible. Citing Jerome, Thorpe distinguishes touching the Bible and reading the Bible and argues, as Stanley Fish might, that “þe lettre þat is touchid wiþ mannes honde is not þe gospel, but þe sentence þat is verily bileued þat is ver[tu of] Goddis word is not in þe leues of a book but it is in þe roote of resoun …” (ll. 1772-1776). In this scene Thorpe resemanticizes the word book to affirm the cognitive ‘text’ in the mind. This argument is in line with Lollards’ critiques of ‘false’ pilgrims, images, interpretation, doctrine, preaching, and so forth. The true gospel, he says, is not “in þe writynge aloone of lettres but þe gospel is in þe marwȝ of þe sentence of scripturis” (ll. 1776-1778). Not the letter but the spirit. Not the material words but the meaning. Not the skin but the marrow of the sentence. Not the explicit utterance but the text as understood and remembered. Thorpe here is making a distinctly pragmatic interpretation of utterance and understanding. He destabilizes the authority of the speech act swearing an oath by shifting the referent of the word gospel or book from inscription to meaning. Saying that a word’s or a book’s meaning is more important than its phonetic or written form might seem obvious or trivial to some, but in the context of his interaction with Arundel, Thorpe’s distinction takes on strategic pragmatic significance and relevance. That Thorpe may have won this struggle on his own terms is suggested by a final bit of dialogue in the scene: “And þanne a while þeraftir þe Archebischop seide to me, ‘Wolt þou not submitte þee to þe ordynaunce of holy chirche?’ And I seide, ‘Sere, I wol ful gladly submytte me as I haue schewid to ȝou’” (ll. 2229-2232). The text leaves Thorpe’s offer still on the table. Thorpe’s critiques of the speech act swearing are in line with some other medieval metapragmatic awareness and critiques, especially regarding intention and ritual. To recall some earlier medieval contexts and arguments on intention and sincerity: We have seen how some thirteenth-century grammarians developed a pragmatic theory of language based on the role of intention in communicative understanding (Chapters 1 and 2). We have also seen how Gui profiled heretic talk by explicitly analyzing and harshly critiquing Waldensians’ spoken syntax, semantics, and gestures in order

How Here tics Talk, According to Bernard Gui and William Thorpe

201

to expose how, he believed, they were intentionally saying one thing but meaning another. For Gui, Waldensians’ intentional equivocation and evasive refusals to affirm the official Church’s doctrine or protocols were signs of their heresy. Later medieval pragmatic explorations of sincerity and intention were informed by earlier medieval logic and philosophy of language and the implications of intention as part of logic and discourse theory. In the twelfth century, for example, Abelard adopted a strong subjective contextual approach to inference and semantics with respect to intention. What a speaker means to say in a particular context, not just the form and order of the words, should guide the listener’s understanding. In Dialectica Abelard develops the topic of equivocatio in ways similar to Augustine and later Bacon but not to those of many medieval logicians. He argues that a particular vocalization or word form can be attached to different meanings depending on the context of use. Abelard even uses the same canis example of equivocatio found in other twelfth- and thirteenth-century logic texts (Dialectica, ed. De Rijk, 1.1 [on vox, intention, and equivocation]; cf. 2.1 [on predications, ubi, de situ, etc.]; 5.1 [oratio and signification]), but he does so to propose a theory of contextual meaning in response to the ‘semantic realism’ of theologians like Bernard of Clairvaux who opposed Abelard’s mode of reasoning dialectically into faith. In his Theologia “Summi boni,” Abelard exploited intention and the logical idea of ​​equivocatio as a positive feature of reasoned religious discourse. According to Abelard, the term persona, for example, has four meanings and therefore different referents depending on the discursive context in which the word is used or received: theological (the being of God in three persons), rhetorical (juridical person), poetic (character or actant), and grammatical (speaking position or mode of address). Not only can some words have different meanings to refer to different things, but in theological discourse the same word can also refer to different things at the same time because the word can implicate sacred and secular contexts simultaneously. Therefore, persona appropriately names the Trinity in that Father, Son, and Holy Spirit are defined differently to one another, yet the names all refer to the one God (Mews 2002: 159-203). Abelard isn’t exactly posing the same question as Frege’s reference and sense, but he is definitely walking on the same side of the semantic/pragmatic street. To understand what is said and meant, we must seek (attend to) and then understand as much as we can the context in which an utterance is situated. Mews (2005) believes that for ‘context’ in written language, Abelard means not only speech situation and textual situation but also how an author typically uses certain words

202 

The Medieval Life of L anguage

and what his motive or intention might be for making the utterance or text in the first place. This is a good description of the dialogism represented in Thorpe’s Testimony. Without explicitly foregrounding the interpreter or reader as the constructor of meaning, Abelard upholds the hermeneutic and pragmatic principle that the meaning of any utterance, sentence or text, is greater than the sum of its individual parts. Meaning is not componential. It’s pragmatic, inferential, and in a sense holistic and contextual. Abelard’s theory of intentionality fully informs his theory of ethics too. Intending to say or do something and then consenting to carry out that intention are more important for discerning a person’s moral status than what the person actually says or does. Individual acts are morally indifferent in themselves. Acts of charity, for instance, can be carried out for different reasons and with different motives, and not all of them necessarily demonstrate one’s love of God: “ita cuiusque intention bona in se vocatur, opus vero bonum non ex se apellatur sed quod ex bona procedit intentione” [so anyone’s intention is called good in itself, although the work (opus) is not called good by itself but because it proceeds from a good intention.] (Ethica seu Scito teipsum, ed./trans. Luscombe 1971: 46/47). Abelard sometimes used the term “consent” (consentio) for what other philosophers and theologians referred to as “will” (voluntas) (cf. Marenbon 1997: 251-264). Implicitly in line with Abelard theories of intention and contextual meaning, Thorpe’s Testimony makes dramatically explicit some of the ethical and pragmatic implications of intentionality. When Thorpe refuses to swear to what Arundel asks him to swear to, he implies that speaker intention governs the meaning of an utterance as opposed to the physical act of pronouncing particular words in a particular order and performing them with certain gestures (kneeling, etc.). At this moment in the text, Thorpe aligns himself with two general pragmatic principles: 1) the meaning of a speech act, or any speech for that matter, is not what is said but what is meant, and 2) the context of an utterance shapes but doesn’t determine an utterance’s intention and shapes but doesn’t determine how an utterance is received and understood by the listener or reader. In the context of his exchanges with Arundel, says Thorpe, coerced behavior and speech are insincere and therefore untrue. When Thorpe critiques Arundel’s use of the speech act swearing, he implies but does not baldly assert that he does not agree with or believe what Arundel wants him to say he believes. In the Testimony this implication turns out to be another effective strategy of evasion. Thorpe tells Arundel he will speak falsely if he swears publicly to what Arundel wants him to. This might

How Here tics Talk, According to Bernard Gui and William Thorpe

203

seem to be an obvious point, given that the Testimony is a personal account of a Lollard’s examination by a hostile religious authority. Nonetheless, the strategy is crucial for how we understand Thorpe’s metapragmatic awareness and how that awareness saturates the narrative and why the Testimony is an important source for medieval pragmatics in practice. Struggles for discursive footing can take place in several ways. Sometimes Thorpe resists Arundel by debating the meaning of terms, but in his critique of swearing he also implicates the presupposition that sincere intention governs the meaning of any utterance. Much of the Testimony recounts how Thorpe and Arundel struggle explicitly over the meanings and implicatures of key terms, conditions, and gestures. I would go further and say their individual discursive and pragmatic strategies and their hostile interaction also represent implicitly a struggle over intention, sincerity, and metapragmatic awareness in public religious discourse. Arundel is demanding submission to the authority of official Church. Thorpe is standing on sincere belief and the integrity of religious purity. Words and meanings matter, and the men’s disagreements as to the meaning of bileve, holi chirche, and other key terms and concepts indicate how their interaction is predicated on a struggle for control of discursive space in a vigorously hierarchical institutional context. Arundel is the examiner and so occupies an established position of power. But the actual discursive exchanges reveal that social and discursive power is more fluid, albeit temporary, in particular contexts. In the Testimony, however self-interested, what that means is that Thorpe struggles successfully for discursive footing. It also means that his oppositional stance during his examination is comprised of different strategies for rebutting or evading Arundel’s efforts to coerce from him a ‘proper’ response. In some contexts, Thorpe appeals to the intent or meaning of utterances, scripture, or doctrine. In other contexts (for instance, his critique of swearing) he appeals to the precise form of words spoken or the literal text of scripture as the criterion for what should constitute Christian belief and practice. If he can’t find a sacrament or practice or doctrine in scripture, then Thorpe won’t accept that practice or belief as a true Christian one. Christian faith is or should be based on sola scriptura, as Wyclif and his immediate circle insisted. Thorpe especially uses this discursive strategy to argue against pilgrimage, transubstantiation, worship of the saints, and tithes. The differences between Thorpe and Arundel’s discursive and pragmatic strategies inscribed in the Testimony are ultimately framed by the institutional context and the Archbishop’s authority. The encounter is hostile and accusatory. Thorpe is on the defensive because he is charged with making

204 

The Medieval Life of L anguage

certain statements from the pulpit which the official Church regards as heretical. The situation implies that Thorpe is being charged with speaking heretically from within, not outside, the prescribed discourse domain of the official Church, unlike many other accounts of heretics. That the dialogue ends inconclusively only adds to the complexity of the situation. After Arundel dismisses Thorpe and leaves for London, the Lollard preacher is left in his cell in Saltwood, alone and thanking God for his deliverance. The narrating Thorpe imagines himself to have been in a quasi-crucifixion scene: “For as a tree leyde vpon anoþer tree ouerthwert on crosse wyse, so weren þe Archebischop and hise þree clerkis alwei contrarie to me and I to hem” (ll. 2245-2247). The text affirms a pragmatics of sincerity and true-speaking with true intention founded on Lollard beliefs and reformist activities. The Testimony makes preaching the sentence of the Gospel the most important kind of speech and provides sympathetic readers with narrative examples of how to talk and interact with authorities who challenge their freedom and beliefs as deviant or heretical. Thorpe deploys a range of exemplary pragmatic strategies for dealing with hostile, face-threatening, even life-threatening speech: equivocatio, evasion, recontextualization, syntactic modification, resemanticization, and at key moments assertive, stubborn, and elaborated modes of involved argument. Thorpe’s pragmatics are disruptive and destabilizing in the service of affirming his beliefs and public religious stance. The text, however, leaves his fate unspoken. Maybe Thorpe successfully evaded further interrogation or punishment, but I’m not sure Arundel heard what he said because, well, that’s how heretics talk.

6

Margery Kempe’s Strategic Vague Language Abstract This chapter continues the previous analysis of heretics’ speech from the perspective of Conversation Analysis. Bakhtin’s theory of dialogism sets Kempe’s pragmatic thinking in a sociolinguistic frame. The narrative of her examinations at the Archbishop of York’s court suggests that people’s thinking about how language defines, expresses, controls, and resists also informed how they pragmatically and metapragmatically constructed their speech for social survival, subjective authority, or agency in asymmetric or hostile interactions. Medieval grammarians’ and logicians’ concerns with reference and equivocatio (ambiguity, polysemy, vagueness) were reinterpreted in controversies about how heretics and nonconformists talk in hostile institutional situations. Kempe’s sophisticated use of evasive, vague, hedged, and recontextualized speech and situational pragmatics proves more than a match for the Archbishop and his clerks. Keywords: vague language, equivocation, echoic speech, pragmatics, Margery Kempe, conversation analysis

Like Thorpe’s Testimony, the Book of Margery Kempe tells us a lot about medieval pragmatics in action and about people’s metapragmatic awareness of speech strategies and discourse practices in different subject positions. Kempe (c. 1373-after 1438) was born into a well-respected merchant family in Bishops Lynn and herself became a member of the powerful guild of the Holy Trinity. She probably composed and dictated her Book over several years in the 1430s. The autobiographical text narrates events from earlier decades up to about 1436. The Book is full of episodes representing Kempe’s controversial, sometimes hostile interactions with clergy and lay people as well as her visions, positive religious connections, and friendships. Several episodes depict her examination by clergy to discern her orthodoxy or

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_ch06

206 

The Medieval Life of L anguage

conformity. Some episodes depict her in sympathetic or positive interactions, but the overall perspective as narrated by Kempe is one of controversy and difficult talk. Kempe narrates in great detail her contests with clergy over her religious conformity and her and their judgments about proper Christian behavior. Kempe dictated her narrative in English to at least two scribes at different times during the 1430s. It’s not altogether clear how the scribes’ work influenced the sole manuscript version once housed in the Carthusian monastery of Mount Grace, Yorkshire. Kempe’s life-writing is a scribally mediated and collaborative text. In this chapter, conversation analysis (CA) is again useful to read slowly a portion of Kempe’s text and explore the narrative’s pragmatics, metapragmatics, dialogism, and relevance. The episode is Kempe’s interactions with the Archbishop of York, Henry Bowet, and his court in early autumn 1417 (Book of Margery Kempe, ed. Lyn Staley [1996], Chapters 50-52. Citations are from Staley’s edition.). The situation is inherently hostile. Kempe has been summoned to the archbishop for questioning, just as William Thorpe had been ten years earlier by Archbishop Arundel. In her talk Kempe deploys a variety of pragmatic and metapragmatic strategies to maintain her positive and negative face, fend off face-threatening acts, and gain or maintain control of the exchange which is stacked against her. She uses vague, evasive, polysemic language and dialogic or embedded syntax to sidestep questions, redirect her responses, shift topics, and assert her beliefs. But in doing so she is suspected of being insincere, uncooperative, and possible heretical, like Thorpe and Waldensians. So how does Kempe get away with it? Or does she?

Cooperate or Else Before reading closely the Kempe at York episode, I want to say a bit more about the relation between discourse strategies, vague language, and relevance. As we have seen, medieval theorists, especially Abelard, Bacon, Magister Johannes, and Pseudo-Kilwardby, show how the intended meaning of many utterances, especially interjective, nonliteral, and elliptical utterances, are not usually interpreted simply from the form of the utterance (virtus sermonis). Contemporary theorists of speech acts, politeness, and relevance theorists generally agree. In many contexts, speaker meaning (intention, implicature) is underdetermined by logical syntax or explicit reference (explicature). One of the most forceful alternatives to Gricean pragmatics is Relevance Theory (RT). Grice asserted one of his maxims as “Be relevant” (1975). But

Margery Kempe’s Str ategic Vague L anguage

207

as with his Cooperative Principle and other maxims, what is ‘relevant’ depends on a number of factors, including who’s asking the question. In the Anglo-American tradition, both Relevance Theory and Politeness Theory (PT) revise Grice’s idea of Cooperation and argue that relevance is the most important principle for accounting for how we understand language. Brown and Levinson (1987) and Sperber and Wilson (1995, 2012) take issue with Grice’s logic-based theory and adopt more sociolinguistic frameworks for pragmatics. In doing so, they open RT to an engagement with what we have explored in this book as Bakhtinian dialogism, conversational difference, and socially-stratified discourse. Brown and Levinson’s politeness theory stays relatively close to Grice’s view of the rational goals of pragmatics and pragmatics research: “We wish to demonstrate the role of rationality, and its mutual assumption by participants, in the derivation of inferences beyond the initial significance of words, tone, and gesture. It is our belief that only a rational or logical use of strategies provides a unitary explanation of such diverse kinetic, prosodic, and linguistic usages” (1987: 56. My emphasis. Cf. Habermas 1998: 23, 43). They stress that speakers use propositional and non-propositional speech (e.g., interjections, questions, non-literal speech) to maintain their positive face (public self-image, esteem) and protect their negative face (freedom to act) by restraining face-threatening acts in speech. They take mutual face-saving and polite cooperation to be the norms for reasonable interaction (1987: 5-6). Unlike Grice, however, Brown and Levinson include as politeness strategies some speech behavior which does not conform with the Cooperative Principle: “Deference is not encoded in language by the use of arbitrary forms, but by the use of motivated forms” (1987: 23. Their emphasis.). Hostile interactions or asymmetric power situations present a challenge for PT in that face-threatening speech and acts disrupt the convention of cooperation and require different strategies from at least some of the speakers in order to maintain their agency, beliefs, or positive face. Sperber and Wilson (1995) take RT further and collapse Grice’s Cooperative Principle and Maxims into one overriding principle, Relevance. Like Brown and Levinson, Sperber and Wilson shift the pragmatic attention away from sincerity and intention as the most significant positions for effective communication and focus more on the dialogic relation between speaker and listener, both of whom check for relevance to interpret utterances. Sperber and Wilson argue that interpretations of relevance are the primary way speakers and listeners produce and understand utterances, manage cognitive load, and facilitate both participants’ successful online processing.

208 

The Medieval Life of L anguage

Pragmatic meaning depends on relevant context which depends on speaker and listener interpretations of context. When utterances are more accessible (relevant), the listener expends less cognitive and inferential energy construing intended meaning. Therefore, those utterances are more quickly grasped by the listener as relevant to the meaning and communication in the talk. It’s important to recognize that relevant contexts and language are adjusted for different audiences and different speech conditions. Therefore, perceived relevant contexts establish different footings for speakers. Sperber and Wilson (2012) take into account non-literal or nonconventional utterances. Such unconventional speech requires the listener to do more cognitive work to construct the/a meaning. The same goes for utterances which presuppose unexpected contexts or which use unfamiliar vocabulary or language codes. Such speech may still be relevant to the interaction, but listeners have to work harder to mentally construct a relevance context. In doing so, the relevant context may shift between speaker and listener, which then puts into question the conversation’s initial footing and frame. Recent analysis of impolite speech has expanded the field of PT and RT to explore how hostile or seemingly non-cooperative speech changes the understanding of relevance and relevance strategies. Is impolite speech structured? Are there scripts for disagreement, and if so, what functions do they have in interactions, and what purposes do they serve? For example, is there a role for (im)politeness in the speech of nonconforming subjects or those defending themselves from accusations? Our readings of Gui’s Practica, Thorpe’s Testimony, and Chaucer’s poetry have suggested how some medieval speakers, as textually represented, use impolite speech in hostile or threatening situations. In hostile situations, what counts as productive talk or interaction can depend not so much on consensus or agreement but on what each participant thinks is a desirable or necessary goal and what specific pragmatic strategies and frames are useful for achieving those goals. As we know, not all participants necessarily have the same goals. It’s a question of relevance and for whom. RT has important implications for Bakhtinian dialogism as a way of thinking about pragmatic strategies, footing, framing, and context. Dialogism doesn’t presume cooperation in speech interactions, but it doesn’t rule it out either. Rather, says Bakhtin, language as lived experience is fundamentally dialogic because language literally moves around among speakers and the subject positions they inhabit. Dialogism encompasses both centripetal and centrifugal forces in language and society. Bakhtin’s theory of dialogism suggests how relevance theory might be incorporated into a critical pragmatic analysis of discourse and Kempe’s Book in particular. Dialogical discourse

Margery Kempe’s Str ategic Vague L anguage

209

may lead to consensus or agreement or cooperation, but it might constitute social struggle, disagreement, and difference. It all depends on the relevant frame and footing of the participants. RT also presents a different take on equivocatio and potential meaning. We have seen that resemanticization or equivocation was a key strategy in medieval texts with pragmatic orientations, but problematic as speech according to many twelfth- and thirteenth-century logicians. Kempe uses equivocation, relevance, vague language, and other pragmatic strategies in her interactions at York, but we can better grasp her metapragmatic awareness if we distinguish vague language and equivocatio. Vague language is semantically underspecified, inexplicit, and sometimes empty (Channell 1994: 34-38). It differs from ambiguous language or polysemy. Ambiguous or equivocal language is speech which has two or more possible meanings attached to it and therefore two or more relevant contexts, but the listener or reader can’t straightforwardly determine which meaning is appropriate or if both are appropriate in different contexts. Ambiguous language is based on a stable or neutral frame, a benchmark for determining meaning, which is why many logicians, medieval and modern, discuss equivocation as a fallacy in need of disambiguation. Vague language works differently (cf. Weinrich 1966: 411; Channell 1994: 35-36). Vague language is slippery, fluid, and can’t be easily disambiguated. According to Weinrich, vague language is characteristic of most conventional language, that is, the sort of language most people use every day. Vague language is part of the life of language. It is also related to fuzzy semantics, whereby word meanings are stored by speakers in prototype ‘folders’ (cf. Rosch 1975; Aitchison 1987: 51-62). Because vague language is more indeterminate and underspecified in contexts, speakers using vague language can seem more suspect or more threatening, especially when an examiner is seeking detailed information or when social standards of conformity, sincerity, and intent are at issue. The exemplary medieval situation for pragmatic questions of equivocatio and vague language is the heresy examination. However, vague language is not always suspect. In some contexts, it’s actually considered to be helpful, effective, efficient, for example, in everyday colloquial speech and diplomatic negotiations. ‘I have to get my books and stuff before we leave’ is a quick way to express my intention without listing everything I need to do before leaving the house. Vague extenders (‘my clothes and stuff’) and approximators (‘Jason is almost six feet tall.’ My emphasis.) allow for expanded nonspecific reference depending on how the listener understands the context and what doesn’t require exact measurement or more detail. In difficult dialogues vague language can

210 

The Medieval Life of L anguage

sometimes bring people with strong opposing views to agreement when they allow the vagueness to cover a degree of disagreement. Vague language can create a more inclusive social space or write acceptable contract language which enables the parties to agree changes when the circumstances change. Sometimes, vague terms enable an audience to more easily buy into, adhere to, the speaker’s interpreted meaning or may allow the speaker to assert something without providing more detailed specifics. In a news broadcast or report, the sentence ‘Yesterday, Saudi planes hit the Houthi rebels with tons of bombs’ may be sufficient and effective enough to describe a military assault for most general information purposes. Vague language usually relies on audiences to use background knowledge to construct more detailed networked meanings and worlds which give coherence, significance, and concreteness to experiences narrated or represented in texts. However, especially in hostile or adversarial situations, vague language can be evaluated negatively. We saw this in Arundel’s responses to Thorpe’s speech. From the standpoint of judicial authority, vague or evasive language is suspect, the kind of speech people accused of crimes or heresy (or steroid use) are likely to use as a tool to survive or escape. Examinations of suspected heretics are inherently asymmetrical, and the person accused or examined might decide evasive or vague responses are a better strategy than outright opposition or compliance or cooperation, whether or not they are guilty of what they are suspected of. Vague language may even be a strategy of resistance, as both Thorpe’s Testimony and Gui’s Practica narratives suggest. Because the power to decide what is and isn’t an acceptable response is all on the examiner’s side, a speaker using vague language can be trying to maintain her discursive agency or footing or trying to negotiate a polite or less face-threatening form of dissent. However, such pragmatic strategies are risky. If the questioner perceives me to be answering vaguely, I may well be judged uncooperative and suspected even more of whatever it is I am suspected of. At the center of medieval pragmatics are perceptions and signifiers of sincerity, truth-telling, and relevance. As a category of discourse, vague language is a moving target whose perceived function and value depend on relations of power, authority, sincerity, and perceived truth-telling as well as on participants’ different expectations for cooperation or survival in particular situations. Vague language depends on relevant contexts. As such, vague language interrupts Grice’s claims that reasonable, cooperative speech is the prerequisite for communication. Even more than equivocation, vague language destabilizes the interpretive univocity of discursive relevance. Interpreting relevance for frame, footing, evidence, and power is in constant tension with textual dialogism.

Margery Kempe’s Str ategic Vague L anguage

211

Vaguing Pragmatics The Book of Margery Kempe foregrounds vague language and other pragmatic strategies as part of Kempe’s discursive practice and metapragmatic awareness. Kempe’s text, like Thorpe’s Testimony, differs from many other late medieval accounts of heresy inquiries, which were composed from the viewpoint of the authorities or their supporters. Near-contemporary transcriptions and records of heresy trials in Norwich (Norwich Heresy Trials, ed. Tanner 1977), Salisbury, and elsewhere give us names, dates, places, charges, and some sociocultural details of the interactions, but those texts often follow pre-established legal formulas for making written records (usually in Latin) of trial discourse exchanges (often in vernacular) rather than record the fluid interactions between examiners and the suspects or accused. The Norwich records of Lollard investigations and the 1498 trial of the suspected Lollard William Ramsbury (see Chapter 5) are formulaic texts controlled by institutional authorities to codify transgressions and punishments and maintain archival knowledge against which to judge individuals’ subsequent behaviors. The recorded individual confessions and judgments constitute hyperliterate archiving, fitting individual content into institutionally sanctioned knowledge templates. These formulaic records help us ascertain who was charged with what, the evidence presented, and when and where a particular arrest or inquiry or confession took place. But as formulaic texts they do not even try to convey the give and take of spoken exchanges. They are judicial records of what the interrogators were looking for, what they believed they discovered, what answers in the vernacular they received to the leading questions they asked, the answers recorded in Latin, and what the examiners decided to do after the questioning. The Book of Margery Kempe, like Thorpe’s Testimony, is a different representation of discourse and speech pragmatics, more reenactment than record. Reading a text like Kempe’s with CA tools, we need to acknowledge the mediated and after-the-fact status of the written text. Still, as we shall see, Kempe’s text presents a vivid narrative of her interactions with Archbishop Henry Bowet (installed as Archbishop of York in 1407, d. 1423) and his court. She uses vague, ambiguous, and polysemic language to maintain her positive and negative face and seize discursive control in explicitly hostile religious and legal situations. Kempe’s narrative at once justifies her beliefs and behavior and displays her complex, sophisticated pragmatic and metapragmatic understanding of public discourse. Kempe’s prose narrative emerges in an ambiguous (equivocal?) zone between fact and fiction, neither an exact representation of actual events

212 

The Medieval Life of L anguage

or a transcription of speech nor entirely made up. The text produces a discursively and pragmatically complex narrative based on at least one eye-witness account, however much that witnessing I is self-interested and mediated through collaborative scribal composition. These textual conditions make the Book of Margery Kempe even more a “novelistic” and “dialogic” text (Bakhtin 1994: 98-122), multivoiced and materializing from multiple speaking positions and deictic stances. In terms of pragmatics and relevance, the text foregrounds Kempe’s narrating and narrated subjectivities and reveals her complex pragmatic understanding of social interactions and discursive agency. Elsewhere, I have discussed Kempe’s moans, cries, and tears as affective utterances with interjective force (Amsler 2018). Those natural or paralinguistic vocalizations in devotional or sacramental situations are related but different to her more prevalent speech in the narrative. In this chapter I read the York episode in terms of Kempe’s colloquial strategic pragmatics. The text clearly presents Kempe as metapragmatically aware of and adept at using two important but potentially dangerous pragmatic strategies: 1) negotiating meaning and maintaining discursive control in the face of coercion and 2) using textual hermeneutics and vague or equivocal language in oral exchanges to rebut accusations, resist challenges, or express her own ideological or religious viewpoint. In addition to pragmatics, the narrative in Chapters 50 through 52 also reveals the connection between an oral inquiry about Kempe’s orthodoxy and broader medieval discursive conflicts over literacy and textual authority. Oral and written authority are entangled in public speech. The collaboratively written Book of Margery Kempe was intended to circulate Kempe’s oral exchanges, beliefs, and religious expression within a wider, if indeterminate, textual field alongside other women’s religious texts. The text puts Kempe’s pragmatic strategies and metapragmatic awareness on full display in the narrative of her hostile examination at the Archbishop of York’s court.

Kempe Comes to the Archbishop Kempe arrived in York in early autumn 1417, after returning from her pilgrimage to Santiago de Compostela. A local anchoress who had previously thought well of Kempe would no longer receive her “for sche had herd telde so mech evyl telde of hir” (ll. 2810-2811). While others in York were more welcoming, the narrative represents Kempe’s visit to York Minster as contentious from the start: “And also sche had many enmyis whech slawndryd hir, scornyd

Margery Kempe’s Str ategic Vague L anguage

213

hir, and despysed hir …” (ll. 2828-2829). Chapters 50 to 52 (224 lines in Staley’s edition) then narrate a series of encounters in York which become increasingly hostile and threatening for her. When Kempe enters the cathedral grounds, she is immediately interrogated by several canons. In these exchanges, she uses evasive, vague, and ambiguous language to try and gain control of the discursive floor and, I believe, protect her security, freedom of movement, and faith. Rather than cooperate with her interrogators or avoid threatening their positive and negative face, she openly and intentionally resists their efforts to interpellate her. When we read the conversation maneuvers in slow motion using CA and explore the relevant micro contexts, we can understand how Kempe’s strategies sustain her agency and why pragmatics and pragmatic awareness matter in her narrative. Her first encounter at York Minster is overtly violent and dangerous. A canon grabs her by the collar of her white gown and asks, “Thu wolf, what is this cloth that thu hast on?” (ll. 2830-2831). Exercising his institutional authority to judge lay people’s religious behavior but perhaps overstepping his authority to lay hands on a person, the canon addresses Kempe with the inferior’s pronoun thu [þu] and challenges her right to wear white clothing as a sign of her spiritual condition. The canon’s pointed epithet (“wolf”), with its implicature of deceiver or imposter, and his use of a WH information-seeking question without force, a rhetorical question, imply his hostile stance. Kempe doesn’t answer him, but some children, presumably choristers walking with the canon, do: “Ser, it is wulle” (l. 2833). The children’s innocent referential and literal answer to the canon’s hostile question with its implied nonliteral referent destabilizes the relevance frame in the episode. The children’s innocent response also contextualizes the next interaction between Kempe and the canon and introduces an alternative response to the canon’s rhetorical question. The children’s answer challenges, however unintentionally, the canon’s aggressive face threat to Kempe and implicitly exposes his nonliteral, hostile intent. However, the canon ignores the children – he couldn’t care less what they say – and reiterates the initial relevance context. He continues to question Kempe and insults her when she doesn’t respond to his question herself. The canon becomes more insulting (“gan to sweryn many gret othis”; l. 2834) because she doesn’t cooperate with his intention to interpellate her into his authority. When Kempe does finally respond, she addresses not the content of the canon’s question but the manner of his speech: “Ser, ye [ȝe] schulde kepe the comawndmentys of God and not sweryn so necgligently as ye [ȝe] do” (ll. 2835-2836). She also models speech propriety by returning polite

214 

The Medieval Life of L anguage

[ȝe] to the canon’s superior [þu], and she hedges by saying he is morally negligent rather than sinful. From the moment the episode begins, individual utterances are readdressed and displaced by other utterances. The children’s, the canon’s, Kempe’s speech, all are pragmatically indirect. Kempe returns a metajudgment of the canon’s discourse to his referential judgment of her dress, and she implicitly accuses him of setting a bad example for the children. The canon regards the children’s literal reply as irrelevant, but Kempe creates a new relevant context by incorporating the children’s speech into the conversational situation and supporting them as participants in the exchange. The canon speaks as if only he and Kempe are involved in the exchange, but she recontextualizes the interaction by acknowledging the children’s active presence. She aff irms their social and pragmatic relevance as auditors who are influenced positively or negatively by how adults talk. Maintaining his socially sanctioned position as Kempe’s superior and authority in spiritual matters, the canon’s next question directly acknowledges Kempe’s voice and addresses her appeal to the “comawndmentys”: “[he] askyd hir hoo kept the comawndmentys.” By repeating the word “comawndmentys,” he makes his question explicitly relevant to the new topic Kempe has introduced. Kempe replies, “Ser, thei that kepyn hem” (ll. 2836-2837). Kempe’s circular or echo response – it’s really a tautology – maintains their exchange as talk about a generic referent, “thei.” Attempting to corner Kempe with accusatory questions, the canon then presses his attack with a direct address and asserts his pronominal power: “Than seyd he, ‘Kepyst thu hem?’” (l. 2837). This time, Kempe accepts the canon’s direct address, but again she evades the canon’s sting by expanding the implication of his question: “Syr, it is my wille to kepyn hem, for I am bownde therto, and so ar ye [ȝe] and every man that wil be savyd at the last” (ll. 2838-2839). Kempe’s strategy of embedding syntax emphasizes intention (“wille”) and reaccents the canon’s hostile speech. She returns their talk’s referent to the generic level (“every man”) and once again evades the canon’s hostile questions. By asserting that she, the canon, and all Christians are judged by God’s law, Kempe avoids making any specific claims about herself and her piety, although that is precisely what the canon wants her to do in his gotcha game. She openly criticizes the canon for publicly swearing but steadfastly refuses to directly attribute any sanctity or righteousness to herself. Instead, she uses vague language and circular or echoic reference – those who keep the commandments are those who keep the commandments – and emphatically affirms her

Margery Kempe’s Str ategic Vague L anguage

215

intention to keep Christ’s commandments rather than declare she in fact does so. God, she implies, will be the judge of that. This initial exchange with a hostile York canon reveals three important pragmatic strategies characteristic of Kempe’s public discourse in difficult or dangerous situations. First, the tension in the exchange, as in her other York encounters, is manifested within a pragmatic code of politeness and presupposed social deference. Earlier, we saw how Thorpe regularly uses the polite and stratified T/V convention for Middle English pronouns of address (þu/ȝe) when interacting with Archbishop Arundel. Similarly, throughout the York episode, Kempe and her interlocutors mostly maintain the conventional T/V structure of Middle English pronouns for acknowledging and constituting social status in spoken or written conversation. In the canon’s initial question, þu marks his social superiority and connotes condescension toward the addressee. Kempe, who would have been considered inferior in social status to the clerics and Archbishop despite her family’s relatively elite secular status in Bishops Lynn (cf. Jurowski 2007), consistently returns ȝe to the canon’s þu. She mitigates her hostility-risking use of strategic vague or indirect language with conventionally polite forms of address (cf. Bergs 2004). Her second pragmatic strategy has to do with predication. Both Kempe and the canon respond to one another by extracting their questions or responses from the lexis of the previous speaker (e.g., “comawndmentys”). While this strategy might seem to make the speakers’ utterances cohesive and relevant to the topic, that is not all that’s going on. Rather, the exchange presents a struggle for relevance. Kempe’s reply to the canon constructs a tautology from the open-ended referentiality of information-seeking questions. Both third-person pronouns (hoo, thei) are contextually anchored by the same predicate (VP), kepyn the comawndmentys. Kempe’s vague responses contrast with the children’s earlier literal reply to the canon’s rhetorical, that is, non-information seeking question, “what is this cloth? …” When Kempe echoes and then takes control of the canon’s question (“Kepyst thu hem?”), she reestablishes her footing by embedding the original question as a statement within a matrix clause of intention (“it is my wille to X”). Kempe’s pragmatic reaccenting of the canon’s speech includes not only vague language and equivocatio but also clausal embedding and recontextualization. Gui described how the Waldensians he and others examined used the same strategy. Kempe’s third pragmatic strategy in her interaction with the canon, related to the second, is to ameliorate or mitigate truth claims. Rather than assert a fact in the absolute (I kepyn the comawndmentys), Kempe shifts the modality implied by the canon’s hostile question to an assertion

216 

The Medieval Life of L anguage

of her intention to do so (“it is my wille to kepyn hem”). The canon wants her to assert a truth claim, which he can then pick apart to question her spiritual condition. But by asserting her intent (“wille”) to do something, Kempe avoids making a truth claim about what she has in fact done or does. The narrative affirms Kempe’s voluntarist position with respect to speech and intention while leaving the evaluation of her actions to God alone. In speaking this way, Kempe implicitly rejects the canon’s authority to judge her behavior, even as she responds to him with the structural politeness, explicit deference, and cohesive repetition considered socially conventional in asymmetrical speech interactions in English at the time. Kempe and the children’s encounter with the first York canon violates several conversation principles and maxims as prescribed in much contemporary pragmatics: cooperative sociability, rational goals, and consensus relevance. Neither Kempe nor the canon is speaking cooperatively. Without intending to do so, the children also violate Grice’s Cooperative Principle with their referential literalism and conversation intervention. Kempe violates the Cooperative Principle and Quantity Maxim when she fails to provide the kind of information the canon demands and when she evades responding directly to the canon’s questions. Her evasive language and echo responses deflect his institutional or conventional authority to question her in spiritual matters. Kempe manufactures the relevance of her echoic speech by appropriating and reaccenting the other’s speech. According to Sperber and Wilson (1995), “echoic” speech achieves discursive relevance by interpreting another utterance and expressing some propositional attitude towards it (occasionally endorsement, but often irony, doubt, ridicule). In Kempe’s case, echoic speech works somewhat differently. She shifts the syntactic modality from assertion to intent and so evades making a truth claim about her spiritual condition. In her vocal exchanges at York Kempe also interrupts some late medieval socially constructed conventions of relevant context or situated meaning. Her responses as represented in the narrative, which I take to be a faithful version of what Kempe thinks has occurred, do not fulfill the canon’s request for information. Kempe doesn’t misunderstand what the canon is asking, but she doesn’t always explicitly disagree with him either. Rather, in her speech, she intentionally violates Grice’s pragmatic Maxims of Relevance, Quality, and Quantity in order to exploit the vagueness inherent in the canon’s language and seize control of the conversation. She evades his accusations but refrains from making specific counterclaims. She remains or tries to remain within the negotiated bounds of social politeness and orthodoxy. The trouble is, such evasion and noncooperation are received

Margery Kempe’s Str ategic Vague L anguage

217

by the accusing canon as disagreement and possibly as evidence of her spiritual transgression. However, because Kempe’s replies are pragmatically indirect, it’s difficult to find an assertion in her speech. The text also exploits the children’s role in this initial episode at York. Their reply to the canon, rather than being extraneous in the narrative, functions as the encounter’s pragmatic barometer. The canon clearly establishes negative affect and threatens Kempe’s face when he grabs her collar, and he interprets Kempe’s responses as evasive. But the children’s innocent and literal reply frames the entire speech event in a different relevant context. The new frame associates (but doesn’t identify) Kempe’s violations of conversation cooperation with innocent literalism rather than with communicative perversity or insincerity. As an alternative narrative frame, the children’s voices constitute a material, literal, denotative utterance and a direct, if uninvited, response to the canon’s question. Their motivation cannot be ascribed to heretical reading, insincerity, or evasive language. The text implies the children interpret the situation not as hostile but according to their pragmatic understanding and their sincere intention to provide the canon with the information his question seeks. But based on the immediate context, we can infer that Kempe sincerely intends not to cooperate or respond directly to the canon’s questions. From the canon’s openly hostile viewpoint, the children’s literal response is irrelevant while Kempe’s are evasive and insincere. In this first encounter at York, all the participants struggle, sometimes unintentionally, over which discourse norms and presuppositions are appropriate, relevant, or authorized. Kempe’s initial encounter at York also raises a question as to what constitutes a linguistic community and pragmatic good faith. Medieval grammarians and philosophers and contemporary philosophers and pragmatists alike appeal to community norms of signification and conventional forms of expression which make communicative talk effective. Olivi argued that a linguistic community is formed around speech used “habitualiter” [habitually, conventionally] and what we can call consensus signification and community-accepted impositions of words on material and immaterial referents. The community of speakers, says Olivi, depends on an agreedupon semantics (communi impositiones) confirmed by the collective will of the community. Olivi and some other late medieval grammarians and philosophers described a functioning linguistic community primarily in terms of conventions of reference and signification and much less in terms of modal or embedded syntax, modes of address, potential meanings actualized in context, or language choice. As we saw in Chapter 1, Olivi characterized flouts and re-imposition as acceptable within a discursive

218 

The Medieval Life of L anguage

community by arguing that some changes in speech norms are acceptable and can be deployed in good faith so long as the speaker alerts her audience to what she is doing in a specific relevant context. No single speaker or writer should violate those norms “unless the speaker [or writer] clearly expresses to the listener her intention [to do so]” (nisi suam intentionem clare exprimat auditori; Quid ponat ius, in Rosier-Catach 2004: 180-81.). Similarly, Grice presumes a speaker’s good faith and therefore lays responsibility for interpreting flouts of the Cooperative Principle and maxims on the listener. In his quodlibetical writing, Olivi explicitly locates the speaker’s intention to alert the audience to a changed relevant context “in an act or performance (in actu aut exercitio), as it occurs when [the name] is actually uttered out of an intention to signify or when it is written in a certain narration (narratione)” (Olivi, QL, q.7; Brown 1986: 353). For Olivi, intentionally deviating from community norms of signification and referential response requires that a speaker produce an explicit justification in order to maintain good faith with the community. From a conventional viewpoint, acceptable speech is a condition for continued community membership. Olivi’s voluntarist orientation to pragmatics means that habitual contexts constrain collective meanings, and participants are generally expected to produce utterances which abide by those signifying dispositions in specific contexts unless they announce their intentions to do otherwise, thereby showing communicative intention and good faith toward their audiences. Olivi’s ideal speaker is an explicit deformer, a creator, not an evader or a liar. Kempe, however, takes linguistic pragmatics and voluntarism in a different direction. Speaking with the first York canon she performs and implies (implicit) rather than expresses (explicit) her intention not to comply with the canon’s questions. She and the canon would seem to participate in the same broad linguistic community. They use the same code (English), syntax, vocabulary, and for the most part, conversation rules. What distinguishes them is the power asymmetry of the speech situation, and that power asymmetry is what Kempe disrupts when she refuses to cooperate with the canon’s authority to question her on her individual religious practice and spiritual condition. Kempe and the canon struggle (“jangelyd”) over discursive footing, pragmatic strategies, relevance, and discursive control throughout their talk. Their exchange ends abruptly when the canon is said to go “awey prevyly er sche was war, that sche wist not wher he becam” (ll. 2830-2840). Kempe’s first hostile episode at York is immediately followed by a second (ll. 2841-2847). This time, she is interrogated by a “gret clerke … askyng thes wordys how thei schuld ben undirstondyn, ‘Crescite et multiplicamini’ [Genesis 1:22]” (ll. 2841-2842). The epithet “gret clerke” suggests that more

Margery Kempe’s Str ategic Vague L anguage

219

important clerics at the Archbishop’s court were becoming involved with Kempe the longer she remains in the minster. The second cleric’s conversation opening introduces a new topic in the York episode: the theme of literacy and textual pragmatics. Readers no doubt recognize the Biblical passage he cites as the same one the Wife of Bath controversially discusses in her Prologue. Perhaps the “gret clerke” is sincerely seeking to understand how Kempe understands the Old Testament text. Perhaps not. In the fourteenth and fifteenth centuries, the verse was a touchstone in clerical and lay cultures because of its special relevance for debates about marriage, divorce, and sexuality. Anne Hudson suggests the verse was used in Kempe’s informal examination “to detect in her any leanings towards the beliefs of the Free Spirit” (1985: 114). The Brethren of the Free Spirit were a loosely affiliated group of nonconformists based on the Continent and among Beguines. The loosely associated group’s moniker Free Spirits was derived from the Compilatio de novo spiritu collated by Albert the Great in the mid 1270s as part of an investigation of suspected heretics in lower Germany. According to the Compilatio, the suspects rejected transubstantiation and clerical authority, claimed the perfected soul in union with God cannot sin, and used especially erotic language to describe union with God. In some people’s minds, the Free Spirits were associated with the mystics Marguerite Porete (c. 1250-1310) and Meister Eckhart (c. 1260-c. 1328; cf. Leff 1967, 1: 308-407). Church authorities claimed that suspected heretics used the Genesis passage to justify sexual immorality. During Kempe’s examination at York, however, the cleric’s citation of Genesis 1:22 in Latin is more directly related to the question of scriptural interpretation and lay literacy. Perhaps he’s asking, in the aftermath of Arundel’s 1407 Constitutions: Does the secular Kempe know Latin scripture? Elsewhere in the narrative, Kempe is shown to be somewhat familiar with scripture because she reads and discusses texts relating the lives of pious women and because she regularly seeks out public preachers. Or perhaps the York cleric is attempting to one-up Kempe by displaying his literate superiority: How can a laywoman know what the Latin verse really means? When constructing such micro contexts for the cleric’s and Kempe’s speech intentions, we need to keep in mind that earlier Arundel himself had examined Kempe and approved her way of life. He allowed her to choose her own confessor and gave permission for her to receive the sacrament every Sunday (ll. 818-847). In the York episode it’s more likely some of the clerics at the Archbishop’s court are attempting to trip Kempe up by outing her knowledge of scripture and textual hermeneutics and putting pressure on her to recognize the Latin text.

220 

The Medieval Life of L anguage

Kempe’s response to the “gret clerke”’s challenge that she interpret the Latin text illustrates not only her ability to interpret scripture but also her sophisticated pragmatic strategies in hostile situations. In her response she indicates some understanding of Latin and exegesis, and she exploits the ambiguity of the intransitive verb construction in the Latin scriptural text. After successfully evading the first York canon’s questions, this time Kempe responds directly to the “gret clerke”’s questioning. Kempe does not say she can’t understand Latin (as she had claimed earlier in Leicester; see ll. 2650-2651), nor does she ask the cleric to translate or paraphrase the scriptural verse into English. Pragmatically, she first seems to cooperate with the “gret clerke” by commenting directly on the passage. Quickly, however, she exploits the vague, not equivocal, language of the cleric’s question, How should the words be understood (“undirstondyn”)?, by metacommenting on the question and saying that a proper reading of scripture includes both literal and spiritual interpretations. Kempe’s hermeneutic gloss establishes the relevance of both a literal reading (recall the young choristers’ reply) and a non-literal reading of the passage. She embeds her response in a not only … but also construction. The scriptural words, she says, should be understood literally “of begetyng of children bodily” and spiritually (figuratively) “be purchasing of vertu, whech is frute gostly” (ll. 2843-2844). Kempe’s parallel clauses subtly foreground the hermeneutic complexity of understanding any scriptural text according to Christian exegesis. Her literal reading of Genesis 1:22 emphasizes the semantics of material reference and the orthodox view that sex is for procreation. Her spiritual interpretation relies on the broad metaphor of children and virtues as one’s “frute.” Kempe then seizes the interpretive floor and asserts her view of piety by exploiting equivocatio in scripture. She multiplies the possible referents of “frute” and the ways one can obtain virtue: “be heryng of the wordys of God, be good exampyl gevyng, be mekenes and paciens, charite and chastite, and swech other, for pacyens is more worthy than myraclys werkyng” (ll. 2844-2846. My emphasis.). The gerunds and nouns in her reply implicate proper spiritual behavior for both ecclesiastics and lay people. Except one. By excepting “myraclys werkyng,” nowhere implied by the scriptural text, Kempe situates her understanding of Christian virtue with voluntary good works, right living, and listening to sermons, not clerical privilege. Kempe’s strategic use of the vague extender “and swech other” expands the referential space of her reply and allows her and the “gret clerke” to add, share, or debate other examples of how to achieve virtue through pious acts and dispositions. Like her manipulation of vague language and equivocatio, Kempe’s use of pragmatic extenders incorporates, in fact contains, her hostile

Margery Kempe’s Str ategic Vague L anguage

221

questioner within her immediate speech context. Thus, Kempe is able to regain control of the discursive domain, even if only for the duration of the immediate talk. With her extended list of material and spiritual “frute,” she explicitly subordinates the controversial question of miracle working to general Christian virtues applicable to clergy and laity alike. By exploiting the semantic vagueness in the “gret clerke”’s language (whether or not that’s what he intended), Kempe inserts her own voice and voluntarist beliefs regarding faith, piety, and scriptural meaning into fifteenth-century debates that engaged clergy and laity about spiritual virtues, the role of clergy, and scriptural access. As recounted elsewhere in the Book, Kempe became familiar with strategies of Scriptural Hermeneutics 101 from her private discussions with at least two clerics, from the small archive of spiritual texts they read together, and from her active listening to sermons and religious public discourse. Preaching friars, monks, and clergy used these interpretive strategies to produce varied and elaborated meanings from scriptural texts, and Kempe, like other pious lay people, paid close attention to these reading strategies. While Kempe’s interpretation of the Genesis 1:22 passage is orthodox and mainstream, she sets her interpretation in a multilingual context to control the immediate discourse at York and to construct and maintain her face and public authority to speak about God. She quickly mobilizes literal and figurative semantics to gloss Latin nouns and verbs in English. Using a standard medieval polysemic hermeneutics – literal (‘literaliter’) / gostly (‘spiritualiter,’ figurative) – she pragmatically shifts her position or footing when responding to the “gret clerke” from that of an inferior on trial, a lay woman being examined, to a more equal speaking position within the linguistic and interpretive community presupposed by the York clerics. Kempe strategically violates what the York clerics might consider maxims of Quality and Quantity when her scriptural interpretation exceeds the cleric’s question and recontextualizes or reaccents his hostile opening rather than cooperatively responding to it. In her hostile encounter with the “gret clerke,” Kempe reverses the pragmatic strategies she used earlier with the first York canon. Arriving at York Minster and attacked by the first canon, she had used the literal understanding of speech as resistance and tautological syntactic echoic responses about “the comawndmentys” and pious behavior as evasion. Now, in this second encounter, Kempe adheres to the more widely relevant context of interpreting scripture spiritually as advocated by the dominant hegemonic religious community, represented by the “gret clerke.” In De signis, Bacon had shown that re-impositio functions pragmatically and semiotically in several ways. Reimposition can construct

222 

The Medieval Life of L anguage

and reconstruct literal and figurative meanings of language as part of mobile polysemic signification. It can also be deployed, amplified, or restricted by speakers as they construct and respond to the speech situation. In her response to the “gret clerke,” Kempe exploits both options. In the second encounter at York, Kempe successfully asserts discursive relevance in direct relation to the “gret clerke.” As I said, her response to the cleric’s question about the Latin scripture passage implies she presents herself as, that is, adopts the footing of an equal reader of scripture. When she finishes, the “gret clerke,” unlike the canon, says he is “wel plesyd” with her interpretation (l. 2847). Kempe manages to defuse the hostility which initiates the second exchange by publicly demonstrating her interpretive competence and doctrinal appropriateness within the relevant discursive frame offered. I imagine the York clerics, having suspected Kempe to be a troublesome nonconformist, expected her to interpret the scripture text literally or without sophistication, but she skillfully avoids that trap. The exchange between Kempe and the “gret clerke,” then, represents in the text an ideal version of how fifteenth-century clergy and laity could interact almost as Christian equals in an everyday, multilingual setting with basic Latin text and oral vernacular interpretation (cf. Amsler 2012). Interestingly, once the “gret clerke” approves of her interpretation of Genesis 1:22, the other potentially difficult fact is dropped altogether. As presented in the narrative, Kempe seems to understand and interpret the cleric’s Latin citation without translation. Read together, Kempe’s first two interactions with York clergy reveal her complex pragmatics of vague and polysemic language, discursive “jangle”-ing, echoic embedding, and dialogic struggles for relevance and footing. She deploys syntactic and semantic literalism and echo response (tautology) to evade the first canon’s hostile questions. She evades the “gret clerke”’s potential trap by adopting the consensus relevant strategy of spiritual interpretation to read Latin scripture as a guide for Christian behaviors. In her second exchange, Kempe discards the innocent literalism of the York choristers in favor of more mainstream clerical reading, one which her immediate audience receives as competent and conventional. In both interactions, she avoids making truth claims about her own spiritual condition. Taken together, the strategies of innocent literalism and adept scriptural exegesis sidestep the loaded and hostile questions of the York authorities. At the same time, Kempe’s speech exploits the fuzzy boundaries of vague language and equivocal nouns and verbs in Latin and English. Her subtle pragmatic use of syntactic embedding, like that of Gui’s Waldensians, enables her to reaccent her interlocutors’ assertions or implicatures and incorporate them into her

Margery Kempe’s Str ategic Vague L anguage

223

own intentions. As speech actions within a speech event, Kempe’s pragmatic strategies implicitly destabilize negative characterizations of lay women and heretics as overly literal, unsophisticated, dangerous, or unruly readers of scripture. The narrative of the first two York encounters represents Kempe as proposing different kinds of discursive relevance in different situations and successfully evading hostile questioners whose face-threatening acts were probably designed to confound or expose her as a pious fraud or worse. By adjusting for different relevance frames in specific speech contexts, Kempe performs a different kind of RT pragmatics and social interaction to those which rely on the ideal speech situation, ‘rational’ communicative cooperation, or maximum cognitive processing in a common relevance context. In disagreements or hostile, potentially dangerous, or face-threatening conversations, proposing and enforcing a particular relevance frame can enable the more socially powerful participant to assert and control the talk agenda and what is ‘truth.’ Difficult as it may be, changing the relevance context or frame can enable a less socially powerful or alternate hegemonic participant to evade the control or authority imposed by the more powerful participant. Alternate relevance frames can also shift the presuppositions and semantics of the interaction so as to constitute more equal participation or a different knowledge context. Power is a presupposition or a predisposition of discourse and context from within the dominant hegemonic order. In practice, however, discursive power and agency can circulate, sometimes quickly, even if temporarily, among participants and situations. Despite the structured power of institutions, class, and gender, some power relations can and do change within dialogic speech events. Pragmatic strategies which enhance footing or agency or reassert relevance remake power relations and meaning within talk. In Kempe’s case, maintaining her footing and relevance is key to asserting her will in social and religious talk. Kempe’s third hostile exchange at York begins with her summons to answer questions in the chapterhouse (ll. 2858-2859) and concludes with her questioning before the Archbishop of York, Henry Bowet, at Cawood Palace, Yorkshire. Summoning her to answer, a cleric implies she has not acted according to her promise: “Damsel, thu seydest whan thu come first hedyr that thu woldyst abydyn her but fourteen days” (ll. 2855-2856). In reply, Kempe exploits the phrase “but fourteen days” as vague or unclear, and so she controls the conversation floor. She boldly and metapragmatically asserts that her reply was vague on purpose: “Ya, ser, wyth your leve, I seyde that I wolde abydyn her fourteen days, but I seyd not that I schulde neithyr abydyn mor her ne les. But as now, ser, I telle yow trewly I go not yet” (ll. 2856-2858). She says decisively: Yeah, more or less, when I’m ready.

224 

The Medieval Life of L anguage

Despite such face-saving and assertive utterances, Kempe remains in a hostile situation. She does as she is ordered to do: appear before the clerics in the York chapterhouse. Not all modern readers think Kempe is as threatened as the text makes out (e.g., Arnold 2004: 82-88). Kempe’s behavior suggests she is. The chapterhouse episode represents a turning point in the narrative, as Kempe begins to attract supporters among the clergy and the townspeople. During her questioning (ll. 2873-2884), she successfully rebuts the accusation she is a married woman traveling alone without her husband’s permission, and she answers well questions regarding the Articles of Faith. Anticipating what her examiners have in mind, she also explicitly (baldly on-record) states, “I wil neithyr meynteyn errowr ne heresy,” and if proven to be in error, she offers to amend any of her religious statements “wyth good wille” (ll. 2878-2882). At this narrative moment, the “seculer pepil” and some York clerics begin to overtly support Kempe, asking her to pray for them and preventing her imprisonment while she waits for the Archbishop. In the York chapterhouse episode Kempe is more supported by clergy and lay people than she is shown to be at any other moment during the York episodes. But when Kempe is escorted to Cawood, the hostility surrounding her once again increases. Members of the Archbishop’s court call her a Lollard and a heretic (l. 2914), to which Kempe responds indirectly by once again criticizing the clerics for their swearing and rude, rough talk (ll. 2916-2918). Just as she had done earlier with the first York canon, Kempe tries to shift the discursive topic from her religious beliefs to the clerics’ manner of speaking, hypocrisy, impiety, and impoliteness. When Archbishop Bowet first arrives, he chastizes her for wearing white (as the first York canon did) and orders her put in manacles “for sche was a fals heretyke” (ll. 2923-2925). The implication is that, like Thorpe’s, Kempe’s reputation has preceded her and that Bowet, a known Lollard hunter, has already formed an opinion about her. The Archbishop’s actions and words increase the threat considerably. Rather than using evasive strategies, this time Kempe responds directly and assertively: “I am non heretyke, ne ye shal non preve me” (l. 2926). Despite (or maybe because of) her defiance, she is made to stand alone in terror and dread in a room in Cawood. Signifying Kempe’s danger, the narrative at this moment shifts to a somatic focalization foregrounding Kempe’s fear as she prays to Christ: “hir flesch tremelyd and whakyd wondirly” (l. 2929). When the Archbishop returns, Kempe breaks into a loud, tearful cry, a vocalized release after her long, solitary torment and uncertainty. Taking charge of the examination, Bowet personally tests Kempe on the Articles of Faith. Whereas the first two York examiners seem to have been operating

Margery Kempe’s Str ategic Vague L anguage

225

informally and perhaps independently, the chapterhouse and Cawood examinations constitute a more formal inquiry. In the latter two, Kempe is directly required to declare her faith. In both interactions her responses are explicit, informed, and assertive. In these contexts, Kempe chooses assertion (bald on-record) over evasion as the more relevant strategy. Unlike in the first two York encounters, the narration attributes Kempe’s success in the third encounter to God’s grace (as in Thorpe’s Testimony), which allows her “to answeryn wel and trewly and redily wythowtyn any gret stody so that he [Archbishop] myth not blamyn hir” (ll. 2945-2946). Since the core of both sets of questions in the third examination is the Articles of Faith, the narrative implies that Kempe is speaking under the sign of grace. At this moment in the text, her pragmatic sophistication takes a backseat to Kempe’s wilful assertiveness and orthodoxy in God’s eyes. After her questioning and despite what the other York clerics might have expected, the Archbishop responds positively to Kempe’s statements regarding the Articles of Faith and belief: “Sche knowith hir feyth wel anow. What schal I don wyth hir?” (ll. 2946-2947). Bowet’s judgment of Kempe’s answers and his open-ended question inviting responses from the clergy sounds a bit like Pilate’s reply to the Pharisees after Jesus’ examination. The phrases “wythowtyn any gret stody” and “wel anow” are pragmatically significant. “wythowtyn any gret stody” recalls Thorpe’s similar, perhaps disingenuous description of his responses when he was ‘interviewed’ by Archbishop Arundel. The phrase probably means ‘without preparation’ (MED), that is, extemporaneously or intuitively. With that more everyday semantic context, Kempe is representing herself as answering questions about her faith without special deliberation and potentially without pragmatic evasion strategies. Reading the York narrative with a pragmatics orientation, we realize Kempe is not the only one deploying vague language as a pragmatic strategy. The Archbishop does too when he judges Kempe’s faith with the overtly vague qualifier “wel anow.” The expression enow is well recorded in fifteenth-century English. Malory, for instance, uses inowe to mean ‘enough’ or ‘sufficient’ (1971: 681, 685 689, 720, etc. The Winchester manuscript is dated between 1470 and 1483. See MED for further citations.) The Archbishop, the highest-ranking religious figure in the city, proposes a vague, general criterion for orthodoxy (“wel anow”) in what is clearly a tense, hostile situation. “Wel” serves to strengthen the claim of “anow,” so “wel anow” carries cognitive and evaluative weight. Bowet examines Kempe a second time later in Beverly and again finds “no defawte in hir.” He even provides her with a letter certifying her orthodoxy (ll. 3090-3189). Bowet’s characterization of Kempe’s examination and replies contrasts sharply

226 

The Medieval Life of L anguage

with other medieval narratives of examination of suspected heretics, for example, that by Walter Map, who focuses on suspected heretics’ mistakes, errors, lack of sophistication, or false reasoning. Gui, as we have seen, reads Waldensian speech under examination as more sophisticated. So Bowet’s vague approval of Kempe’s responses may have other motivations. Although Bowet was well recognized as a Lollard persecutor, at the time of Kempe’s examination he was old and in ill health. His vague but positive judgment suggests less his support for Kempe and more his unwillingness to increase the hostility at York. Doctrinally, he doesn’t see a reason to condemn her, but that doesn’t mean he wants her to stay around. Kempe and the Archbishop collectively, but not collaboratively, seek to defuse the hostile and face-threatening discourse which had surrounded Kempe since she arrived in York. But their attempts are conspicuously unsuccessful. When the Archbishop delivers his judgment (ll. 2946-2947), the underlying hostility of the York clergy toward Kempe reemerges. But actually, it never really went away. Neither the Archbishop nor Kempe can completely control most of the clergy’s general hostility toward the woman. But when the clerics complain that Kempe “myth pervertyn” the people (a common rationale among heresy examiners to justify their efforts to identify and remove suspected heretics because they threaten the spiritual health of the community), Bowet seems to reverse course and encourage criticism of Kempe (ll. 2947-2950). So, what Kempe does next to counter the face threats and implicit danger to herself is pragmatically extraordinary and complex, a highpoint in a narrative with many peaks. Despite the Archbishop’s judgment that Kempe’s replies are acceptable (“wel anow”), he picks up on the York clergy’s fear over Kempe’s influence on the local laity and declares in highly emotional terms his dislike for her attitudes and behavior: “Than the Erchebischop seyd unto hir, ‘I am evyl enformyd of the; I her seyn thu art a ryth wikked woman’” (ll. 2950-2951). Based on people’s reports about Kempe’s behavior, the Archbishop orders her not to preach in York but to leave the city. Bowet’s orders and threats at this point imply that regardless of whether or not she’s a heretic, Kempe is a threat to social and religious order. However, rather than go quietly, Kempe stages a confrontation with the Archbishop. Her stance suggests she has gained confidence as her examination has proceeded. As she did earlier with the first York canon, she echoes Bowet’s words in her response and reaccents them in a bald, on-record criticism: “And sche seyd ageyn, ‘Ser, I her seyn that ye arn a wikked man. And, yyf ye ben as wikkyd as men seyn, ye schal nevyr come in hevyn les than

Margery Kempe’s Str ategic Vague L anguage

227

ye amende yow whil ye ben her’” (ll. 2951-2953). Like the Archbishop, Kempe says she hears reports rather than has direct evidence of the Archbishop’s moral state. Pragmatically, she exploits the semantic generality of the insult “wikked” and matches the Archbishop’s order that she leave York with her counterjudgment that Bowet will not be admitted to heaven unless he mends his ways. Kempe rejects implicitly both Bowet’s speech act command (‘Leave’) and its grounds, but she doesn’t say so explicitly. With Kempe’s echoic reply, the Archbishop takes the bait. He “ful boistowsly” asks her what men say about him, to which Kempe replies vaguely, allusively, and perhaps with a dialogic smirk, “Other men, syr, can telle yow wel anow” (l. 2954). From a medieval pragmatic perspective, the textual representation combines speech in the mode of actus conceptus (“ful boistowsly”) with another of Kempe’s echoic responses using Bowet’s own words, the salient “wel anow.” In this highly emotional exchange, we can read the resistance strategy in Kempe’s echoic response and reaccenting of the Archbishop’s words (cf. Bakhtin 1986: 87, 91). Kempe appropriates the Archbishop’s original vague words and reproduces them in a new discursive context with more immediate and concrete relevance. Whereas the Archbishop’s “wel anow” connoted vague approval, Kempe’s echoic return is overtly impertinent. Her echoic response matches the Archbishop’s vague language, and Kempe adopts a distinctly powerful footing in a contested and dialogic context. Pragmatic hedging destabilizes propositional certainty and can function to protect the speaker’s position or claim to certain truth as well as protect the listener’s face. Kempe uses hedging strategically to mitigate the speaker’s obligation or responsibility for the certainty of any claims. The Archbishop wants Kempe out of York, but she hedges her speech to avoid being held accountable and fully subordinate to the Archbishop. By refusing to promise (“sweryn”; l. 2963), Kempe implicitly refuses to leave the city as ordered. She also refuses to agree to the Archbishop’s command that she not “techyn ne chalengyn the pepil in my diocyse” (l. 2964). When she hedges on complying with these orders, Kempe refuses to verbally recognize or cooperate with the Archbishop and his institutional authority. As the York episode repeatedly shows, Kempe’s refusal to cooperate with the York clergy is anchored to the way she exploits vague language and equivocatio, deflects face-threatening acts, and reaccents others’ speech. Her pragmatic strategies serve to qualify or resist institutional authority and conventional or hegemonic meanings and so to seize discursive power and strengthen her footing in hostile disagreement. Kempe is struggling to sustain an alternate world view with immediate consequences.

228 

The Medieval Life of L anguage

The Archbishop’s command that Kempe not teach or preach functionally introduces a new topic in the York episode: lay preaching. Kempe responds to the Archbishop’s order by expanding the topic in a startling way with metapragmatic implications. Based on several accounts of Waldensian and Lollard practices and examinations (e.g., Knighton’s Chronicle, Thorpe’s Testimony), we might expect Kempe to claim as a devout, enthusiastic, sometimes nonconforming Christian that scripture gives her the authority to preach or teach. Instead, Kempe argues she doesn’t preach or teach at all. Her reply is more than a hedge. Kempe reframes the Archbishop’s command by resemanticizing the terms he uses. First, Kempe claims that rather than teach or preach, “I schal spekyn of God … for God almythy forbedith not, ser, that we schal speke of hym” (ll. 2965, 2967-2968. My emphasis.). The first-person plural pronoun includes lay people like Kempe. She claims that any Christian person has a right to “speke of hym,” not just the clergy. She backs up her assertion by reciting a passage from the Gospel of Luke (11:27-28), with a possible allusion to her earlier encounter with the first York canon: And also the gospel makyth mencyon that, whan the woman had herd owr Lord prechyd, sche cam beforn hym wyth a lowde voys and seyd, “Blyssed be the wombe that the bar and the tetys that gaf the sowkyn.” Than owr Lord seyd agen to hir, “Forsothe so ar thei blissed that heryn the word of God and kepyn it.” (ll. 2968-2971)

Comparing the Gospel passage quoted in Kempe’s book with the English version in the Wycliffite English Bible (Later Version) reveals an interesting change in the narrative’s subject orientation: And it was don, whanne he hadde seid these thingis, a womman of the cumpanye reride hir vois, and seide to hym, Blessid be the wombe that bare thee, and blessid be the tetis that thou hast soken. And he seide, But ȝhe blessid be thei, that heren the word of God, and kepen it. (WB/F-M, 4: 185)

Whereas the Wycliffite English Bible orients the narrative to Jesus speaking (“whanne he hadde seid these thingis”), Kempe’s version reorients the passage explicitly to the woman hearing Jesus preaching (“whan the woman had herd owr Lord prechyd”) and then replying to what she has heard. The passage demonstrates Kempe’s familiarity with and ability to recite from memory scripture in English translation, another bone of contention between her and the York clerics. Her knowledge of scripture in part was

Margery Kempe’s Str ategic Vague L anguage

229

acquired by hearing sermons preached. The pragmatic point, whether or not Kempe intentionally adapts the scriptural text to the immediate debate about female preaching, is that the passage she quotes is especially relevant in the discursive context. However, Kempe doesn’t stop there. Shge goes on to make a further claim based on the Gospel passage she quotes, although that further claim does not necessarily follow from what she has just said: “And therfor, sir, me thynkyth that the gospel gevyth me leve to spekyn of God” (ll. 2968-2972). At this moment Kempe has it both ways: she follows up the topic of lay preaching by introducing scripture which narrates Jesus preaching (“herd owr Lord prechyd”) and then claims she is speaking (“spekyn”) about God. By continuing to hedge between preaching and speaking, Kempe adroitly sidesteps the Archbishop’s command that she neither teach nor preach publicly in his diocese. By now in the dialogue, the York clerics can’t contain themselves. Kempe has directly recited and glossed scripture, and she hedges her reference to preaching with the equivocal word speaking. The clerics recognize a potentially heretical hedge when they hear one. According to the text, immediately (“swythe”) a “gret clerke” (the same one?) brings out a book and cites I Corinthians 14:34-35 against Kempe: “that no woman schulde prechyn” (l. 2975). At this narrative moment, the conflict between Kempe and the York clergy is explicitly about lay and female preaching, and that puts the discursive interaction represented in the text squarely within the frame of English Lollardy. Kempe and the clergy are implicitly disputing the difference between speaking and preaching with Gospel and Pauline authorities, and their contest has broader implications for the status of neo-apostolic and scriptural authority and women’s public subjectivities. As the clerics and Kempe match scriptural citations and inferences in a hyperliterate contest, Kempe’s performance implicitly affirms her to be a capable interpreter of scripture. And her interpretive competence causes trouble. The exchange among Kempe, Archbishop Bowet, and the clerics at Cawood complicates the semantics of the term “spekyng” and its hyponyms. The semantic slippage puts into question the power and social status of the one who is said to be “spekyng.” Distinguishing between speaking (hypernym), a general vocal act, and preaching (hyponym), a narrower, institutionally regulated vocal act, Kempe carefully exploits vague language and semantic categories within a hostile speech event. Rather than explicitly opposing preaching to speaking, Kempe uses the more general term speaking as a non-transgressive alternative to preaching, teaching, or admonishing (l.

230 

The Medieval Life of L anguage

2964). In effect, she hedges the semantics, and in so doing, she strategically and pragmatically renders the Archbishop’s command irrelevant in the conversation. In this context, the categorical term “spekyng” is vague, giving Kempe the discursive space to claim she has in fact not usurped the clergy’s authority. Scripture, she says, not the clergy, “gyveth me leve to spekyn of God” (l. 2972). She asserts that the warrant for her assertion is a book, not the clergy, thereby bypassing institutional power. Kempe relies on her utterance’s implicature, the semantic vagueness of a superordinate term, and the generic applicability of the scriptural example to carry her defense. But the power relations in the episode are asymmetric, so Kempe is damned if she does and damned if she doesn’t. When she cites scripture in English to legitimize her distinction between speaking and preaching, she leaves herself open to the charge she is working the clergy’s land without a permit. In fifteenth-century England that is just what many clergy, but fewer and fewer lay people, feared would happen if scripture became more accessible in vernacular translation. When Kempe responds openly to questions about scriptural meaning, her interpretations, their conversational relevance, and her discursive authority, not to mention her stance as a public speaker/preacher, she makes her understanding public and challengeable. Historically, the important point is that Kempe is participating as a pious lay person in a discourse of scriptural interpretation to justify her behavior, and that act can be dangerous or threatening (cf. Watson 1995; Somerset 1998; Kerby-Fulton 2005). Arundel would not approve, but the translators of the Wycliffite English Bible certainly would. In this tense exchange of questions, citations, and accusations, the York clerics try to reassert what they regard as the more relevant context for interpreting scripture and ordinary language. They seek to regain control of the floor and contain Kempe’s speech and actions. They burst out with insults and accuse her of heresy, unauthorized preaching, and demon possession (“sche hath a devyl wythinne hir”; l. 2973). Kempe continues to defend herself and dramatically contest the relevance frame the clerics propose with a combination of scriptural authority, vague language, hedging, and hyponymic distinctions. Kempe caps this performance with a final, empowering, electrifying metapragmatic move. Her speech and actions demonstrate how she marshals pragmatic vagueness with deeper metapragmatic implications for discursive agency. Consensus and agency are not always in sync. In a given situation, different speakers and listeners do not necessarily share the same conventions, meanings, and contexts. In her defense against the Archbishop and clerics’ attacks, Kempe deploys fuzzy semantic category distinctions and

Margery Kempe’s Str ategic Vague L anguage

231

material contexts to impose pragmatic relevance: “I preche not, ser, I come in no pulpytt. I use but comownycacyon and good wordys, and that wil I do whil I leve” (ll. 2975-2977. My emphasis.). As she starts her defense, Kempe carefully hedges her claim by asserting that her speech doesn’t really qualify as the more specialized kind of speaking referred to as preaching. Rather, she says, she is using ordinary speech, the everyday language of vernacular interaction: “I use but comownycacyon and good wordys.” But just as quickly, she drops her hedging and baldly asserts her intention to continue doing what she has been doing while implicitly rejecting the Archbishop’s command: “that wil I do whil I leve.” Kempe’s semantic hedge here is tendentious but effective. Gertz-Robinson thinks Kempe’s teaching and talking constitute de facto preaching, but I don’t think that captures the subtlety of Kempe’s pragmatic maneuvering here (cf. Gertz-Robinson 2005, 2012: 61-71; Craun 2010). Kempe implicitly denies that the hypernym or superordinate term, speaking, necessarily entails the hyponym or subordinate term, preaching. Functionally, Kempe collocates “comownycacyon” with “good wordys” to create a vague term, whose semantic wiggle room she exploits to her rhetorical advantage. The broad Middle English meaning of comownycacyon is ‘conversing’ or nominally ‘a conversation,’ but the MED cites only the Book of Margery Kempe for that meaning. However, we can corroborate Kempe’s use in this passage with Pecock’s near contemporaneous (c. 1445) use in his The Donet: ”Þis teching he schal make openly in pulpyt … and priuely in comunicacioun wiþ persoonys desiring and asking to be tauȝt” (Pecock 1921: 72). For both Kempe and Pecock, the meaning of comownycacyon is broadly social and interpersonal, but it doesn’t directly implicate the official Church’s notion of the public act of preaching and the masculine-marked pulpit as the preaching site. Kempe expands the semantic relevance context of comownycacyon. In this episode Kempe also works metapragmatically by invoking material conditions for pragmatic relevance and shifting the frame to regain her footing. She cannot be preaching, she asserts, because she does not occupy the physical site, a pulpit, which conventionally signifies and stages the act of preaching. This is an extraordinary pragmatic maneuver on her part. Medieval preaching was a fluid, if institutionally sanctioned, speech event. Different material and semiotic markers in different settings were used to assert the preacher’s authority and pragmatic status. The base performance context for medieval preaching was the material sign of a formal pulpit inside a church, raised above the congregants. Another, less formal material sign was the preacher positioned above the audience, for instance, at a window or an upper floor of a barn or inn. Medieval accounts of popular and itinerant

232 

The Medieval Life of L anguage

preachers such as friars often refer to them standing on an ad hoc dais or speaking from a window. Kempe’s familiarity with the conventions of preaching, well documented in her Book, enables her to use the preaching performance code to help defuse some of the hostile charges against her own public speaking. And Kempe was surely familiar with the hexagonal pulpit with panels depicting Latin Patristic Fathers in St. James Church (Castleton Acre, Norfolk) near King’s Lynn (Figure 3). Kempe defends her speaking about God by simply saying she has not spoken or conversed about spiritual or religious matters in any location which can be described as a pulpit. Implicitly, Kempe characterizes preaching in a pulpit as the clergy’s role, whereas she claims to be talking about God the way Jesus did, without a pulpit and on the same level as his listeners. A later Hans Cranach woodcut in Luther and Melanchton’s Passional Christi und Antichristi (1521) depicts Jesus speaking with a group seated around him on the ground (Figure 4). The scene connotes simplicity, humble conversation, and close interpersonal affect. I think Kempe implies a similar set-up for herself when she evokes a favorite trope of fifteenth-century nonconformists and reformers: Jesus did not preach or teach from a pulpit the way contemporary priests do. Implicitly, she invokes a neo-apostolic vision of public Christian discourse, one which destabilizes the traditional hierarchy of clergy/laity and distributes the responsibility for preaching the gospel to all believers. So maybe she is saying she preaches after all. She’s vaguing. The extraordinary exchanges at the Archbishop’s court at Cawood Palace in August 1417 – almost ten years exactly after Thorpe’s examination – are full of intimidation, hostility, tears, manacles, charges and countercharges, scriptural interpretation, and accusations of heresy. Through it all, Kempe defends herself with semantic, syntactic, and material pragmatic and metapragmatic strategies. She exploits vague category terms and material signifiers to evade the clerics’ accusations. She counters the clerics and the Archbishop by hedging, mitigating, or echoing and reframing their terms and relevance contexts to articulate her own views and faith “wel anow.” Sometimes she baldly asserts her own interpretations of scripture or Christian virtue. She reaccents the Archbishop’s and clerics’ words and metapragmatically asserts that she will continue her public speaking about God in her “comownycacyon and good wordys,” as a righteous Christian woman should do. Moreover, she says, speaking about God does not meet the performance criteria for preaching. In her pragmatic and metapragmatic discourse, Kempe claims that all Christians should speak about God, whether through preaching, scripture reading (which Kempe regularly seeks out), or in more informal conversation.

Margery Kempe’s Str ategic Vague L anguage

233

Image 6.1  Oak pulpit

Image from St. James Church, Castleton Acre, Norfolk. Available at: http://www.norfolkchurches. co.uk/castleacre/images/dscf0260.jpg. Accessed 1 May 2020

Image 6.2 Luther, Passional Christi und Antichristi (Wittenberg, 1521), pp. 18-19

Woodcuts by Lucas Cranach the Elder. Internet Archive. Available at: https://archive.org/details/ passionalchristi00luth/page/n17/mode/2up. Accessed 1 May 2020

234 

The Medieval Life of L anguage

However, despite her pragmatic strategies and discursive maneuvers, Kempe remains vulnerable because of the institutionally sanctioned power asymmetry between clergy and laity. In fact, Kempe’s position is made even riskier because of her struggle to control her discourse, maintain her face, and assert her views before the York clergy. Her seeming pragmatic effectiveness challenges the clerics’ presumptive institutional and sex-based authority. During her examination at York, Kempe’s discursive struggles pit her sophisticated pragmatic and metapragmatic expertise against the Church’s institutional power to regulate and prescribe lay people’s discourses.

Kempe Tells a Tale Kempe and the Archbishop might appear to be at a standoff, but the York narrative doesn’t end with Kempe’s spectacular metapragmatic defense of her public speaking. There’s one more narrative sequence before she leaves the city. The final scene in this complex narrative episode begins when a “doctowr” who had examined her earlier responds to Kempe’s pragmatic refusal to be accused of preaching. He tells the Archbishop, “Syr, sche teld me the werst talys of prestys that evyr I herde” (l. 2978). This sentence is another topic opener, and sure enough, the Archbishop once again rises to the bait. Bowet uses the authoritative speech act command (“comawndyd”) and orders Kempe to repeat to him the story, the tale of the Bear, the Priest, and the Pear Tree. Kempe’s command performance is more than he asked for. Before retelling the tale, Kempe reframes her earlier tale-telling with more semantic vague language while continuing to address the Archbishop with excessive politeness: “Sir, wyth yowr reverens, I spak but of o preste be the maner of exampyl” (ll. 2979-2980). Again, Kempe exploits the potential vagueness of reference to defend herself from clerical accusations. She hedges any reference in her tale to clerical misbehavior and punishment by saying the tale did not refer to all “prestys” but to a certain cleric (“o preste”). In fact, the priest may or may not be referentially real because, she says, she told the tale “be the maner of exampyl.” Kempe switches the relevance for her speech. Addressing Bowet, she reframes her earlier intention in terms of the relevance of a genre (exemplum) and then hedges on the distribution of the ‘example’ to the entire set of priests. Is the character in the tale a token of a type or a discrete singular? Kempe’s implicit theory of reference here is difficult but cagey. The semantics and signification questions implied in Kempe’s response were also taken up by medieval logicians and semioticians, sometimes with

Margery Kempe’s Str ategic Vague L anguage

235

pragmatic implications. In late medieval logic, reference in a given utterance was often analyzed in terms of the relation between a genus word/name and a species or individual. In terministic logic, the noun “preste” is a general term because it names a category of something and can be predicated of many things. When the Middle English quantifier/adjective “o” (one, a) is attached to “preste,” the term is indefinitely restricted (restrictio). Attaching ‘every’ to “preste” would explicitly expand (ampliatio) the term’s reference to the entire class of priests. So “o” is a vague quantifier because it doesn’t necessarily make the term a singular with reference to one and only one thing, in the way substituting ‘John’ or adding ‘this’ to “preste” would (see, e.g., Bacon, Summulae dialectices, 1986: 1.1.1-5). We can analyze Kempe’s utterance as an “attributive use” rather than “referential use” of descriptive language. For Donellan (1966), the phrase might imply something like ‘Whoever fits this description is who I am referring to.’ Vague but still referential. Medieval logicians, however, usually discussed the semantic relation between the singular name (Socrates) and the more general word homo in terms of predication. Peter of Spain, for instance, discussed whether in the expression Homo currit I can substitute Socrates for homo and ‘mean’ the ‘same’ thing? In one sense, yes, in that Socrates refers to a being for whom homo can be predicated. In another, more denotative sense, no, in that homo does not refer to Socrates and only Socrates (Tractatus, ed. de Rijk, 2.2-10; 1972: 17-19). The scholastic distinction does not map exactly onto the Fregean distinction between sense and reference, but the point is clear enough. Moreover, the immediate context of Kempe’s speech complicates the semiotic type/token relation she relies on. By refusing to agree that “o preste” implies all priests, even by example, she hedges the reference for singularity and classes. Kempe recounts for the Archbishop the story of the bear, the priest, and the pear tree, a tale which deploys the scatological language of clerical and fraternal condemnation found in later medieval satire (cf. Zittya 1986). A priest takes himself into the woods “for the profite of hys sowle.” When night falls, he goes to sleep in a “fayr erber” under a pear tree in full blossom. A bear comes and shakes the tree and then stuffs himself with fruit and flowers. He turns and shits on the priest, who has “gret abhominacyon of that lothly syght” (ll. 2981-2988). In the morning, the humiliated and troubled priest wanders with “dowte what it myth mene.” He meets a “semly agydd man lych to a palmyr er a pilgrime” (l. 2990). The palmer tells the humiliated priest that his encounter with the bear who eats flowers and then shits all over him signifies the priest’s lechery, greed, backbiting, impiety and incoherent mumbling through Mass. The tale symbolically signifies through

236 

The Medieval Life of L anguage

metaphorical transference that the priest produces spiritual and moral shit because of his sinfulness and failure to give full attention to his religious duties as priest. The pilgrim-exegete’s words echo Kempe’s criticisms of the clergy elsewhere in her narrative. However, Kempe’s use of reported speech in her exemplum (“he said”) distances the teller from the tale. The “agydd man,” not Kempe, voices the interpretation of the priest’s experience. While the tale’s referents are concrete, the speaker’s context for the tale is indefinite. Further, Kempe claims to be telling the narrative as an “exampyl,” invoking the moral genre exemplum rather than directly reporting an actual or even imagined event. After Kempe’s earlier exchanges about what constitutes preaching and ordinary speaking, her subtle pragmatic manipulation of genre context here should not surprise us. Reference is complicated by both genre conventions and the logico-semiotic problem of how category terms refer in particular instances. When she f inishes recounting her tale for the Archbishop, Kempe’s exemplum prompts two affective evaluations and a metacommentary on pragmatic uptake from the audience. Reflecting perhaps his responsibility for the quality of pastoral care, the Archbishop “likyd wel the tale and comendyd it, seyng it was a good tale.” The cleric, however, “whech had examynd hir befortyme in the absens of the Erchebischop” says, “Ser, this tale smytyth me to the hert” (ll. 3008-3011). The narrative arc of Kempe’s examination at York comes full circle here. The conscience-stricken cleric addresses the Archbishop, but it is Kempe who comments metapragmatically on the nature of his response. She refers indirectly to her own discourse with language that is generic, ambivalent, and referentially vague. Replying to the cleric, she quotes her “worschipful doctowr, ser, in place wher my dwellyng is most, … a worthy clerk, a good prechar,” who said “many tymes in the pulpit, ‘Yyf any man be evyl plesyd wyth my prechyng, note hym wel, for he is gylty’” (ll. 3011-3015). Before beginning her tale, Kempe had distanced herself from clerical criticism and accusations of preaching. But when she finishes the tale for the Archbishop and cites her “doctowr … a good prechar,” she indirectly associates her own exemplum with clerical preaching. Earlier, she denied she preached, but now she implicitly identifies herself with the “worschipful doctowr … whech boldly spekyth ageyn the mysgovernawns of the pepil and wil flatyr no man” (ll. 3012-3013). Two can play the self-identification game, she implies. While the cleric identifies himself with the sinful or wayward priest in the story, Kempe indirectly characterizes herself as a preacher. Kempe uses the adverb “boldly” (l. 3012) to describe the “worschipful doctowr”’s preaching, which might also implicitly apply to Kempe herself.

Margery Kempe’s Str ategic Vague L anguage

237

Kempe confirms her identification with the actions and speech of the “worschipful doctowr” when she turns to the York cleric and addresses him by reaccenting the “good prechar”’s speech: “‘And ryth so, ser,’ seyd sche to the clerk, ‘far ye be me, God forgeve it yow’” (l. 3015). If any listener is chastened by her exemplum, that is, her “comownycacyon and good wordys,” that’s how it goes. If the shoe fits…. Kempe’s attributive reference in a pragmatic context depends on the listener’s uptake as to whether or not he is actually being referred to. The cleric who had chastized and accused her when she arrived at the minster now feels chastened: “The clerk wist not wel what he myth sey to hir.” The narrative then affirms Kempe’s stronger discursive position when the cleric asks Kempe to forgive him and “… Also he preyid hir specyaly to prey for hym” (ll. 3016-3018). Kempe’s exchanges with the Archbishop and others at York shift between vague, indirect, oblique discourse and more direct, accusatory discourse. Her interactions and discursive practice reveal her sophisticated linguistic, pragmatic, and metapragmatic understanding in hostile or controversial situations. Throughout the narrative Kempe is represented as causing discomfort or trouble, but in the York episode her discourse strategies manifest her ability to speak publicly and persuasively as a lay woman about religious and ethical matters. She discursively matches men with greater authority, education, and status not only by citing and interpreting scripture but by pragmatically shifting her footing and using evasive or reaccenting strategies to counter their accusations while at the same time critiquing clerical behavior explicitly, implicitly, and narratively. Kempe’s metapragmatic awareness leads to her self-defense. The end of the York examination section in the Book of Margery Kempe confirms her hardwon discursive success. She has promoted her positive face and maintained her negative face and perhaps her freedom of movement. The York episode ends with a generally positive reversal of Kempe’s reputation as it was represented at the beginning of Chapter 50. When she returns from Cawood Palace to York, “mech pepil and […] ful worthy clerkys” receive her enthusiastically, thanking God that this “not lettryd” woman could “answeryn so many lernyd men wythowtyn velani or blame” (ll. 3028-3031). The different responses from different subgroups of townspeople illustrate how contexts of “structuring structurations” of power can shift within and because of alternate hegemonic discursive events and speech, pragmatic contests, and public pragmatic acts (cf. Laclau and Mouffe 2000; Bourdieu 1990: 52). Some clerics at the Archbishop’s court are repentant, the Archbishop is annoyed but less accusatory than before, and the York lay people and some of the clergy celebrate Kempe’s empowered and face-saving

238 

The Medieval Life of L anguage

talk with the learned. The Archbishop even asks Kempe “to preye for hym” (ll. 3027-3028), so long as she leaves. The episode concludes with the Archbishop asking, perhaps with some peevishness, “Wher schal I have a man that myth ledyn this woman fro me?” (ll. 3018-3019). For five shillings, someone does. The narrative of Kempe’s three hostile encounters at the Archbishop of York’s court (Chapters 50-52) foregrounds key pragmatic ideas and practices and metapragmatic awareness in medieval interpersonal exchanges. Kempe at York is defined by asymmetric power relations, but within that dialogic context she shares with her interlocutors a speech code (English), knowledge of scriptural text in English and Latin, and a set of conversation strategies for handling disputation, conflict, and argument. In the face of hostile accusations and physical threats, Kempe uses various pragmatic and metapragmatic strategies to evade charges of heresy and having transgressed religious authority. She strategically asserts her agency with the consistent use of polite address, fuzzy semantics (hyper/hyponyms), potential equivocation, echoic responses, and embedded syntax to shift relevance, shifting reference between literal and figurative language, vague extenders, hedging, genre codes, indirect discourse, and material signifiers of authority – a veritable catalogue of pragmatic discourse features. These pragmatic strategies enable Kempe to partially sidestep the Archbishop’s control, seize the conversation floor, recontextualize terms and utterances, and distance herself from some potentially dangerous utterances while making claims and direct accusations of her own. In addition, she counters some clergy’s accusations and scriptural citation with her own citations. At York, she uses scripture to support her claim to “spekyn of god” in pious colloquial “comownycacyon.” Denying that she preaches, Kempe nonetheless adopts several preaching stances, including telling an exemplum to support a moral lesson. Kempe’s textual pragmatics in the York examinations are as subtle and multidirectional as her spoken pragmatics. Language in use, expressive, affective, representational, and symbolic, are all in play in Kempe’s account of her examinations at York. Virtus sermonis, the force of words, is the power linguistic forms and signs have and which speakers have at their disposal to construct things, knowledge, and worlds. But that power to mean is always situated in relation to speakers and interpreters, the contexts we make. Authoritative discourse largely enacts and maintains “structuring structures of disposition, the habitus, which is constituted in practice and is always oriented towards practical functions” (Bourdieu 1990: 52). The world as given. But when deployed as alternate hegemonic practice, linguistic force realizes other potential agencies and

Margery Kempe’s Str ategic Vague L anguage

239

a different world. Reimposition exploits the meaning potential not yet actualized, the life within language and discourse. Some philosophers and pragmatists, Habermas (1998: 67), for example, try to parse praxis by focusing on one action or form to the exclusion of others and by differentiating ‘sentence’ from ‘utterance’ in order to establish the primacy of the first; that is, what is collectively meant or agreed to is the ground for understanding what is said as meant. Habermas’ analytic strategy may help distinguish form from function, but it gives us a lopsided understanding of what actually happens in speech as praxis. As reimposing and reaccenting agents, speakers make and reframe contexts, clarify or hedge reference, and establish or shift relevance to maintain, gain, or recover symbolic and literal power. Kempe’s struggles for discursive relevance disrupt established power, at least temporarily, while protecting her life and social status and, most important, giving her a public voice. Many readers, myself included, have focused on – Will we admit ‘been attracted to’? – Kempe’s disruptive behavior and vocalizations and her affective and literate piety and what they suggest about medieval affectivity and embodied experience. But Kempe’s pragmatic sophistication, her discursive competence, also calls for our attention, even if, perhaps especially if, she is not always cooperative. Kempe’s narrative and pragmatic actions resonate with Olivi’s and Bacon’s theories of imposition and equivocatio, but the York episode presents Kempe’s speech, intention, and dialogic interaction in a more grounded, sociolinguistic, and pragmatic context. Kempe’s text presents some contested or counter hegemonic linguistic acts and pragmatic speech – evasive or equivocal speech, pragmatic gesture, lay scriptural interpretation, hedging before asserting – more positively than we find them described in many official Church accounts. Kempe’s narrative may not reproduce her life and speech entirely accurately. It may not summarize the exchanges which actually took place. Nonetheless, her account, like Thorpe’s Testimony, presents in rich detail how fifteenth-century pragmatic strategies and metapragmatic awareness were understood and used dialogically in an English socio-religious context. In the Book those strategies and metaawareness effectively enable an antagonistic, secular, pious, and discursively sophisticated woman to assert her heartfelt and relatively orthodox views. Her pragmatic savvy opens up a possible alternate hegemony of devout lay agency. It may also have saved her life. Barry Bonds was convicted of using evasive language in his grand jury testimony, but he is not in Kempe’s league. Not even close.



One More Thing Abstract A medieval pragmatic orientation in grammar emerged from both theoretical and practical knowledge. Medieval pragmatic approaches to language actualized an alternate (Laclau and Mouffe [2000] would say “antagonistic”) hegemony, a discursive space of thinking differently about language, meaning, and communication. Pragmatically oriented grammarians reanalyzed many examples of (propositional) speech found in medieval logic texts and more formalist grammatical writing and arrived at different conclusions. They found descriptive value and analytic productivity in what others regarded as fallacies and errors. In effect, pragmatically oriented grammarians actualized as meaningful potential alternatives in the linguistic system. Some of those alternatives are exemplified in poetic discourse and how heretics talk. Keywords: medieval pragmatics, modern pragmatics, virtus sermonis, alternate hegemony (Laclau and Mouffe), grammar and nature

At the start of this book I described pragmatics as an approach to language from the point of view of users and practice. That starting point shapes this critical history of medieval pragmatics theory, ideas, and metapragmatic awareness. Investigating medieval linguistics, philosophy of language, and narrative representations of speech interaction demonstrates the concrete workings of that definition of pragmatics in intellectual and imaginative discourse. But can we say that the historicity of medieval pragmatic ideas and metapragmatic awareness are epiphenomena attached to some essential core? ‘Pragmatics is pragmatics,’ right? Not exactly. Contemporary pragmatics is a broad interdisciplinary field comprising linguistics, philosophy of language, cognitive studies, and more. Pragmatic theory and analysis and the disciplinary definitions and strategies which structure them are motivated and shaped by the alliances and antagonisms in broader contexts of language, knowledge, and social

Amsler, Mark, The Medieval Life of Language: Grammar and Pragmatics from Bacon to Kempe. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463721929_epilogue

242 

The Medieval Life of L anguage

interaction. Paradigm theory (Kuhn), archaeology and genealogy analysis (Foucault), and historical ontology (Hacking) suggest how an intellectual discourse such as pragmatics or linguistics is a networked practice that establishes and reestablishes its questions, terms, and evidence so as to assert or maintain legitimacy and authority. A discourse is historically situated. It emerges and can be displaced or superseded by another or maintained as the way things are. A dominant hegemony is always under construction and maintainence, however powerful or given. Hegemony is always dynamic and therefore changeable. Alternate hegemonies are possible. Dialogism holds the possibility that things can be otherwise. Contra Kuhn’s theory of a dominant “paradigm,” a discourse such as linguistics or pragmatics often constructs more than one theoretical model at a time. Medieval linguistics manifests a multiplicity of models which interrelate around the question ‘What is grammatica?’ A pragmatic orientation is just one. Traditional grammatica had focused on the parts of speech. After about 950 CE grammar became less clearly separate from logic and dialectic. As grammarians gave more attention to syntax and the logic of propositions, they enlarged the field of grammar as an explanatory discourse with analytic tools. Grammar became more closely aligned with logic, as it is for some modern theorists of pragmatics. But that still didn’t necessarily make for a pragmatic theory or perspective on language. Reading thirteenth- and fourteenth-century grammatical texts as intellectual practice shows us that pragmatics is not just pragmatics anywhere, anytime. A medieval pragmatic orientation in grammar emerged from both theoretical and practical knowledge. Medieval pragmatic approaches to language actualized an alternate (Laclau and Mouffe [2000] would say “antagonistic”) discursive space of thinking about language, meaning, and communication. Augustine’s theory of language and signs was the principal progenitor of late medieval semiotics, but after 1200 Augustinian semiotics competed in the schools with Aristotle’s account of language and cognition in the introduction to De interpretatione, De sophisticis elenchis, William of Moerbeke’s (c. 1220/1235-1286) revised Latin translation of De anima, and Averroist and Avicennian commentaries on Aristotle. Working with Augustine’s semiotics, Kilwardby, Bacon, and some of the anonymous grammarians discussed in this book highlighted language as the principal sign-making activity, as do twentieth-century theorists such as Bakhtin, Eco, and Barthes. Bacon started with signification and signifying practices and then described language as the principal, but not the only, mode of signifying. A pragmatic theory of language and meaning, derived in part from Augustinian semiotics, represented a distinct alternative to grammatical theory based around Aristotle and modism.

One More Thing

243

We have seen how pragmatically oriented grammarians reanalyzed many examples of (propositional) speech found in medieval logic texts and more formalist grammatical writing and how they analyzed and concluded differently. They found descriptive value and analytic productivity in what others regarded as fallacies and errors. In effect, pragmatically oriented grammarians actualized as meaningful potential alternatives in the linguistic system. They subverted, perhaps without explicitly planning to, many of the received assumptions and analyses of grammatica as they had been worked out in the twelfth- and early thirteenth-century schools. The pragmatic grammarians also asked different questions about situated language, considered types of vocalizations which more formalist grammarians excluded, and repurposed some metalanguage (e.g., conceptus, impositio, equivocatio) for talking about language as an ordered system for expression and communication. The pragmatic grammarians restructured language theory by emphasizing signification and meaning as socially, dialogically, and inferentially constructed. They gave more attention to everyday uses and forms of speech and social contexts. These grammarians and philosophers argued that a speaker’s intention alone does not necessarily determine all an utterance can mean. As social discourse, an utterance was meaningful in multiple relevant contexts, including speaker’s intent, the listener’s inferences about what the speaker means to say, and the way both speaker and listener construe the relevance of the utterance. Positing a more fluid notion of communicative context for meaning, the pragmatic grammarians showed the constitutive and dynamic role of context in cognitive and linguistic interactions between speaker and listener. Meaning is social and dialogic. The context of an exchange, not just the formal utterances, and the audience’s uptake were central to medieval pragmatic theory, pragmatic and metapragmatic ideas, and of course social practice. Medieval pragmatists displaced the form of an utterance as the only or final determination of what an utterance ‘means.’ They effectively reframed utterances as holding the potential for multiple signif ications depending on how participants assess relevant contexts. In their writings, the grammarians discuss how communication is accomplished not only with ‘complete’ propositions but also with elliptical expressions, ‘natural’ vocalizations, figurative speech, and novel utterances in different contexts. They afforded theoretical significance to interjections, actus exercitus, and performative or non-referential aspects of language. Their analyses of the vocabulary of ‘pain’ (dolor, heu, etc.) differed from those of modistic grammarians and concretely illustrate the multiplicity of models in medieval grammatical theory. The pragmatic grammarians’ model and

244 

The Medieval Life of L anguage

theory incorporated affect and performativity, feeling in language. Medieval pragmatic theory was a revolution in grammatical analysis and intellectual knowledge about language and had important implications for theological, exegetical, and poetic discourse as well as grammar. In effect, the semiotically and pragmatically oriented grammarians produced an ‘alternate hegemony’ within late medieval grammar, if only temporarily. As we have seen, Kilwardby, Bacon, Pseudo-Kilwardby, and Magister Johannes reevaluated received ideas about language and meaning, especially those relating to meaning and form, and argued that: 1 the lexicon of natural language includes both natural and conventional vocalizations which are understood as meaningful in particular contexts; 2 the role of concepts in cognition and emotional speech is complex in that mental concepts are not always similitudines [likenesses] of the things referred to, as Aristotle would have it (in Boethius’ translation), and that concepts do not necessarily refer to mental universals which mediate between a speaker’s understanding and a listener’s understanding of what is ‘meant’ by an utterance; 3 natural language is a sign-making system with an elastic set of grammatical rules and an open-ended lexicon which speakers and listeners deploy within a speech community to produce meaningful utterances responsive to particular contexts, expectations, and intentions. These linguistic positions focused on context, intention, and affect and entailed that medieval pragmatic theory articulated an incoherent concept of natura with respect to language. On the one hand, a natural language, primed by biological (‘God-given’) endowment, is mostly a conventional system of rules and vocabularies for expressing thought and feeling. On the other, a natural language is a product of reason (ratio) and the intellect, which medieval grammarians regarded as God-given and therefore natural to what it is to be human and created in God’s universe. With little hyperbole, we can say that medieval pragmatic theory rejected significant aspects of Aristotelian and Augustinian doxa with respect to how those philosophies handled language as a formally ordered, contextually meaningful signifying practice. How we mark the stability of meaning sets the parameters for semantics and pragmatics. Pragmatics maintains the openness of language as meaningmaking practice, meaning as potential and constitutive as well as actual. Medieval pragmatic grammarians elide what current linguistics calls the ‘semantics-pragmatics interface.’ Contemporary linguistics usually analyzes the semantics-pragmatics interface in terms of how an utterance conveys

One More Thing

245

information based on inferences from the literal meaning of the words (including conventional figurative meanings) in the speech community as well as the listener’s inferencing of the speaker’s motives and contexts which cause the utterance to appear in the form the speaker uses. The relation between semantics and pragmatics varies depending on the presuppositions, implicatures, and scalar implicatures considered cognitively in play and inferred at a particular moment (cf. Horn 1984, 2001). Medieval grammar and logic articulated a different version of the semantics-pragmatics interface in the debates around the concept of virtus sermonis. Generally, virtus sermonis refers to the form of the words themselves understood independently of a speaker’s intention or authority. Virtus sermonis presupposes that language speaks itself, that the text is uttered by the conceptualized ‘speech community’ as a set of meanings and norms. However, grammarians and philosophers of language such as Bacon adopted a semiotic framework that sometimes overrode logicians’ realist distinction between significatio and suppositio. The pragmatic grammarians redescribed utterance context as a dynamic field of potential verbal meaning and including both speaker and audience intentions, shared or competing presuppositions and conventions, audience expectations, and novel contexts. Many of these grammarians represented the speech context as dialogic, co-created or conflictual between speakers and listeners. Theoretically and narratively, medieval pragmatic discourse situated language in social space and consistently represented language as dialogic social interaction whose potentially multiple meanings are activated by users in their relations with others and other contexts. As part of the field of medieval pragmatics, literary texts, life narratives, and other accounts of dialogic interactions instantiate rather than ‘reflect’ pragmatic ideas and metapragmatic awareness. For all the differences between Gui, Chaucer, Thorpe, and Kempe, their writings reveal how broadly many pragmatic ideas and metaknowledge of pragmatic strategies were understood and how those ideas ands strategies were the focus of linguistic analysis in various textual discourses, not just grammatica. In grammar and poetry, medieval pragmatics, like vernacular theology, illustrates well Gramsci’s notion that “although one can speak of intellectuals, one cannot speak of non-intellectuals because non-intellectuals do not exist” (1971: 9). The contemporary discipline of historical pragmatics largely focuses on pragmatic and discourse markers as used in particular texts and on changes in discourse markers over time. A history of pragmatics, on the other hand, at least this history of pragmatics, reads literary and historical texts for what the uses of pragmatic and discourse markers and strategies

246 

The Medieval Life of L anguage

reveal about the writers’ understanding of pragmatics and language as social and discursive practice. A critical history of pragmatics finds traction in historical and literary narrative representations and poetic discourse as much as in grammatical theory and philosophy of language. Bacon’s theory of tacit re-impositio, equivocatio, and intentional signification questions the presumed fixity of lexical meanings over time and across speakers, highlighting instead the relationality of speakers and interpreters in ongoing sign-making and social understanding. Writers such as Thorpe and Kempe narrate their pragmatic strategies to construct alternate subjectivities in their encounters with powerful agents of dominant hegemonic order and the official Church. Dialogically, they deploy equivocation and vague language as strategies for social and cognitive agency. Their texts effectively model dissenting speech not only as expressing alternative views to mainstream Christianity but also as struggling to gain status and public voice within a dominant institutional discourse seeking to marginalize them or coerce them to speak in more conformist ways. Furthermore, their narratives reveal the knowledge and use of a different kind of modal syntax from that which logicians such as William of Sherwood (1190-1249) and Peter of Spain discussed in terms of truth conditions (cf. Knuuttilo 1993: 99-175). Waldensians, Thorpe, and Kempe use echoic syntax and recontextualization to reaccent and evade their examiners’ implicit accusations or to speak sincerely on their own terms. No less than Bacon and Magister Johannes, Thorpe and Kempe embrace in their practice a different understanding of equivocatio and meaning potential from that of many medieval logicians. These literary texts and life narratives constitute a different but no less critical part of medieval pragmatic discourse. In the order of pragmatic discourse, they participate as another intellectual mode alongside theory for thinking within a discursive practice. That is, pragmatic theory and metapragmatically situated literary and life-writing make up an alternate hegemony within the medieval discursive field of grammar and philosophy of language. Literary practice can, but doesn’t necessarily, extend the conventional boundaries of what constitutes the shape of language and potential meaning and the social order of knowledge and power that meaning sustains. Roman Jakobson generalized the question semiotically: “Equivalence in difference is the cardinal problem of language and the pivotal concern of linguistics” (1987: 430). Pragmatics as social practice in theory and literature takes Jakobson’s structuralist point further. The central problem of language and subjectivity are the alliances and antagonisms which produce new imaginaries and social spaces with what are presumed

One More Thing

247

to be stable equivalents. The life of language is embedded in change and reformation. Literary discourse opens other pragmatic potentials. A pragmatic reading of Chaucer’s texts suggests how a single interjection such as allas or a single word such as water are meaningful utterances in particular contexts. Allas and Water are affectively and referentially polysemous rather than semantically fixed. By dislocating linguistic form as stable performance, Chaucer’s poems semiotically interrogate the conditions for meaning in particular contexts. Many of his writings imply that meaning is use and that use is always contextual and open- ended, even when explicit referential meaning is prescribed. The poems’ textual manipulations of interjective and affective speech, rather than being marginal to utterances, cause utterances to swerve away from generalizations and universalisms about meaning, subjectivity, and intention. Chaucer’s poems generate complex textual and discursive contexts, dialogism, which in effect affirm the life of language as variable, productive, and potentially anti-authoritarian at least as much as they constitute individual speakers as autonomous ‘subjects’ ‘or characters.’ A critical pragmatics points to the way speakers in Chaucer’s texts (and for that matter in Gui, Thorpe, and Kempe’s writing) seem to inhabit language rather than that a particular language or mode of speaking belongs to the speakers. It’s not that the texts rule out will, choice, or intention with respect to narrative representations. Rather, the characters’ textual or deictic positions are a function of utterances in a context. Practices make subjects; subjects inform practices. Speakers, including narrators, drag language into context, where meanings are repositioned. Vague language, equivocation, affective utterance, resemanticization, reaccenting, all are features of dialogic discourse which forms subjects and dissent in discursive practices. Medieval and modern pragmatic understanding and metapragmatic awareness both disrupt, but in different ways, the “conduit metaphor” for ‘rational’ speech, that is, the assumption that language carries thought directly and words (should) express discrete thought clearly and unambiguously to a reasonable person (Reddy 1979). In a pragmatic approach to language ‘the medium’ is definitely part of ‘the message.’ We can assert the alterity of historical distance as a precondition for accessing the past in difference, but we also need to acknowledge that contemporary pragmatics thinking, its metalanguage and theory, can help us access medieval pragmatic thinking, whether by continuity or by contrast and difference. Not every discursive formation is a rupture. Another important point to press here relates to medieval linguistics and natural language. Much has been made of the fact that medieval

248 

The Medieval Life of L anguage

grammarians almost always analyzed Latin as the stand in for Langage, leading to the cliched notion that Latin grammatica was an abstracted discourse not fully engaged with natural language as used by people in the everyday real world. Yes, Latin in the Middle Ages was mostly an elite mode of social and intellectual discourse. However, one of the striking things uncovered in the medieval pragmatics theory and metapragmatic practices and texts discussed in this book is how many of the grammarians’ examples, situations, and analyses are focused on real world, everyday, or homely experiences and utterances. Although couched in Latin, many of the grammarians’ examples seem to have more to do with real world speech than do the logicians’ propositions or sophisms, Homo currit or Ergo malum est bonum. The same applies to some of the accounts of suppositio outside logical texts and commentaries. Grammarians’ discussions of affective speech and modal syntax as the counterpart of cognitive speech shift intellectual attention to natural language in various contexts, not only formal or cognitive ones. When medieval grammarians pay attention to concept formation and affect in speech, when they consider performatives in addition to more explicitly referential utterances, when they unpack the role of audiences in constructing meaningful speech, their theoretical and analytic accounts seem to focus more directly on the particularities and potentials of natural language rather than on propositions constructed to exemplify or interrogate a problem with predication, anaphora, or contradiction. In other words, while writing in Latin, the grammarians discussed here were paying attention to the sort of pragmatic practices people were already using and engaging with in both their Latin and vernacular speech, as the medieval literary, life-writing, and chronicle texts make clear. For many medieval grammarians, pragmatic understanding of language in social interactions had real world and theological implications outside the academic discipline of philosophy. The implications of pragmatic thinking about language are as varied as the threads which make up pragmatics discourse. Pragmatic ideas about language also had some specific implications for medieval theology and worldview. For example, Nature was a complicated, even contradictory concept in the Middle Ages. Theologians and philosophers understood natura to be that which is necessary, not just possible. At the same time, they identified natura as God’s agent, a force as creative as human will, reason, and imagination. Nature produces order as much as does human reason. Boethius mapped the nature/convention dichotomy onto the logical distinction between genus and species: “nam quod apud omnes idem est natura est, consuetudinis vero est quod apud aliquos permutatur” (De

One More Thing

249

divisione, ed Magee; 1998: 12-13) [For that which is the same for all is by nature, while that which changes from one people to another is a matter of custom]. Traditionally, Latin grammatical discourse had distinguished between natura and consuetudo, with emphasis falling on language as an ordered conventional system for expressing and communicating meaning. However, many medieval pragmatists implicitly adopted Bacon’s distinction between natural phenomena and natural signs. Something is a sign within a semiotic context when someone perceives something as a sign. Interjections, natural vocalizations (sighs, moans, giggles), affective or performative speech, implicatures, audience motivations, pragmatic action, and so forth called into question the rigid separation of natural sound and conventional language. Human expressivity and understanding brings together in the pragmatic moment, contingency and natura. Speech is both willed and natural. Medieval pragmatics and pragmatic ideas account for how natural and conventional vocalizations are expressive, ordered, and meaningful within semiotic and social space. Historicizing medieval or contemporary pragmatics reveals a way of thinking, speaking, and behaving. Historicizing also impels us to acknowledge that a discourse, however dominant, is not permanent. Continuities and ruptures, dominant and alternate hegemonies, exist side by side within dialogic spaces. Medieval grammar oriented to pragmatics and language in use articulated an alternate hegemony but was swept aside, along with Aristotelian modism, by the emergence of humanist grammar and the ideals of elegantia. The life of language and discourse is not fixity but change and reformation. With pragmatic ideas and practices, some medieval grammarians, religious reformers, and writers sought to actualize an alternate hegemony for meaning and belief, one which recognized the field of discourse as political, not altogether given, and full of possibilities.

Bibliography Primary Abelard, Peter. 1970. Dialectica. Ed. Lambert M. de Rijk. 2nd ed. Assen: Van Gorcum. —. Ethics. 1971. Ed. and trans. David E. Luscombe. Oxford: Clarendon Press. Ælfrics Grammatik und Glossar: Text und Varianten. 2001. Ed. Julius Zupitza. 3rd ed. Hildesheim: Georg Olms. Alexander of Hales. 1924-1948. Summa universae theologiae. Ed. Bernardini Klumper and the Quarracchi Fathers. 4 vols. Quarrachi: Collegium S. Bonaventurae. Anselm of Canterbury. 1946-1961. Opera Omnia. Ed. Francis S. Schmitt. 6 vols. Edinburgh: Thomas Nelson. Aquinas, Thomas. 1964-1980. Summa theologiae. 61 vols. London: Eyre and Spottiswoode. Aristotle. 1929. Physics Books 1-4. Trans. Philip H. Wicksteed and Francis M. Cornford. Cambridge, MA: Harvard University Press. —. 1975. Categories and De interpretatione. Trans. J. L. Ackrill. Oxford: Oxford University Press. —. 1986. De anima. Trans. Hugh Lawson-Tancred. New York: Penguin Classics. Augustine. 1962. De doctrina christiana. Ed. Joseph Martin, CCSL 32. Online: LLT-A. http://clt. brepolis.net. Accessed 10 February 2019. —. 1970. De magistro. Ed. K-D Daur, CCSL 29. Online: LLT-A. http://clt.brepolis.net. Accessed 10 February 2019. —. 1975. De dialectica. Ed. Jan Pinborg. Trans. Darrell Jackson. Dordrecht: Reidel. —. Regulae, in KGL, 5:533-47. Bacon, Roger. 1978. De signis. Ed. Karin M. Fredborg, Lauge Nielsen, Jan Pinborg. “An Unedited Part of Roger Bacon’s Opus maius: ‘De signis’.” Traditio, 34: 75-136. —. 1986. Summulae dialectices, I-II. Ed. Alain Libera. Archives d’histoire doctrinale et littéraire du Moyen Age, 53: 139-289. —. 1987. Summulae dialectices, III. Ed. Alain Libera. Archives d’histoire doctrinale et littéraire du Moyen Age, 54: 171-278. —. 2009. The Art and Science of Logic. Trans. Thomas S. Maloney. Toronto: Pontifical Institute of Mediaeval Studies. —. 2013. On Signs. Trans. Thomas S. Maloney. Toronto: Pontifical Institute of Mediaeval Studies. Boethius, Anicius Manlius Severinus. In Categorias Aristotelis Libri Quatuor. PL 64: 159A-294C. Online: LLT-A. http://clt.brepolis.net. Accessed 10 February 2019. —. 1998. De divisione liber. Ed. and trans. John Magee. Leiden, Boston, Köln: Brill. —. In Librum Aristotelis De Interpretatione Libri Duo. PL 64: 294D-639A. Online: LLT-A. http:// clt.brepolis.net. Accessed 1 June 2019. Boethius of Dacia. 1969. Modi significandi sive quaestiones super Priscianum maiorem. Ed. Jan Pinborg and Heinrich Roos. Corpus Philosophorum Danicorum Medii Aevi, vol. 4. Copenhagen: G. E. C. Gad. Book of Margery Kempe. 1985. Trans. Barry R. Windeatt. London: Penguin. Book of Margery Kempe. 1996. Ed. Lyn Staley. Kalamazoo, MI: Medieval Institute Publications. The Cambridge Translations of Medieval Philosophical Texts: Volume 1: Logic and the Philosophy of Language. 1988. Ed. Norman Kretzmann and Eleonore Stump. Cambridge: Cambridge University Press. Charisius, Ars grammatica. In KGL, 1:1-296.

252 

The Medieval Life of L anguage

Chaucer, Geoffrey. 1987. The Riverside Chaucer. Ed. Larry Benson et al. 3rd ed. Boston: Houghton Mifflin. —. 1989. The Ellesmere Manuscript of Chaucer’s Canterbury Tales: A Working Facsimile. Cambridge: D. S. Brewer. —. Hengwrt Manuscript. Online: http://www.sd-editions.com/AnaServer?HengwrtEx+0+start. anv. Accessed 1 February 2020. Clanvowe, John. 1975. The Works of Sir John Clanvowe. Ed. V. J. Scattergood. Cambridge: D. S. Brewer. Corpus of Middle English Prose and Verse. In Medieval Compendium. Ann Arbor: University of Michigan Library, 2000-2018. Online: http://quod.lib.umich.edu/m/middle-english-dictionary. Online: https://quod.lib.umich.edu. Dante Alighieri. 1961. Inferno. Ed. Giuseppe Vandelli. Trans. John D. Sinclair. New York: Oxford University Press. Diomedes, Ars grammatica. In KGL, 1:299-529. Donatus, Aelius. 1981. Ars maior, in Louis Holtz, Donat et la Tradition de l’Enseignement grammatical. Paris: Centre National de la Recherche Scientifique. —. 1981. Ars minor. In Holtz, Donat et la Tradition de l’Enseignement grammatical. English Wycliffite Sermons. 1987. Ed. Pamela Gradon. Vol. 2. Oxford: Clarendon Press. —. 1990. Ed. Anne Hudson. Vol. 3. Oxford: Clarendon Press. Gosvin of Marbais. 1998. Tractatus de constructione. Ed. Irène Rosier-Catach. Nijmegen: Ingenium. Grammatici latini. Ed. Heinrich Keil. 8 vols. (Vol. 8: Anecdota Helvetica, ed. Hermann Hagen). Leipzig: B. G. Tuebner. Rpt.: Hildesheim: Georg Olms, 1961. (Originally published 1855-1880). Gui, Bernard. 1964. Practica inquisitionis heretice pravitatis (Part 5). Ed. and trans. G. Mollat. 2 vols. Paris: Société D’Edition ‘Les Belles Lettres.’ —. 2006. The Inquisitor’s Guide: A Medieval Manual on Heretics. Trans. Janet Shirley. Welwyn Garden City, NJ: Ravenhall Books. Helias, Peter. 1993. Summa super Priscianum. Ed. Leo Reilly. 2 vols. Toronto: Pontifical Institute of Mediaeval Studies. Heresies of the High Middle Ages. 1991. Trans. Walter W. Wakef ield and Austin P. Evans. New York: Columbia University Press. Heresy Trials in the Diocese of Norwich, 1428-1431. 1977. Ed. Norman P. Tanner. London: Royal Historical Society. Holkham Bible Picture Book. 1971. Ed. F. P. Pickering. Anglo-Norman Texts, vol. 23. London: Blackwell. The Holy Bible, Containing the Old and New Testaments, with the Apocryphal Books, in the Earliest English Versions Made from the Latin by John Wycliffe and His Followers. 1850. Ed. Josiah Forshall and Frederic Madden. 4 vols. Oxford: Clarendon Press. Isidore of Seville. 1911. Etymologiae sive origenes. Ed. Wallace Martin Lindsay. 2 vols. Oxford: Clarendon Press. Jordanus of Saxony. Notulae super Priscianum Minorem. In Sirridge 1980. Kilwardby, Robert. Super Priscianum Minorem (selections). In Rosier 1994. —. 1984. In Donati Artem maiorem III. Ed. Laurence Schmücker. Brixen/Bressanone: A. Weger. —. Pseudo-Kilwardby. 1975. Super Priscianum Maiorem (Cambridge, MS Peterhouse 191, ff.1r-229v). In Fredborg et al. CIMAGL, 15: 1-146. Lambert of Ligny-le-Chatel (aka Lambert of Auxerre, Lambert of Lagny). 1971. Logica. Ed. Franco Alessio. Florence: La Nuova Italia Editrice. —. 2015. Logica, or Summa Lamberti. Trans. Thomas Maloney. Notre Dame, IN: University of Notre Dame Press. Luther, Martin and Philipp Melanchthon. 1521. Passional Christi und Antichristi. Wittenberg: Johann Rhau Grunenberg.

Bibliogr aphy

253

Malory, Thomas. 1971. Works. Ed. Eugene Vinaver. 2nd ed. Oxford: Oxford University Press. Map, Walter. 1914. De nugis curialium. Ed. M. R. James. Oxford: Oxford University Press. Nicholas of Paris. 1979. Sincategoreumata. In H. A. G. Braakhuis, De 13de Eewse Tractaten over Syncategorematische Termen. Inleidende Studie en uitgave van Nicolaas van Parijs’ Sincategoreumata. 2 vols. PhD Dissertation, Leiden. (Vol. 2: Nicholas’ Sincategoreumata.) Olivi, Peter John. 1986. “Quaestiones logicales: Critical Text.” Traditio, 42: 335-388. —. 2016. “Quid ponat ius vel dominium.” Oliviana. Online: http://journals.openedition.org/ oliviana/882. Accessed 11 Nov 2019. Patrologiae Cursus Completus: Series Latina. 1844-1855, 1862-1865. Ed. Jacques-Paul Migne. Paris: Imprimerie Catholique. Pecock, Reginald. 1921/1996. The Donet. Ed. Elsie V. Hitchcock. EETS, vol. 156. Oxford: Oxford University Press. Peter of Celle. 1987. Selected Works. Trans. Hugh Feiss. Kalamazoo, MI: Cistercian Publications. Peter of Spain. 1972. Summulae logicales (Tractatus). Ed. Lambert M. de Rijk. Assen: Van Gorcum. —. 1990. Language in Dispute (Summulae logicales). Trans. Francis Dinneen. Amsterdam and Philadelphia: John Benjamins. —. 1992. Syncategoreumata. Ed. Lambert M. de Rijk. Trans. Joke Spruyt. Leiden: Brill. Pompeius, Commentum in arte Donati. In KGL, 5:95-312. Priscian, Institutiones grammaticae. In KGL, 2:1-597, 3:1-377. Probus, Instituta artium. In KGL, 4:47-192. Pseudo-Robert Grosseteste. 1976. Tractatus de Grammatica: Eine fälschlich Robert Grosseteste zugeschreibene spekulative Grammatik. Ed. Karl Reichel. Paderbrn: Schöningh. Remigius of Auxerre, In Artem maiorem Donati commentarium. In KGL, 8:219-274. Sacerdos. Ars grammatica. In KGL, 6:427-546. Servius, Commentatrius in Artem Donati minorem. In KGL, 4:405-428, 443-448. Smaragdus. 1986. Liber in partibus Donati. Ed. Bengt Löfstedt, Louis Holtz, and Adele Kibre. CCCM 68. Turnhout: Brepols. Summum Virtutum de Remediis Anime. 1984. Ed. and trans. Siegfried Wenzel. Athens: University of Georgia Press. Terence, Eunuchus. Perseus Project. www.perseus.tufts.edu. Accessed 14 Feb 2018. Thorpe, William. 1993. Testimony. In Two Wycliffite Texts: The Sermon of William Taylor, 1406. The Testimony of William Thorpe, 1407. Ed. Anne Hudson. EETS, vol. 301. Oxford: Oxford University Press. Varro. De lingua latina. 1910. Ed. Georg Goetz and Friedrich Schoell. Leipzig: Teubner. Wyclif, John. 1858/2010. Fasciculi Zizaniorum Magistri Johannis Wyclif cum Tritico. Ed. W. W. Shirley. Roll Series. Cambridge Library Collection: Rolls. Cambridge: Cambridge University Press. doi:10.1017/CBO9780511695490. Accessed 11 November 2019. —. De Contrarietate Duorum Dominorum. 1883. In Polemical Works in Latin. Ed. Rudolph Buddensieg. 2 vols. London: Truebner and Co. —. De veritate sacrae scripturae. 1905. Ed. Rudolph Buddensieg. Vol. 1. London: Truebner and Co.

Secondary Aitchison, Jean. 1987. Words in the Mind: An Introduction to the Mental Lexicon. Oxford: Blackwell. American Heritage Dictionary. 2011. Online: https://www.ahdictionary.com. Accessed 1 November 2019.

254 

The Medieval Life of L anguage

Amsler, Mark. 1987. “The Wife of Bath and Women’s Power.” Assays, 4: 67-83. —. 2011. Affective Literacies: Writing and Multilingualism in the Late Middle Ages. Turnhout: Brepols. —. 2012a. “Mikhail Bakhtin, Langston Hughes, and the Poetic Utterance.” Macabéa: Revista Eléctronica do Netlli, 1: 43-57. —. 2012b. “Poverty as a Mobile Signifier: Waldensians, Lollards, Dives and Pauper.” In Experiences of Poverty in Late Medieval and Early Modern England and France. Ed. Anne Scott. Farnham, England and Burlington, VT: Ashgate. 227-251. —. 2018. “More Than Words Can Say? Late Medieval Affective Vocabularies.” In Later Middle English Literature, Materiality, and Culture. Ed. Brian Gastle and Erick Kelemen. Newark: University of Delaware Press. 3-24. —. 2019. “Affectus in Medieval Grammar.” In Before Emotion: The Language of Feeling, 400-1800. Ed. Juanita Feros Ruys, Michael W. Champion, and Kirk Essary. New York and London: Routledge. 26-37. Andresen, Julie. 1985. “Why Do We Do Linguistic Historiography?” Semiotica, 56: 357-370. Archéologie du signe: colloque tenu du 2 au 12 août 1977. 1983. Ed. Eugene Vance and Lucie BrindAmour. Toronto: Pontifical Institute of Mediaeval Studies. Arnold, John. 2004. “Margery’s Trials: Heresy, Lollardy and Dissent.” In A Companion to the Book of Margery Kempe. Ed. John Arnold and Katherine Lewis. Cambridge: D. S. Brewer. 75-94. Austen, J. L. 1962. How to Do Things with Words. Ed. J. O. Urmson and Marina Sbisà. 2nd ed. Cambridge, MA: Harvard University Press. Bakhtin, Mikhail. 1981. The Dialogic Imagination. Trans. Michael Holquist and Caryl Emerson. Ed. Michael Holquist. Austin: University of Texas Press. —. 1986. Speech Genres and Other Late Essays. Ed. Caryl Emerson and Michael Holquist. Trans. Vern McGee. Austin: University of Texas Press. —. 1994. The Bakhtin Reader: Selected Writings of Bakhtin, Medvedev, Voloshinov. Ed. Pam Morris. London: Arnold. Beidler, Peter. 2015. The Lives of the Miller’s Tale: The Roots, Composition and Retellings of Chaucer’s Bawdy Story. Jefferson, NC: McFarland. Bell, Allan. 1984. “Language Style as Audience Design.” Language in Society,13: 145-204. Bergs, Alexander. 2004. “Address Pronouns in Late Middle English.” In Voices on the Past: Studies in Old and Middle English Language and Literature. Ed. Alicia Rodríguez Álvarez and Francisco Alonso Almeida. La Coruña: Netbiblo. 127-38. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad, and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. Harlow: Pearson Education. Biletyzki, Anat. 1996. “Is There a History of Pragmatics?” Journal of Pragmatics, 25: 455-470. “Bonds Avoids Prison Time for Giving Evasive Testimony.” 2011. New York Times, 16 December. Bourdieu, Pierre. 1990. The Logic of Practice. Trans. Richard Nice. Stanford: Stanford University Press. Boureau, Alain and Sylvain Piron. 1999. Pierre de Jean Olivi (1248-1298). Pensée scolastique, dissidence spirituelle et société. Paris: Vrin. Brooks, Cleanth. 2007. “Irony as Principle of Structure.” In The Critical Tradition: Classic Texts and Contemporary Trends. Ed. David Richter. 3rd ed. Boston and New York: Bedford/St. Martin’s. 799-806. (Brooks’ essay first appeared 1971.) Brown, Penelope and Stephen Levinson. 1987. Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Brungs, Alexander and Frédéric Goubier. 2009. “On Biblical Logicism: Wyclif, ‘Virtus sermonis,’ and Equivocation.” Recherches de théologie et philosophie médiévales, 76: 199-244.

Bibliogr aphy

255

Bryan, Jennifer. 2016. “‘A berd! A berd!’: Chaucer’s Miller and the Poetics of the Pun.” Studies in the Age of Chaucer, 38: 1-37. Burr, David. 1971. “Petrus Ioannis Olivi and the Philosophers.” Franciscan Studies, 31: 41-71. —. 1976. The Persecution of Peter Olivi. Transactions of the American Philosophical Society, New Series, Vol. 65, Pt. 5. Philadelphia: American Philosophical Society. The Cambridge History of Later Medieval Philosophy. 1982. Ed. Norman Kretzmann, Anthony Kenny, and Jan Pinborg. Cambridge: Cambridge University Press. Cameron, Euan. 2000. Waldenses: Rejections of Holy Church in Medieval Europe. Oxford: Oxford University Press. Cameron, Margaret. 2015. “Meaning: Foundational and Semantic Theories.” In The Oxford Handbook of Medieval Philosophy. Ed. John Marenbon. Oxford: Oxford University Press. 342-362. Carroll, Ruth, Matti Peikola, Hanna Salmi, Mari-Liisa Varila, Janne Skaffari, and Risto Hiltunen. 2013. “Pragmatics on the Page.” European Journal of English Studies, 17: 54-71. DOI: 10.1080/13825577.2013.755006. Accessed 25 June 2019. Carruthers, Mary. 1979. “The Wife of Bath and the Painting of Lions.” PMLA, 94: 209-222. Cesalli, Laurent and Irène Rosier-Catach. 2018. “Roger Bacon’s Semantics Revisited.” Oxford Studies in Medieval Philosophy, 6: 62-99. Chafe, Wallace. 1982. “Integration and Involvement in Speaking, Writing, and Oral Literature.” In Spoken and Written Language: Exploring Orality and Literacy. Ed. Deborah Tannen. Norwood, NJ: Ablex. 35-53. Channell, Joanna. 1994. Vague Language. Oxford: Oxford University Press. Colish, Marcia. 1983. The Mirror of Language: A Study in the Medieval Theory of Knowledge. Rev. ed. Lincoln: University of Nebraska Press. Colombat, Bernard. 2016. “L’interjection dans la tradition grammaticale latine, de l’Antiquité à l’Humanisme.” In History of Linguistics 2014. Ed. Carlos Assunção, Gonçalo Fernandes, and Rolf Kemmler. Amsterdam and Phildelphia: John Benjamins. 83-96. A Companion to The Book of Margery Kempe. 2004. Ed. John H. Arnold and Katherine J. Lewis. Cambridge: D. S. Brewer. Copeland, Rita. 2001. Pedagogy, Intellectuals, and Dissent in the Later Middle Ages. Cambridge: Cambridge University Press. —. “Affectio-affectus in Latin Rhetoric up to c. 1200.” In Before Emotion: The Language of Feeling, 400-1800. Ed. Juanita Feros Ruys, Michael W. Champion, and Kirk Essary. New York and London: Routledge. 38-50. Courtenay, William. 1984. “Force of Words and Figures of Speech: The Crisis Over virtus sermonis in the Fourteenth Century.” Franciscan Studies, 44: 107-128. Craun, Edwin. 2010. Ethics and Power in English Reformist Writing. Cambridge: Cambridge University Press. Cullen, Christopher M. 2006. “Alexander of Hales.” In Companion to Philosophy in the Middle Ages. Ed. Jorge J. E. Gracia and Timothy B. Noone. Oxford: Blackwell. 104-109. Culpepper, Jonathan. 2011. Impoliteness: Using Language to Cause Offence. Cambridge: Cambridge University Press. — and Merja Kytö. 2010. Early Modern Dialogues: Spoken Interaction as Writing. Cambridge: Cambridge University Press. Curtius, Ernst R. 1953. European Literature and the Latin Middle Ages. Trans. Willard R. Trask. Princeton: Princeton University Press.

256 

The Medieval Life of L anguage

Damasio, Antonio. 1996. Descartes’ Error: Reason and the Human Brain. London: Macmillan. Davis, Norman et al. 1979. A Chaucer Glossary. Oxford: Clarendon Press. de Rijk, Lambert M. 1962. Logica Modernorum: A Contribution to the History of Early Terministic Logic. Vol. 1: On the Twelfth Century Theories of Fallacy. Assen: Van Gorcum —. 1967. Logica Modernorum: A Contribution to the History of Early Terministic Logic. Vol 2.1: The Origin and Early Development of the Theory of Supposition; Vol. 2.2: The Origin and Early Development of the Theory of Supposition – Texts and Indices. Assen: Van Gorcum. Derrida, Jacques. 1982. Margins of Philosophy. Trans Alan Bass. Chicago: The University of Chicago Press. Donaldson, E. Talbot. 1970. Speaking of Chaucer. London: Athlone. Donnellan, Keith. 1966. “Reference and Def inite Descriptions.” Philosophical Review, 75: 281-304. Ebbesen, Sten. 1977. “Can Equivocation be Eliminated?” Mediewistyczne, 18: 103-124. —. 1980. “Is ‘Canis currit’ Ungrammatical? Grammar in Elenchi Commentaries.” Historiographia Linguistica, 7: 53-68. Eelen, Gino. 2001. A Critique of Politeness Theories. Manchester: St Jerome Publishing. Elder, John Petersen. 1947. “The Missing Portions of the Commentum Einsidlense on Donatius’ Ars grammatica.” Harvard Studies in Classical Philology, 56: 129-160. Engels, Joseph. 1963. “Origine, sens et survie du term boécien ‘secundum placitum’.” Vivarium, 1: 87-113. Felski, Rita. 2011. “Context Stinks!” NLH, 42: 573-591. First Person Singular II: Autobiographies by North American Scholars in the Language Sciences. 1991. Ed. E. F. K. Koerner. Amsterdam: John Benjamins. First Person Singular III: Autobiographies by North American Scholars in the Language Sciences. 1998. Ed. E. F. K. Koerner. Amsterdam: John Benjamins. Fleischmann, Suzanne. 1990. “Philology, Linguistics, and the Discourse of the Medieval Text.” Speculum, 65: 19-37. Fleming, J. V. 1966. “The Antifraternalism of the ‘Summoner’s Tale’.” JEGP, 65: 688-700. Forrest, Ian. 2005. The Detection of Heresy in Late Medieval England. Oxford: Clarendon Press. Foucault, Michel. 1980. Power/Knowledge: Selected Interviews and Other Writings. Ed. Colin Gordon. New York: Pantheon. —. 1997. Ethics: Subjectivity and Truth. Ed. Paul Rabinow. Trans. Robert Hurley et al. Vol. 1. Essential Works of Foucault, 1954-1984. 3 vols. New York: The New Press. Fredborg, Karin M., J. Green-Pedersen, Lauge Nielsen, and Jan Pinborg. 1975. “The Commentary on ‘Priscianus Maior’ Ascribed to Robert Kilwardby.” CIMAGL, 15: 1-46. Fulton, Helen. 1996. “Punctuation as a Semiotic Code: The Case of the Medieval Welsh cywdd.” Parergon: Bulletin of the Australian and New Zealand Association for Medieval and Renaissance Studies, New Series, 13: 21-35. Genette, Gérard. 1980. Narrative Discourse: An Essay in Method. Trans. Jane Lewin. Ithaca: Cornell University Press. Gertz-Robinson, Genelle. 2005. “Stepping into the Pulpit? Women’s Preaching in The Book of Margery Kempe and The Examinations of Anne Askew.” In Voices in Dialogue: Reading Women in the Middle Ages. Ed. Linda Olson and Kathryn Kerby-Fulton. Notre Dame, IN: University of Notre Dame Press. 459-482. —. 2012. Heresy Trials and English Women Writers. Cambridge: Cambridge University Press.

Bibliogr aphy

257

Goffman, Ervin. 1967. Interaction Ritual. New York: Doubleday. —. 1974. Frame Analysis New York: Harper and Row. —. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press. Goubier, Frédéric. 2014. “Dire et vouloir dire dans la logique médiévale : Quelques jalons pour situer une frontière.” Methodos. Online: http://journals.openedition.org/methodos/3790. Accessed 15 July 2019. — and Nausicaa Pouscoulous. 2011. “Virtus sermonis and the Semantics-Pragmatics Distinction.” Vivarium, 49: 214-239. Gramsci, Antonio. 1971. Selections from the Prison Notebooks. Ed. and trans. Quintin Hoare and Geoffrey Nowell Smith. New York: International Publishers. Grice, H. Paul. 1975. “Logic and Conversation.” In Syntax and Semantics, vol. 3. Ed. Peter Cole and Jerry Morgan. New York: Academic Press. 41-58. Habermas, Jűrgen. 1979. Communication and the Evolution of Society. Trans. Thomas McCarthy. Boston: Beacon. —. 1998. “Was heißt Universalpragmatik?” In Sprachpragmatik und Philosophie. Ed. K.-O. Apel. Frankfurt am Main: Suhrkamp. 174-272. Hacking, Ian. 2002. Historical Ontology. Cambridge, MA: Harvard University Press. —. 2006. “Making Up People.” London Review of Books, 17 August. 23-26. Halliday, M. A. K. 1978. Language as Social Semiotic. London: Edward Arnold. Hobson, Jacob. 2018. “Ælfric’s Interjections: Learning to Express Emotion in Late Anglo-Saxon England.” Review of English Studies, New Series, 69: 815-831. Holsinger, Bruce. 2011. “Historical Context: Surface, Depth, and the Making of the Text.” NLH, 42: 593-614. Hoose, Adam. 2016. “The ‘Sabotati’: The Significance of Early Waldensian Shoes, c. 1184-c. 1300.” Speculum, 91: 356-373. Horn, Laurence. 1984. “A New Taxonomy for Pragmatic Inference: Q-based and R-based Implicature.” In Meaning, Form and Use in Context (GURT ’84). Ed. Deborah Schiffrin. Washington, D.C.: Georgetown University Press. 11-42. —. 2001. A Natural History of Negation. 2nd ed. Chicago: The University of Chicago Press. Howard, Donald. 1976. The Idea of the Canterbury Tales. Berkeley: University of California Press. Hudson, Anne. 1985. Lollards and Their Books. London and Ronceverte: The Hambledoon Press. —. 1986. “Wyclif and the English Language.” In Wyclif in his Times. Ed. Anthony Kenny. Oxford: Clarendon Press. 85-103. —. 1988. The Premature Reformation: Wycliffite Texts and Lollard History. Oxford: Clarendon Press. Hymes, Del. 1962. “The Ethnography of Speaking.” In Anthropology and Human Behavior. Ed. T. Gladwin and W. C. Sturtevant. Washington, D.C.: Anthropological Society of Washington. 13-53. Irvine, Martin. 1985. “Medieval Grammatical Theory and Chaucer’s House of Fame.” Speculum, 60: 850-876. Jager, Eric. 2014. Blood Royal: A True Crime and Detection in Medieval Paris. New York and London: Little, Brown. Jakobson, Roman. 1987. Language in Literature. Ed. Krystyna Pomorska and Stephen Rudy. Cambridge, MA: Harvard University Press. Janik, Allan and Stephen Toulmin. 1973. Wittgenstein’s Vienna. New York: Simon and Schuster.

258 

The Medieval Life of L anguage

Jucker, Andreas. 2000. “English Historical Pragmatics: Problems of Data and Methodology.” In English Diachronic Pragmatics. Ed. G. Di Martino and M. Lima. Napoli: Cooperativa Universitaria Editrice Napoletana. 17-55. —. 2012. “Pragmatics in the History of Linguistic Thought.” In The Cambridge Handbook of Pragmatics. Ed. Keith Allan and K. Jaszczolt. Cambridge: Cambridge University Press. 495-512. —. and Irma Taavitsainen. 2013. English Historical Pragmatics. Edinburgh: Edinburgh University Press. Jurowski, Maureen. 2002. “The Arrest of William Thorpe in Shrewsbury and the Anti-Lollard Statute of 1406.” Historical Research, 75: 273-295. —. 2007. “Lollardy and Social Status in East Anglia.” Speculum, 82: 120-152. Kelly, Louis. 1975. “Saint Augustine and Saussurean Linguistics.” Augustinian Studies, 6: 45-64. —. 1985. “St. Augustine’s Theories of the Linguistic Sign and the grammatica speculativa of the Thirteenth Century.” Studia Patristica, 16: 517-523. —. 2002. Mirror of Grammar: Theology, Philosophy and the Modistae. Amsterdam and Philadelphia: John Benjamins. Kerby-Fulton, Kathryn. 2005. “When Women Preached.” In Voices in Dialogue: Reading Women in the Middle Ages. Ed. Linda Olson and Kathryn Kerby-Fulton. Notre Dame, IN: University of Notre Dame Press. 31-55. King, Peter and Andrew Arlig. 2018. “Peter Abelard.” SEP (Fall 2018 Edition). Ed. Edward N. Zalta. Online: https://plato.stanford.edu/archives/fall2018/entries/abelard. Accessed 2 March 2019. Kneepkens, C. H. 1983. “The Quaestiones grammaticales of the Ms Oxford, Corpus Christi College 250: An Edition of the First Collection.” Vivarium, 21: 1-34. —. 1990. “On Medieval Syntactic Thought with Special Reference to the Notion of Construction.” Histoire, Epistémologie, Langage, 12: 139-176. —. 2013. “Robert Kilwardby on Grammar.” In A Companion to the Philosophy of Robert Kilwardby. Ed. Henrik Lagerlund and Paul Thom. Leiden: Brill. 17-64. Knuuttila, Simo. 1993. Modalities in Medieval Philosophy. London and New York: Routledge. Kristeva, Julia. 1981. Le langage, cet inconnu: Une initiation à la linguistique. Paris: Seuil. Kuhn, Thomas. 1996. The Structure of Scientific Revolutions. 3rd ed. Chicago and London: University of Chicago Press. Lacan, Jacques. 1977. Ếcrits: A Selection. Trans. Alan Sheridan. London: Tavistock Publications. —. 1993. The Seminar, Book 3: The Psychoses, 1955-1956. Trans. Russell Grigg. London: Routledge. Laclau, Ernesto and Chantal Mouffe. 2000. Hegemony and Socialist Strategy: Toward a Radical Democratic Politics. 2nd ed. London and New York: Verso. La Spiritualité de Pierre de Celle. 1946. Ed. Jean Leclerq. Paris: J. Vrin. Leech, Geoffrey. 2014. The Pragmatics of Politeness. Oxford: Oxford University Press. Leff, Gordon. 1967. Heresy in the Later Middle Ages: The Relation of Heterodoxy to Dissent, c. 1250-c. 1450. 2 vols. Manchester: Manchester University Press. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Libera, Alain. 1981. “Le traité De Appellatione de Lambert de Lagny (Lambert d’Auxerre).” Archives d’Histoire Doctrinale et Littéraire du Moyen Age, 48: 227-285. Lindahl, Carl. 1989. Earnest Games: Folkloric Patterns in the Canterbury Tales. Bloomington: Indiana University Press. Lovejoy, Arthur O. 1948. Essays in the History of Ideas. Baltimore: The Johns Hopkins University Press.

Bibliogr aphy

259

Magee, John. 1989. Boethius on Signification and Mind. Leiden: Brill. Maloney, Thomas. 1995. “Is the De doctrina christiana the Source for Bacon’s Semiotics?” In Reading and Wisdom: The De Doctrina Christiana of Augustine in the Middle Ages. Ed. Edward D. English. Notre Dame, IN and London: University of Notre Dame Press. Pp. 143-171. McNamer, Sarah. 2010. Affective Meditation and the Invention of Medieval Compassion. Philadelpha: University of Pennsylvania Press. Marenbon, John. 1997. The Philosophy of Peter Abelard. Cambridge: Cambridge University Press. —. 2013. “Gilbert of Poitiers’ Contextual Theory of Meaning and the Hermeneutics of Secrecy.” In Logic and Language in the Middle Ages: A Volume in Honour of Sten Ebbesen. Ed. Jakob Leth Fink, Heine Hansen, and Ana Maria Mora-Márquez. Leiden: Brill. 49-64. Martin, André. 2020. “Peter John Olivi on Sense Perception, Attention, and the Soul’s Orientation towards the Body.” In Medieval Perceptual Puzzles: Theories of Sense Perception in the Thirteenth and Fourteenth Centuries. Ed. Elena Băltuță. Leiden: Brill. 304-334. Meier-Oeser, Stephan. 2011. “Medieval Semiotics.” Stanford Encyclopedia of Philosophy. (Summer 2011 Edition). Ed. Edward N. Zalta. Online: https://plato.stanford.edu/archives/sum2011/ entries/semiotics-medieval. Accessed 10 October 2019. Mews, Constant. 2002. Reason and Belief in the Age of Roscellin and Abelard. Burlington, VT: Ashgate. —. 2005. Abelard and Heloise. Oxford: Oxford University Press. Mey, Jacob. 2013. “A Brief Sketch of the Historic Development of Pragmatics.” In Oxford Companion to the History of Linguistics. Ed. Keith Allan. Oxford: Oxford University Press. 587-611. Middle English Dictionary. Ed. Robert E. Lewis et al. Ann Arbor: University of Michigan Press, 19522001. In Middle English Compendium. Ed. Frances McSparran et al. Ann Arbor: University of Michigan Library, 2000-2018. Online: http://quod.lib.umich.edu/m/middle-english-dictionary. Accessed 1 February 2020. Monagle, Clare. 2017. The Scholastic Project. Kalamazoo, MI: ARC Humanities Press. Moore, R. I. 1987. The Formation of a Persecuting Society. Oxford and Cambridge, MA: Blackwell. Mora-Márquez, Ana Maria. 2011. “Pragmatics in Peter John Olivi’s Account of the Significance of Common Names.” Vivarium, 49: 150-164. Muscatine, Charles. 1957. Chaucer and the French Tradition. Berkeley: University of California Press. —. 1976. “The Social Background of the Old French Fabliaux.” Genre, 9: 1-19. Nerlich, Brigitte. 2006. “Pragmatics: History.” In Encyclopedia of Language and Linguistics, 2nd ed. Ed. Keith Brown et al. 14 vols. Oxford: Elsevier. 10: 37-44. Online: www.sciencedirect.com. —. and D. D. Clarke. 1996. Language, Action, and Context: The Early History of Pragmatics in Europe and America, 1780-1930. Amsterdam: John Benjamins. Nuchelmans, Gabriel. 1988. “The Distinction actus exercitus/actus significatus in Medieval Semantics.” In Meaning and Inference in Medieval Philosophy. Ed. Norman Kretzmann. Dordrecht: Kluwer. 57-90. Neuss, Paula. 1974. “Double Meanings: I. Double-Entendre in The Miller’s Tale.” Essays in Criticism, 24: 325-340. On the Medieval Theory of Signs. 1989. Ed. Umberto Eco and Costantino Marmo. Trans. Shona Kelly. Amsterdam: John Benjamins. Oxford English Dictionary. 2012. 3rd ed. Oxford: Oxford University Press. Online: https://www. oed.com. Accessed 1 February 2020.

260 

The Medieval Life of L anguage

Panaccio, Claude. 1991. Les mots, les concepts, les choses. Montréal and Paris: J. Vrin. Parkes, M. B. 1992. Pause and Effect: An Introduction to the History of Punctuation in the West. Aldershot, Hampshire: Scolar Press. Passnau, Robert. 1999. “Olivi on Human Freedom.” In Pierre de Jean Olivi (1248-1298): Pensée scolastique, dissidence spirituelle et societé. Ed. Alain Boureau and Sylvain Piron. Paris: Librairie Philosophique J. Vrin. 15-26. —. and Juhana Toivanen. 2018. “Peter John Olivi.” The Stanford Encyclopedia of Philosophy. Ed. Edward N. Zalta. Online: https://plato.stanford.edu/archives/win2018/entries/olivi. Patterson, Lee. 1978. “The ‘Parson’s Tale’ and the Quitting of the ‘Canterbury Tales’.” Traditio, 34: 331-351. —. 1991. Chaucer and the Subject of History. Madison: University of Wisconsin Press. Peikola, Matti. 2000. Congregation of the Elect: Patterns of Self-Fashioning in English Lollard Writings. Turku: University of Turku. Peirce, C. S. 1940. Philosophical Writings of Peirce. Ed. Justus Buchler. New York: Routledge and Kegan Paul. —. 1991. Peirce on Signs. Ed. James Hoopes. Chapel Hill: The University of North Carolina Press. Reddy, Michael. 1979. “The Conduit Metaphor: A Case of Frame Conflict in our Language about Language.” In Metaphor and Thought. Ed. Andrew Ortony. Cambridge: Cambridge University Press. 284-324. Rosch, Eleanor. 1975. “Cognitive Representations of Semantic Categories.” Journal of Experimental Psychology, 104: 192-233. Rosier (Rosier-Catach), Irène. 1983. La grammaire spéculative des Modistes. Paris: PUF. —. 1994. La parole comme acte: Sur la grammaire et la sémantique au XIIIe siècle. Paris: J. Vrin. —. 1999. “Modisme, pré-modisme, proto-modisme: vers de définition modulaire.” In Medieval Analyses in Language and Cognition. Ed. Sten Ebbesen and Russell Friedman. Copenhagen: C. A. Reitzels Forlag. 45-82. —. 2004. La parole efficace: Signe, rituel sacré. Paris: Seuil. —. 2010. “Speech Act and Intentional Meaning in the Medieval Philosophy of Language.” Bulletin de philosophie mediévale, 52: 55-80. —. and Laurent Cesalli. 2018. “‘Signum est in praedicamento relationis’: Roger Bacon’s Semantics Revisited in the Light of His Relational Theory of the Sign.” Oxford Studies in Medieval Philosophy, 6: 62-99. Sapiro, Gisèle. 2015. “Habitus: History of a Concept.” In International Encyclopedia of the Social and Behavioral Sciences. Ed. James D. Wright. 2nd ed. Amsterdam and Oxford: Elsevier. 484-489. Schegloff, Emanuel and Harvey Sacks. 1973. “Opening up Closings.” Semiotica, 7: 289-327. Schiffrin, Deborah. 1994. Approaches to Discourse. Oxford and Cambridge, MA: Blackwell. Schirmer, Elizabeth. 2009. “William Thorpe’s Narrative Theology.” Studies in the Age of Chaucer, 31: 267-299. Sidnell, Jack. 2010. Conversation Analysis: An Introduction. London: Wiley-Blackwell. Simone, Raffaele. 1975. “Theorie et Histoire de la Linguistique.” HL, 2: 353-378. Sirridge, Mary. 1980. “Jordanus, Notulae super Priscianum minorem Magistri Jordani.” CIMAGL, 36: 1-108. Somerset, Fiona. 1998. Clerical Discourse and Lay Audience in Late Medieval England. Cambridge: Cambridge University Press. —. 2002. “Excitative Speech: Theories of Emotive Response from Richard Fitzralph to Margery Kempe.” In The Vernacular Spirit. Ed. Renate Blumenfeld Kosinski, Duncan Robertson, and Nancy Warren. London: Palgrave. 59-79.

Bibliogr aphy

261

Sperber, Dan and Deirdre Wilson. 1996. Relevance: Communication and Cognition. Rev. ed. Hoboken, NJ: John Wiley. —. 2012. Meaning and Relevance. Cambridge: Cambridge University Press. Spruyt, Joke and Catarina Dutilh Novaes. 2015. “‘Those Funny Words’: Medieval Theories of Syncategorematic Terms.” In Linguistic Content: New Essays on the History of the Philosophy of Language. Ed. Margaret Cameron and R. J. Stainton. Oxford: Oxford University Press. 100-120. Strang, Barbara M. H. 1970. A History of English. London and New York: Routledge. Strohm, Paul. 1995. “Chaucer’s Lollard Joke: History and the Textual Unconscious.” Studies in the Age of Chaucer, 17: 23-42. —. 2006. England’s Empty Throne: Usurpation and the Language of Legitimation, 1399-1422. New Haven: Yale University Press. Studies in Medieval Linguistic Thought. 1980. Ed. E. F. Konrad Koerner, Hans-Josef Niederehe, and Robert H. Robins. Amsterdam: John Benjamins. Sullivan, Karen. 2011. The Inner Lives of Medieval Inquisitors. Chicago: The University of Chicago Press. Taavitsainen, Irma. 1994. “Subjectivity as a Text-type Marker in Historical Stylistics.” Language and Literature: International Journal of Stylistics, 3: 197-212. —. 1995. “Narrative Patterns of Affect in Four Genres of the ‘Canterbury Tales’.” Chaucer Review, 30:191-210. —. 1997. “Exclamations in Late Middle English.” In Studies in Middle English Linguistics. Ed. Jacek Fisiak. Berlin: De Gruyter Mouton. 573-608. —. and Andreas Jucker. 2010. “Trends and Developments in Historical Pragmatics.” In Historical Pragmatics. Ed. Irma Taavitsainen and Andreas Jucker. Berlin and New York: De Gruyter Mouton. 3-32. Thomas, Jenny. 1996. Meaning in Interaction: An Introduction to Pragmatics. London: Longman. Thomas, Margaret. 2011. Fifty Key Thinkers on Language and Linguistics. New York: Routledge. Thomson, J. A. F. 1989. “Orthodox Religion and the Origins of Lollardy.” History, 74: 39-55. Thurot, Charles. 1964. Extraits de divers manuscrits latins pour servir a l’histoire des doctrines grammaticales. Rpt. Frankfurt-am-Main: Minerva (first published 1869.) Traugott, Elizabeth Closs. 2006. “Historical Pragmatics.” In Handbook of Historical Pragmatics. Ed. Laurence Horn and Gregory Ward. Malden, MA and Oxford: Blackwell. Uckelman, Sara. 2013. “A Quantified Temporal Logic for Ampliation and Restriction.” In Medieval Supposition Theory Revisited. Ed. E. P. Bos. Leiden: Brill. 485-510. Van Dijk, Teun. 1993. “Principles of Critical Discourse Analysis.” Discourse and Society, 4: 249-283. Verscheuren, Jef. 1999. Understanding Pragmatics. London: Arnold. Voloshinov, V. N. 1973. Marxism and the Philosophy of Language. Trans. Ladislav Matejka and I. R. Titunik. New York: Seminar Press. Watson, Nicholas. 1995. “Censorship and Cultural Change in Late-Medieval England: Vernacular Theology, the Oxford Translation Debate, and Arundel’s Constitutions of 1409.” Speculum, 70: 822-864. Weinrich, Uriel. 1966. Languages in Contact. The Hague: Mouton. Wenzel, Siegfried. 1981-1982. “Notes on the Parson’s Tale.” Chaucer Review, 16: 237-256. —. 1994. Macaronic Sermons: Bilingualism and Preaching in Late-Medieval England. Ann Arbor: University of Michigan Press.

262 

The Medieval Life of L anguage

Wetherell, Margaret. 2012. Affect and Emotion: A New Social Science Understanding. London: SAGE. Wierzbicka, Anna. 2003. Cross-cultural Pragmatics: The Semantics of Human Interaction. 2nd ed. Berlin and New York: Mouton de Gruyter. Wooffitt, Robin. 2005. Conversation Analysis and Discourse Analysis. London: SAGE. Zittya, Penn. 1986. The Antifraternal Tradition in Medieval Literature. Princeton: Princeton University Press.

Index Index of Names Abelard, Peter (see also intentio) on intentio 40, 45, 72, 74, 86, 97, 105, 201-202 Aristotle 21, 24, 31, 39-40, 70, 86, 242, 244 Categoriae 48 De anima 39 De interpretatione 39-40, 48 Metaphysics 69 Arundel, Thomas (Archbishop of Canterbury) examination of William Thorpe 165-204 Austen, J.L. on speech acts 33 Bacon, Roger 40, 43-72 definition of sign 46-47 on actus exercitus 34 on equivocatio 67-69 on impositio 55, 61-69 on Relation 47-48 Bakhtin, Mikhail and dialogism 51, 57, 124-125, 139-142, 185 and reaccentuation 141-142, 148-149. 227 Boethius, Anicius Manlius Severinus commentary on Aristotle’s Categories 48, 69-70 De divisione 50, 63, 87, 249 on impositio 62-64 Boethius of Dacia on interiectio 100 Bourdieu, Pierre and ‘field’ 25 and structuring structures 237-238 on habitus 25 Channell, Joanna and vague language 209 Chaucer, Geoffrey Knight’s Tale 117-118 Merchant’s Tale 133-136 Miller’s Tale 129-130, 139-164 Monk’s Tale 120-122 Nun’s Priest’s Tale 136 Pardoner’s Prologue and Tale 124-128 Parson’s Tale 122-124 Reeve’s Tale 133 Troilus and Criseyde 118-120 Wife of Bath’s Prologue and Tale 130-133 Clanvowe, John 198 Derrida, Jacques on context 108 Diomedes (grammarian) on interiectio 88 Donatus, Aelius on interiectio 88, 90

Genette, Gérard and focalization 141 Goffman, Erving on footing 152, 167, 188-189 on frame 188 Gosvin of Marbais and performatives 34 Gramsci, Antonio 15-16 Grice, J. Paul 23, 28, 157, 218 on Cooperative Principle 77, 106 on pragmatic Maxims 165-166, 206 Gui, Bernard Practica and pragmatics 170-183 Habermas, Jürgen on speaking in good faith 76, 239 Hacking, Ian and ‘historical ontology’ 175, 242 Halliday, M.A.K. on context 107 Helias, Peter 68 on interiectio 59, 87, 93 Jakobson, Roman 246 Jordanus of Saxony 93-94, 96 Jucker, Andreas and history of pragmatics 19-20, 110 Kelly, Louis 46 Kempe, Margery and strategic pragmatics 205-239 Lacan, Jacques and point de capiton 159 Laclau, Ernesto (see also Chantal Mouffe) 25 antagonisms and crisis 242 theory of hegemony and alternate hegemony 237 Lambert of Ligny-le-Chatel (Auxerre) Logica 40-41, 64-66, 77 Magister Johannes (grammarian) 42, 94-95, 97-100, 125, 140, 167, 206, 244 Mey, Jacob and history of pragmatics 19-20 on Communicative Principle 78 Mouffe, Chantal (see Ernesto Laclau) Nicholas of Paris 35 De Syncategoreumata 36, 41-42 on transcasus 81-82 Nuchelmans, Gabriel 22, 33-34 Olivi, Peter (of) John 15, 26, 43, 72-83 and pragmatic ethics 79-80, 83 and Will (voluntas) 74-77 on context 74, 77-78, 82 on semantics 80-83

264  Peirce, C.S. 19-20, 108 and theory of signs 45-47, 49, 53, 59 Peter of Spain 34, 64, 77, 105, 246 and Summulae logicales 34, 36, 51-52, 65, 105 and Syncategoreumata 26, 34, 36, 57, 62, 71, 92, 146 on semantics 235 Pompeius (grammarian) on interiectio 89

The Medieval Life of L anguage

Priscian Caesariensis on interiectio 90-91, 146 Rosier (-Catach), Irène and history of pragmatics 18, 22-23, 33-34, 86 Schiffrin, Deborah 106 Searle, John 23 Taavitsainen, Irma and history of pragmatics 109-110, 137 Thorpe, William Testimony 183-204

Index of Subjects actus significatus/actus exercitus 33-39, 82, 95, 100, 108, 132, 243 ampliatio 43, 69-70, 235; see also restrictio analogia 43, 72 concept (conceptus) in language 36, 39-42, 45, 49-50, 65, 67, 71, 79-80, 99-100, 248 Conversation Analysis (CA) 12, 76, 106-107, 152, 166, 188, 206, 213 Critical Discourse Analysis (CDA) 12, 106-107, 166 Dialogism see Bakhtin equivocatio 26-27, 43, 51, 67, 69, 76, 81, 141-142, 149, 189, 191, 193, 196, 203, 215, 220, 227, 239, 246 face threatening acts (Goffman) 129, 141, 148, 150, 152-153, 169, 171, 184, 204, 206-207, 210, 213, 223, 226-227 flouts (Grice) 78 footing (Goffman) 152, 188, 208, 210 frame (Goffman) 188 and relevance 198, 213, 224, 230 habitus 60 hedging 148, 152-156, 164, 171, 214, 227, 229-231, 234-235, 239 impositio 43, 64-67; see also reimposition intentio 13, 26, 45, 72,76-78, 86, 171 and communi intentiones 75-76, 183 interiectio medieval grammatical theory of 85-101 natura in medieval pragmatics 53-55, 72, 244, 248-249 performatives see speech acts; Gosvin of Marbais

politeness (Brown and Levinson) 148-149, 207 polysemy 27-28, 50-51, 124, 147, 149, 157-158, 170, 173, 175, 177, 180, 182, 189, 196, 206, 209, 211, 221-222, 247 pragmatics and literature 246-247 reaccentuation (Bakhtin) 141-142, 148-149 reimposition and Bacon 66-67, 157, 221, 239 relevance and relevance frame 198, 213, 224, 230 Grice’s maxim of 216 Sperber and Wilson’s theory of 56-57, 206-208, 210, 223 restrictio 43, 58, 69, 235; see also ampliatio significatio 51, 245 signum conventional signs 45, 52, 97 medieval theories of 41, 44-48, 53, 55, 60, 62, 97, 134, 172 natural signs 45, 52 sinceritas 13, 26, 76, 82, 166 speech acts see actus significatus/actus exercitus suppositio 13, 26, 49-51, 63, 245, 248 T/V pronouns Middle English pragmatics of 199, 215 vague language (Channell) 209 and heretics 171, 173, 177, 181-182 and relevance 206-207, 209 Margery Kempe’s use of 205-239 vs ambiguity and equivocation 209-210 virtus sermonis 37, 58, 104, 157, 238, 245