This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Contents Introduction Isabel Moskowich and Begoña Crespo
7
Part 1. Linguistic aspects of early English The (im)possibility of stacking adjectives in Early English Agnieszka Pysz
15
Lists in letters: NP-lists and general extenders in Early English correspondence Ruth Carroll
37
Middle English medical books as examples of discourse colonies: G.U.L Hunter 307 Francisco Alonso-Almeida
55
The second-person pronoun in late medieval English drama: The York Cycle (c. 1440) Rosa Eva Fernández-Conde
81
Different paths for words and money: The semantic field of “Commerce and Finance” in Middle English Isabel Moskowich and Begoña Crespo
101
Part 2. Language and culture How might Everyman have been performed? John McKinnell
119
Shift of meaning in the animal field: Some cases of narrowing and widening Isabel de la Cruz-Cabanillas
139
Different aspects of the specialised nomenclature of ophthalmology in Old and Middle English María José Esteve-Ramos
151
Complex predicates in early scientific writing Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
169
Sixteenth-century glosses to a fifteenth-century gynaecological treatise (BL, MS Sloane 249, ff. 180v-205v): A scientifically biased revision Mª Victoria Domínguez-Rodríguez and Alicia RodríguezÁlvarez
179
Part 3. Philology and the study of medieval texts Rewriting eleventh-century English grammar and the editing of texts Donald Scragg
195
DCL, B IV, 24: A palaeographical and codicological study of Durham’s Cantor’s Book Francisco José Álvarez-López
209
The four-wheeled quadriga and the seven sacraments: On the sources for the ‘dedication’ of the Ormulum Nils-Lennart Johannesson
227
Verbal confrontation and the uses of direct speech in some Old English poetic hagiographies Juan Camilo Conde-Silvestre
247
Tolkien, medievalism, and the philological tradition Tom Shippey
265
Introduction* Isabel Moskowich and Begoña Crespo The delicate situation of the Humanities in Europe and elsewhere in recent times has persuaded us of the importance of a collection of papers showcasing present trends in English cultural studies, linguistic surveys and methodological tools. Though conscious of the many specialised conferences and publications already in existence, we felt it was not irrelevant to put together a miscellany of papers that would encompass a whole spectrum of interdisciplinary approaches. This collection contains fifteen essays, each of which represents a stimulating perspective on English, across a variety of current theoretical trends. In putting together Bells Chiming From The Past it has been our aim to provide the reader not only with what we regard as valuable samples of the research being conducted in the field of English medieval studies, but also with high-quality academic writing. Topics in the volume range from the most traditional philological to the most modern perspectives of linguistics today applied to early English texts and contexts. The absence of any narrow thematic restriction for this book has actually proved a blessing in disguise, allowing academic excellence and innovative character to serve as our only criteria in selecting the articles. We believe that to understand the characteristics of presentday English language and culture we should have some knowledge about the earlier stages of language use and this volume traces the early phases of the development of English, investigating diverse *
The collection of essays here presented has been partially funded by the Spanish Ministry of Education and the University of A Coruña. These grants are hereby gratefully acknowledged.
8
Isabel Moskowich and Begoña Crespo
aspects of the theme in original work on English Historical Linguistics and related issues. Most of the papers selected are based on empirical research and represent theoretical and descriptive contributions over a broad range of topics to our knowledge of English, both written and spoken. We maintain that studying the past of a language improves competence in that language and its culture and favours a better understanding of the present. Lessons learnt from past experiences are always useful in trying to comprehend the present: How do we explain the presence of so many French borrowings in English for Specific Purposes? Why do we use you instead of thou? Why do we feed pigs but eat pork? History (in its primitive Greek sense of research understanding) answers these questions. What appear to be exceptional or rare cases turn out actually not to be so, if we look into their past. The general focus of the book is the relationship and interaction of language and culture during the Middle English period. However, the contents of the book slot comfortably into three segments. The first group of articles could be described as linguistically oriented. A second group looks at both language and the cultural milieu in which linguistic facts took place, while a third fits into a wider philological school of thought. Part 1, “Linguistic aspects of early English”, encompasses papers based mainly on the application of current linguistic trends and research techniques to earlier stages of the English language. Pursuing this approach, we find the papers by Pysz, Carroll, Alonso-Almeida and Fernández-Conde. They all resort to the analysis of data to describe different aspects of the linguistic situation in medieval England. While some also mention extralinguistic factors, the theme is touched on only in passing by comparison with the essays in part 2. Syntactic concerns are common to the papers of Ruth Carroll (U. Turku) and Agnieszka Pysz (Adam Mickiewicz U.) while discourse analysis provides the linguistic framework for those of Francisco Alonso-Almeida (U. Las Palmas de Gran Canaria) and Rosa Eva Fernández-Conde (U. Oviedo). Ruth Carroll’s work examines elided NP elements such as verb objects and begins an inventory of those forms together with additional comments. Her findings are part of a more comprehensive research project on general extenders in the
Introduction
9
history of English. Pysz’s article focuses on the multiple occurrence of adnominal adjectives in early English prose texts taken from the YorkToronto-Helsinki Parsed Corpus of Old English Prose. The articles on discourse analysis include the interpretation of Middle English medical manuscripts as discourse colonies, as part of a larger, more ambitious project (Alonso-Almeida), and an analysis of the powerbased relationships that can be inferred from the use of the secondperson pronoun in The York Cycle (Fernández-Conde). The essays in “Language and culture” (part 2) look at the interaction between external events and linguistic facts. John McKinnell’s (Durham U.) paper presents some thoughts about the production of Everyman on the back of its performance by the Durham Medieval Theatre Company in June 1999 and considers their success in recreating as far as possible the effect the play might have had for its first audiences, both in actual performance and on the contemporary imagination. Isabel de la Cruz’s (U. Alcalá) study is devoted to semantic change in the animal lexical field. To account for the type of change discussed in each instance, she looks at the sociohistorical environment in which particular shifts took place. We cannot forget, of course, that it is in the lexicon that the relationship between language and society is most easily observed. In the same vein, Begoña Crespo and Isabel Moskowich’s contribution explores the two-way nature of the interaction between language and history in relation to the lexical field of commerce and finance. The next three papers featured in this section relate to the scientific register, a topic which is at the forefront of historical linguistic research. María José Esteve (U. Jaume I) studies the vocabulary of ophthalmology texts in a paper introduced by an external contextualisation of the theme, and highlights the lexical choices and the etymological information which characterise this type of text. The next offering on this topic is Nuria Bello-Piñón and Elva Méndez-Souto’s (U. A Coruña) linguistic study of complex predicates. The paper includes a categorisation of the predicates from a syntactic point of view and provides details of the prevailing etymological origin of the accompanying nouns. Finally, Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez (U. Las Palmas) discuss a revision of the fifteenth-century gynaecological treatise The
10
Isabel Moskowich and Begoña Crespo
Sekenesse of Wymmen, with specific attention to the theme of lexical substitutions. The last part in the book is devoted to “Philology and the study of medieval texts”. Philology is understood as the scientific study of language through texts, tracing linguistic developments over time and investigating the related literary and cultural phenomena. In this third part, Donald Scragg (U. Manchester) demonstrates the value of using diverse editions as a source of new grammatical hypotheses. Old English grammars focus on the patterns of early West-Saxon manuscripts and regard all others as “dialectal”. Scragg manifests his disagreement with the criteria of those canonical grammars, especially taking into account the variation in the period which is not recorded in those manuscripts. Francisco José Álvarez-López (U. Vigo/Manchester) studies the main palaeographical and codicological features of the tenthcentury manuscript, Durham’s Cantor’s Book, looking in particular at the Old English copy of the Rule of Saint Benedict as its possible origin, and at its relation to the original translation of the Latin Regula, probably by Æthelwulf. Also working with older texts, Nils-Lennart Johannesson (U. Stockholm) provides an interesting description of Orm’s manuscript. His paper focuses on the “Dedication” and examines the sources from which The four-wheeled quadriga and the seven sacraments image may have come. A number of texts are proposed as possible origins in the purest philological style. Juan Camilo Conde-Silvestre (U. Murcia) explores the possible parallels between heroic flyting contests and rhetorical duelling in three poetic hagiographies, and appraises the role of eloquence as a measure of dignity in the characterisation of both heroes/saints and their antagonists. Finally, the philological paper par excellence is Thomas Shippey’s (St. Louis U.) discussion on Tolkien, medievalism and the philological tradition. Shippey analyses the way in which this modernday author used his knowledge of medieval culture and mythology to create an imaginary world in his literary work. The main idea behind the article is the importance of Tolkien’s philological background to his development as a creative writer. In line with our claim about the crisis in the Humanities, Shippey criticises the fact that university interest in this topic has thus far fallen short of that shown by the general public, and that the growing love of books, video games and
Introduction
11
films set in the Middle Ages and worlds of ancient wisdom and knowledge of traditions is not matched in the academic world. Many hours and more effort have gone into compiling and editing the works here presented. Certain elements of the task, involving tables and graphs, special fonts and special characters, posed something of a challenge to our modest skills. However, since one of the main interests of the editors has been less to provide a large number of essays than to achieve a volume that is consistent in quality, the papers ensuing have been treated carefully and painstakingly to ensure that no detail has been overlooked. Any mistakes or misprints are, therefore, unintended, and we hope the reader will regard them as such. As editors we wish to thank all the authors for their kind participation in this volume, those scholars responsible for peerreviewing the articles here collected, and Ms Aoileann Lyons and Mr Gonzalo Camiña for their conscientious and invaluable help.
Contents Introduction Isabel Moskowich Spiegel-Fandiño and Begoña Crespo García Part 1. Linguistic aspects of early English The (im)possibility of stacking adjectives in early English Agnieszka Pysz Lists in letters: NP-lists and general extenders in early English correspondence Ruth Carroll
Part 1 Linguistic aspects of early English
The (im)possibility of stacking adjectives in Early English1 Agnieszka Pysz Adam Mickiewicz University
Abstract This paper discusses the issue of adjective stacking in early English, specifically in the Old English (OE) period. It presents two of the main approaches to stacking, i.e. by Spamer (1979) and by Fischer (2000, 2001), according to which OE adjectives (either strong only or both weak and strong) are constrained by the rule of nonrecursiveness. The paper sets these approaches against a selection of OE data, which show that neither Spamer’s nor Fischer’s account stands up to empirical scrutiny. An alternative, ‘liberal-radical’ approach is therefore suggested, which revises the traditional views on stacking OE adjectives.
1. Introduction The starting point for this paper is the claim advanced by Spamer (1979), according to which the occurrence of adnominal adjectives in Old English (henceforth OE) is constrained by the rule of nonrecursiveness. Spamer’s proposal rests on the assumption that the traditional distinction between weak and strong adjectives2 in Germanic stems from a misconception. In his view, weak adjectives do not count as adjectives proper but belong to the group of elements 1
Earlier versions of this paper were presented at the 17th SELIM International Conference in A Coruña (2005) and the 4th Medieval English Studies Symposium in PoznaĔ (2005). I would like to thank the audiences for their feedback. 2 In the remainder of the paper the labels ‘weak adjectives’ and ‘strong adjectives’ will be employed as shorthand for adjectives with weak inflection and adjectives with strong inflection, respectively.
16
Agnieszka Pysz
which he refers to as adjuncts. Strong adjectives, on the other hand, are bona fide adjectives and fall under the rule of non-recursiveness, which precludes the appearance of more than one adjective in a single nominal phrase. The question of the (im)possibility of stacking adjectives in OE has recently been taken up by Fischer (2000, 2001), who suggests that the rule of non-recursiveness might be extended to weak adjectives (i.e. Spamer’s adjuncts) as well. The aim of this paper is to reconsider the issue of the multiple occurrence of adnominal adjectives in early English prose texts. The discussion focuses on nominal constructions involving more than one adnominal adjective. The empirical data are drawn from the YorkToronto-Helsinki Parsed Corpus of Old English Prose YCOE (Taylor, Warner, Pintzuk and Beths 2003). The paper is organised as follows. Section 2 presents an outline of the phenomenon of adjective stacking in OE. Following a general overview (section 2.1) it presents two specific approaches to the issue in question (section 2.2) and the problems posed by each of them (section 2.3). Section 3 contains the results of an empirical analysis of nominal phrases with two or more adjectives occurring in an uninterrupted string. Finally, section 4 reconsiders the issue of stacking and suggests an alternative approach to the multiple occurrence of adjectives in OE. The alternative approach takes into account the linguistic data presented in section 3 and departs from the two particular approaches reviewed in section 2.2. 2. Adjective stacking in OE: Standard views 2.1. General remarks The first point that needs to be made with regard to adjective stacking in OE concerns the permissible number of adjectives per nominal phrase. It has been widely acknowledged that OE nominal phrases do not abound in adjectival modifiers. In the vast majority of cases, a head noun is modified by one adjective; modification by more than one adjective occurs on a considerably smaller scale. This empirical observation is usually interpreted as a symptom of constraints on the number of adjectives which may occur in a nominal phrase. The second point, inextricably connected with the first, concerns the nature of stacking. OE adnominal adjectives are usually
The (im)possibility of stacking adjectives in Early English
17
characterised in terms of their highly limited potential for being stacked, as a result of which they hardly ever occur one after another in an uninterrupted string. It is generally maintained that whenever a given nominal phrase contains more than one adjective they cannot be situated next to each other. Instead, it becomes necessary to place them in a discontinuous string, which can be achieved in one of the following ways: (a) splitting adjectives on both sides of the head noun, as in (1) (b) separating adjectives by means of a conjunction, as in (2) (c) combining strategies (a) and (b), as in (3) mid soðum geleafan untweogendum ‘with true staunch faith’ (Fischer 2001: 264, her 17c) bill…brad and brunecg ‘broad and bright-edged sword’ (Quirk and Wrenn 1957: 88) berenne kyrtel oððe yterenne ’a bear- or otter-skin tunic’ (Traugott 1992: 283, her 286) At a more individual level, there is no unanimity among scholars on the issue of whether the ban on stacking constrains the distribution of all adnominal adjectives in OE or whether some more detailed qualifications should be made in this respect. In the next section, we will look at two proposals that have been put forward in relation to this point: Spamer (1979) and Fischer (2000, 2001). 2.2. The two views on adjective stacking in OE This section presents two specific views on the issue of adjective stacking in OE. The crux of Spamer’s (1979) proposal is discussed in section 2.2.1, whereas that of Fischer (2000, 2001) is discussed in section 2.2.2. 2.2.1. Spamer (1979) On the face of it, the proposal advanced by Spamer (1979) seems straightforward in so far as it states that the impossibility of stacking applies to all OE adjectives, with no exceptions. This unequivocal
18
Agnieszka Pysz
generalisation, however, turns out to be more complex if we consider that what Spamer refers to as ‘adjectives’ is not synonymous with what traditional OE grammars mean when they employ the same term. Crucially, Spamer explicitly rejects the traditional division of Germanic adjectives into strong and weak inflectional patterns. For him, the elements which most grammarians regard as strong and weak adjectives do not actually belong to the same categorial class. In Spamer’s approach, the term ‘adjective’ is reserved for what traditional grammars regard as strong adjectives, in addition to demonstrative pronouns3 and genitives (either pronominal or substantive).4 Weak adjectives, on the other hand, are not counted as adjectives proper, but as belonging to the group of substantive elements referred to as adjuncts.5 In light of the above, Spamer’s statement that the ban on stacking applies to all OE adjectives acquires a different meaning. Given, firstly, that the group of true modifiers includes strong adjectives, demonstrative pronouns and genitives6 and, secondly, that true modifiers cannot be iterated, it follows that nominal phrases containing an uninterrupted string of modifiers should be precluded by 3
Spamer’s arguments for treating demonstratives and strong adjectives as members of the same categorial class fall into two groups: inflectional and distributional. The former rely on similarities between the desinences of the two paradigms. The latter are determined by the impossibility of co-occurrence. 4 Although the morphology of ‘genitives’ resembles neither strong adjectives nor demonstrative pronouns, Spamer regards them as adjectives for the following reasons: (i) genitives are restricted to the initial position (ii) the use of genitives precludes the use of other ‘true adjectives’ (i.e. traditionally, strong adjectives and demonstratives) That the argument in (i) does not stand up to scrutiny is shown by the existence of patterns such as (5), i.e. sAdj+Poss and (8), i.e. Dem+Poss. The argument in (ii) is, in turn, belied by the existence of patterns such as (5), i.e. sAdj+Poss, (8), i.e. Dem+Poss, (10), i.e. Poss+sAdj or (11), i.e. Poss+Dem. 5 Fischer (2000: 178, fn. 8) says that “Spamer’s use of the term ‘adjunct’ is rather idiosyncratic, and is not really described clearly in his article” (cf. Fischer 2000: 163, 2001: 258). Indeed, Spamer’s discussion of adjuncts is not particularly expansive. He restricts himself to enumerating and comparing the basic properties shown by adjuncts in OE and PDE. We should clarify that the meaning with which Spamer uses the label ‘adjunct’ is not synonymous with the very same term used in the generative framework of the X-bar Theory of phrase structure. 6 From now on, we shall disregard substantive genitives and restrict ourselves to pronominal genitives (labelled here as Poss).
The (im)possibility of stacking adjectives in Early English
19
OE grammar. Thus, Spamer’s approach predicts that the strings in (412) are ungrammatical. (4) *sAdj+Dem (5) *sAdj+Poss (6) *sAdj+sAdj (7) *Dem+sAdj (8) *Dem+Poss (9) *Dem+Dem (10) *Poss+sAdj (11) *Poss+Dem (12) *Poss+Poss On the other hand, given that OE weak adjectives are not adjectives proper, the ban on stacking does not apply to them, so they can freely appear one after another, as in (13). What is more, the sequence of a weak adjective and an adjective proper is perfectly acceptable on the assumption that weak adjectives are adjuncts and may therefore licitly appear in the immediate vicinity of a true modifier. Consequently, the schematic constructions in (14-19), which contain an adjunct and an adjective, are grammatical according to Spamer’s approach. (13) wAdj+wAdj (14) wAdj+sAdj (15) sAdj+wAdj (16) wAdj+Dem (17) Dem+wAdj (18) wAdj+Poss (19) Poss+wAdj 2.2.2. Fischer (2000, 2001) Let us consider briefly the stance on the issue of OE adjective stacking that has been expressed by Fischer (2000, 2001).
20
Agnieszka Pysz
On the one hand, Fischer agrees with Spamer’s proposal that what are traditionally called strong adjectives cannot be stacked recursively. On the other hand, however, she argues against the idea of weak adjectives being recursive. Examples (20-22) can be adduced to illustrate her point (examples taken from Fischer 2000: 168, Haumann 2003: 63-64): (20) Se legfamblawenda seað & se fula…wæs helle tintreges muð ‘That vomiting-fire hole and that foul…was hell’s torture’s mouth’ (21) for þære micclan and stiðan drohtnunge ‘because-of the intense and severe way-of-life’ (22) þæt bio sio soðe & sio fulfremede gesælð ‘that be the true and the perfect fortune’ In (20), there are two weak adjectives, one of which occurs before the noun seað and the other after it. In (21-22), two weak adjectives precede the noun, but they do not occur in the immediate vicinity of each other: the conjunction intervenes. If the data just cited are set against Spamer’s proposal, one might rightly wonder why the weak adjectives in (20-22) occur in a discontinuous string. Given that adjuncts can be iterated freely, there should be no reason (barring some extragrammatical motivation) for which these adjectives could not appear one after another. In Fischer’s opinion, examples such as those in (20-22) are sufficiently indicative to warrant the claim that in OE weak (i.e. adjunctive) adjectives cannot be stacked either.7 She, therefore, proposes that the ban on stacking applies equally to strong and weak adjectives. 2.3. Problems with the two views on adjective stacking in OE In this section we will see that neither Spamer’s nor Fischer’s approach to the issue of adjective stacking yields satisfying results when confronted with OE data. Section 2.3.1 focuses on pinpointing 7
Fischer (2000: 171) says that weak (i.e. adjunctive) adjectives are difficult to string together owing to their compound characteristics.
The (im)possibility of stacking adjectives in Early English
21
the problems posed by Spamer’s approach, while section 2.3.2 concentrates on the weaknesses inherent in Fischer’s approach. 2.3.1. Problems with Spamer (1979) Spamer’s proposal can be undermined on two fronts. The first includes less robust arguments, which cast more or less serious doubts on the Spamerian approach. Three of them have already been mentioned in section 2.2.2 (examples 20-22). Three others appear in Fischer (2000: 166-168), where she considers the following constructions: (23) þone ilcan ceaddan iungne ‘the same Chad young’ (Fischer 2000: 167, her 7) (24) gyldenne wingeard trumlicne and fæstlicne ‘golden vineyard durable and firm’ (Fischer 2000: 167, her 8) (25) gersuman unateallendlice ‘treasures uncountable’ (Fischer 2000: 167, her 9d) Regarding the construction in (23), Fischer argues that there should be no need to place a strong adjective after the noun. Crucially, according to Fischer (2000: 173) adjectives such as ylca and agen/self form a unit with the determiner (cf. ME thilke) so in example (23) [þone ilcan] is treated as an indivisible element. When it comes to the construction in (24), Fischer says that a conjunction is needed after the noun in order to separate it from the strong adjective. As regards the construction in (25), Fischer sees no reason for placing an adjective after the noun in the absence of any prenominal adjective. We will not be taking a stance on Fischer’s line of reasoning adduced above. However, the second group of arguments seems to be more damaging for Spamer’s view on adjective stacking. Above all, his proposal whereby OE adjectives are treated as non-recursive elements envisages that OE grammar should rule out any sequence containing a string of two adjectives in a row. Thus, as mentioned in section 1.2.1, all 9 sequences in (4-12) should be excluded on grounds of their being ungrammatical. Yet, out of the 9 patterns which his proposal rules out, only one (i.e. pattern 12) is not attested in OE texts.
22
Agnieszka Pysz
Ad (4) sAdj+Dem (26) & he geseah Drihten on [ufeweardre þære] hlæddre, NODE (11 NP-DAT (12 ADJ^D ufeweardre) (13 D^D þære) (14 N^D hlæddre)) (ID cootest,Gen:28.13.1157)) Ad (5) sAdj+Poss (27) Ic ðæs þoncunge do Greca herige & swyðost þæm mægene þære iuguþe & þæm [unforswyþdum urum] weorode, for þon on ieþum þingum hie me mid wæron & on þæm earfeðum no from bugon. (NODE (27 NP-DAT (28 D^D þæm) (29 ADJ^D unforswyþdum) (30 PRO$^D urum) (31 N^D weorode)) (ID coalex,Alex:5.1.19)) Ad (6)8 sAdj+sAdj (28) & þæt is þæt [riht cristen] mann ne beode ænigum oðrum butan þæt he wille þæt man him beode. (NODE (9 NP-NOM (10 ADJ^N riht) (11 ADJ^N cristen) (12 N^N mann)) (ID cowulf,WHom_18:141.1507)) Ad (7) Dem+sAdj (29) & cwæd luddor stefne. ðancod wurð hit [þon hæge] ælmihti God þis wurðscipe þæt her is gedon. (NODE (15 NP-DAT (16 D^I þon) (17 ADJ^D hæge) (18 ADJ^D ælmihti) (19 NR^D God)) (ID cochronE-INTERPOLATION,ChronE_[Plummer]: 656.35.413)) 8
Fischer (2000: 163-164; her Table 2, row 6) finds 8 examples of [sAdj+sAdj+N], such as (i) and (ii) below. (i) mid ofermæte unclæne luste (Fischer 2000: 165, her 6a) (ii) of surre rigenre grut (Fischer 2000: 165, her 6b) There are also 2 tokens which feature the sequence [Q+sAdj+sAdj] (see Fischer 2000: 164; her Table 2 row 7).
The (im)possibility of stacking adjectives in Early English
23
Ad (8) Dem+Poss (30) Him þa [se heora] arwyrða bisceop eadiglice & halwendlice geðeaht forðbrohte, (NODE (7 NP-NOM (8 D^N se) (9 PRO$ heora) (10 ADJ^N arwyrða) (11 N^N bisceop)) (ID coblick,LS_25_[MichaelMor[BlHom_17]]:205.153.2624)) Ad (9) Dem+Dem (31) Hy habbaþ [þæs þe] leohtran gang. (NODE (6 NP-ACC (7 ADJP-ACC (8 D^G þæs) (9 D^I þe) (10 ADJR^A leohtran)) (11 N^A gang)) (ID coquadru,Med_1.1_[de_Vriend]:4.15.176)) Ad (10) Poss+sAdj (32) Se ðe slea [his agenne] þeowne esne oððe his mennen, & he ne sie idæges dead, ðeah he libbe twa niht oððe ðreo, ne bið he ealles swa scyldig, forþon þe hit wæs his agen fioh. (NODE (13 NP-ACC (14 PRO$ his) (15 ADJ^A agenne) (16 ADJ^A þeowne) (17 N^A esne)) (ID colawafint,LawAfEl:17.43)) Ad (11) Poss+Dem (33) & þær is ece blis & ece gefea mid þam ecan fæder & mid ðam efenecan suna & mid [his þam] efenecan haligan gaste a butan ende. (NODE (34 NP-DAT (35 PRO$ his) (36 D^D þam) (37 ADJ^D efenecan) (38 ADJ^D haligan) (39 N^D gaste)) (ID coverhom,HomS_38_[ScraggVerc_20]:199.A.2658)) The data adduced above speak volumes and clearly demonstrate that, after all, the apparent non-recursiveness of OE adjectives is not as radical as Spamer would have us believe. Naturally, one might say in Spamer’s defence that the overall number of constructions like those exemplified by (26-33) may be statistically negligible. Yet, even if their occurrence is indeed marginal, the fact that such sequences are encountered at all cannot simply be brushed aside.
24
Agnieszka Pysz
2.3.2. Problems with Fischer (2000, 2001) By way of recall, in the view proposed by Fischer (2000, 2001) the ban on stacking applies not only to strong adjectives but is extended to weak adjectives. This claim poses some not insignificant problems given that the occurrence of constructions involving an uninterrupted string of weak adjectives is well attested in OE prose. Witness the following examples (taken from Haumann 2003: 64): (34) þæt ofstandene þicce sliþige horh ‘the remaining thick slimy mucus’ (35) wæs writere on þysre halgan Romaniscean cyrcean ‘was writer on this holy Roman church’ To account for such examples, Fischer resorts to an explanation according to which weak adjectives are allowed to occur in a stack provided that they form an idiomatic expression (Fischer 2000: 173, after Haumann 2003: 64, fn. 20). On the basis of this argument, the reason why the uninterrupted sequence of weak adjectives in examples (34) and (35) does not result in ungrammaticality is that the sequences þicce sliþige horh ‘thick slimy mucus’ and halgan Romaniscean cyrcean ‘holy Roman church’ constitute fixed expressions. Nevertheless, Fischer appears to handle the issue of weak adjective stacking on an ad hoc basis. One might reasonably ask what criteria a given expression must meet in order to be licitly regarded as idiomatic. Even if one were willing to consider the expressions in (34) and (35) idiomatic, it is not immediately obvious on what grounds expressions like those in (36) and (37) below might be regarded as such: (36) On þam æftemestan mæran freolsdæge stod se hælend (NODE (5 NP-DAT (6 D^D þam) (7 ADJS^D æftemestan) (8 ADJ^D mæran) (9 N^D freolsdæge)) (ID cowsgosp,Jn_[WSCp]:7.37.6335)) (37) Ĉæt clæne hwætene corn þe Crist þa embespæc tacnæð hine sylfne þe sealde his lif for us,
The (im)possibility of stacking adjectives in Early English
25
(NODE (3 NP-NOM (4 D^N Ĉæt) (5 ADJ^N clæne) (6 ADJ^N hwætene) (7 N^N corn) (8 CP-REL (9 WNP-1 0) (10 C þe) (11 IP-SUB (12 NP *T*-1) (13 NP-NOM (14 NR^N Crist)) (15 ADVP-TMP (16 ADV^T þa)) (17 RP+VBDI embespæc)))) (ID covinceB,[Vincent]:304.14)) No attempt will be made here to render a verdict on whether a given expression may be judged to be idiomatic. It is highly unlikely, though, that every single construction in a fairly numerous group of phrases containing a stack of two weak adjectives (in my corpus I have found 103 examples) will qualify as an idiomatic or fixed expression. 3. The data This section contains the results of an empirical analysis of the constructions containing a stack of adnominal adjectives. The data are taken from the York-Toronto-Helsinki Parsed Corpus of Old English Prose YCOE (2003), which contains complete versions of the main OE prose texts as well as selected fragments from specialised genres. The full list of texts is provided at the end of the paper. Lack of space and time preclude a detailed examination of all the constructions with a stack of adjectives. We shall therefore restrict ourselves to adumbrating the most prevalent tendencies and the most untypical phrases in so far as they are likely to have a bearing on the results of the current study. The parsed files of the corpus have been searched automatically using the CorpusSearch (Version 1.1) search engine. The query was formulated in such a way as to retrieve all nominal phrases in which there is a sequence of at least two adjectives occurring side by side. The query was as follows:
26
Agnieszka Pysz
(38) node: NP* query: (ADJ* iPrecedes ADJ*) where a node specifies the search domain and a query consists of one search function (iPrecedes), taking two search terms as its arguments (ADJ*, ADJ*).9 The output of the search resulted in a total of 465 hits and 458 tokens containing the relevant items. Closer scrutiny revealed that some of the data which matched the query had to be eliminated because, for instance, the adjectives were not accompanied by a noun, as in (39), (39) and hi hæfdon twegen suna, me and oþerne gingran, (NODE (15 NP-ACC (16 ADJ^A oþerne) (17 ADJR^A gingran)) (ID coeust,LS_8_[Eust]:321.342)) or because there were some mistakes in annotation, as in (40), where a demonstrative pronoun is wrongly tagged as an adjective (40) Hwæt $is $ðis deadlice lif elcor nymðe hit is se weg þe we sculon $on $faran æghwylc man to swa hwæðerum swa he sylfa þurh $his $gewyrhtum $sie $gedemed & þurh his agene dæde geearnað, swa to $ecum $life & to ecum gefean, swa to ecum deaðe & to ecre $forwyrde. (NODE (8 NP-NOM (9 ADJ^N $ðis) (10 ADJ^N deadlice) (11 N^N lif)) (ID coverhom,HomM_11_[ScraggVerc_14]:26.1784)) or because a noun had been mistakenly labelled as an adjective, as in (41). (41) Ac se ælmihtiga God geeowde, hu myccles mægnes he wæs (NODE (4 NP-NOM (5 D^N se) (6 ADJ^N ælmihtiga) (7 ADJ^N God)) (ID cogregdC,GDPref_and_3_[C]:6.187.18.2352)) 9
‘*’ used in the query denotes a wild card, representing any character or sequence of characters in the search. Thus, ‘ADJ*’ stands for any label which starts with ‘ADJ’, including ‘ADJ’ itself; in other words, ‘*’ stands for anything and nothing. In all other cases ‘*’ is used to indicate ungrammaticality.
The (im)possibility of stacking adjectives in Early English
27
These instances amounted to 152 constructions. Having sieved them out, the corpus of data shrank to 313 constructions of the relevant type. These constructions were, in turn, divided into three groups: (a) constructions involving a stack of two strong adjectives, as in (42) (b) constructions involving a stack of two weak adjectives, as in (43) (c) the remaining constructions (42) sAdj+sAdj+N And rice hæþene men hine snidon þæt he dranc ator, on þam was ælces cynnes wirm oððe ban oððe blod, (NODE (4 NP-NOM (5 ADJ^N rice) (6 ADJ^N hæþene) (7 N^N men)) (ID comart1,Mart_1_[Herzfeld-Kotzor]:De27,A.5.93)) (43) wAdj+wAdj+N Ac se gooda heofenlica fæder: forgifð us geleafan. & hiht. & þa soðan lufe: (NODE (4 NP-NOM (5 D^N se) (6 ADJ^N gooda) (7 ADJ^N heofenlica) (8 N^N fæder)) (ID cocathom1,ÆCHom_I,_18:322.150.3542)) The group subsumed under (c) embraces constructions of three subtypes, namely: (ci) constructions involving a stack of adjectives which cannot be definitely classified as strong or weak, as in (44)10 (cii) constructions involving one strong and one weak adjective, in either order, as in (45) (ciii) constructions involving a stack of three adjectives, as in (46) and (47) (44) oAdj+oAdj+N Ĉonne synd þis þa land þe minæ yldran becwædon into oþrum halgum stowum, þæt is þonne into Cantwarabyrig to Cristæs circan þan hired to brece þes landes æt Illanlege & into Paules mynstre into Lundene þes landes æt Hedham to biscophame & es 10
This subgroup includes constructions in which none of the adjectives can be definitely classed as strong/weak and those in which only one adjective does not conform to the strong/weak inflectional pattern.
28
Agnieszka Pysz
landes æt Tidwoldingtune þan hirede to brece into Paules mynstre & into Beorcingan þam hirede to brece þes landes æt Babbingþyrnan. (NODE (23 NP-DAT (24 ADJ^D oþrum) (25 ADJ^D halgum) (26 N^D stowum)) (ID codocuæ,Ch_1486_[Whitelock_15]:19.287)) (45) sAdj+wAdj+N Raþe æfter þæm on þara twegea consula dæge, Claudius, þe oðre noman hatte Marcellus, & Ualerius, þe oðre noman hatte Flaccus, þa gewearð hit, þeh hit me scondlic sie, cwæð Orosius, þætte sume Romana wif on swelcum scinlace wurdon, & on swelcum wodan dreame, þæt hie woldon ælcne mon, ge wif ge wæpned, þara þe hie mehton, mid atre acwellan, & hit on mete oþþe on drynce to geþicgenne gesellan. (NODE (95 NP-DAT (96 ADJ^D swelcum) (97 ADJ^D wodan) (98 N^D dreame)) (ID coorosiu,Or_3:6.60.14.1159)) (46) Adj+Adj+Adj+N We habbað nu gesæd be ðam circlicum bocum on þære ealdan æ & eac on þare niwan: ða synd þa twa gecyðnyssa be Cristes menniscnysse & be þære halgan þrinnysse on soðre annysse, swa Isaias geseah on his gastlican gesihðe, hu God sylf gesæt & him sungon abutan duo seraphin, þæt sind twa engla werod: Sanctus, sanctus, sanctus, Dominus Deus Sabaoth, þæt ys on Englisc: Halig, halig, halig, Drihten weroda God. (NODE (142 NP-NOM (143 ADJP-NOM (144 ADJ^N Halig) (145 , ,) (146 ADJ^N halig) (147 , ,) (148 ADJ^N halig)) (149 , ,) (150 NR^N Drihten) (151 NP-NOM-PRN (152 NP-GEN (153 N^G weroda)) (154 NR^N God))) (ID colsigewZ,ÆLet_4_[SigeweardZ]:1154.564)) (47) Adj+Adj+Adj+N Halig, halig, halig, Drihten God ælmihtig, $se $ðe wæs, and $se $ðe nu is, and $se $ðe towerd is.
The (im)possibility of stacking adjectives in Early English
29
(NODE (2 NP-NOM-VOC (3 ADJP-NOM (4 ADJ^N Halig) (5 , ,) (6 ADJ^N halig) (7 , ,) (8 ADJ^N halig)) (9 , ,) (10 NR^N Drihten) (11 NP-NOM-PRN (12 NR^N God) (13 ADJ^N ælmihtig)) (ID coaelive,ÆLS_[Mark]:209.3331)) The exact figures corresponding to each group are presented in Table 1 below: Table 1. Adnominal adjectives occurring in a stack (data from 100 texts included in The York-Toronto-Helsinki Parsed Corpus of Old English Prose; Taylor, Warner, Pintzuk and Beths 2003). (a) sAdj+sAdj+N 100 (b) wAdj+wAdj+N 103 (ci) oAdj+oAdj+N 96 (cii) sAdj+wAdj+N/wAdj+sAdj+N 12 (ciii) Adj+Adj+Adj+N 2 Total 313
The most relevant constructions are those subsumed under (a) and (b). The former testify to the legitimacy of stacking strong adjectives, contra Spamer. The latter testify to the legitimacy of stacking weak adjectives, contra Fischer. Both groups are comparable in number (100 and 103 items respectively) and make a fairly convincing case against the view that OE adjectives are radically precluded from stacking. 4. Adjective stacking in OE: a Reconsideration Having looked at the linguistic data from OE, let us now proceed to reconsider the main question which this paper addresses: whether indeed OE adjectives should be viewed as non-recursive and whether the distinction between weak and strong declensions plays any role here. The preceding sections have demonstrated that neither the view of Spamer (1979), precluding strong adjectives from appearing side by side, nor that of Fischer (2000, 2001), precluding stacking both strong and weak adjectives, can be fully sustained. Given this, we
30
Agnieszka Pysz
can draw the conclusion that the impossibility of stacking OE adjectives does not really exist. In other words, OE grammar seems to have no principled mechanism which would explicitly prohibit the occurrence of adjectives in a stack. If such a mechanism did exist, one would never encounter sequences of two adjectives in a row. The empirical part of the paper has shown that such sequences do appear and they should not be denigrated by exception-to-the-rule ‘explanations’. In the light of the above, a reconsideration of adjective stacking in OE is needed. The alternative approach presented here makes no pretence of being authoritative or exhaustive. Rather, it is meant as an improvement, however slight, on the previous views on the issue in question. The approach proposed here is informally termed ‘liberal-radical’ and its main points are set out below. From the cross-linguistic perspective, the operation of adjective stacking is usually viewed as unconstrained in the sense that adjectives may in theory be iterated ad infinitum. In practice, however, some restrictions appear to hold. Thus, in most languages the attested sequences of adjectives hardly ever exceed a maximum of six or seven uncoordinated modifiers (Cinque 1995: 299). There are, nevertheless, languages in which the phenomenon of stacking appears to be more constrained than in others. Thus, for instance, in Romanian11 adnominal adjectives are precluded from occurring in each other’s immediate vicinity, as shown in (48). (48) *o carte mare groasă a book.F.SG big thick.F.SG ‘a big thick book’ In order to prevent the occurrence of adjectives one after another, similar strategies to those used in OE are used, namely: (a) splitting adjectives on both sides of the head noun, as in (49) (b) separating adjectives by means of a conjunction, as in (50) 11
The data from Romanian are taken from Herdan (2004).
The (im)possibility of stacking adjectives in Early English
31
(49) un biet student roman a poor student Romanian ‘a poor Romanian student’ (50) o carte mare úi groasă a book.F.SG big and thick.F.SG ‘a big and thick book’ Although the situation in Romanian is reminiscent of OE, it is not identical. The difference is that, whereas in Romanian the occurrence of two adjectives in a string is excluded, in OE such sequences do appear, though they are not particularly numerous. Thus, in comparison with Romanian, the grammar of OE seems to be more liberal. This is indeed what is meant by the liberal aspect of the approach towards stacking proposed here. In addition, the approach may be viewed as liberal because it weakens the assumptions present in both Spamer’s and Fischer’s accounts. It liberalises Spamer’s approach because it treats weak and strong adjectives as equals, as members of the same categorial class. It liberalises both Spamer and Fischer in that it tolerates the occurrence of an uninterrupted string of two adnominal adjectives, whether weak or strong, whether in a pre- or postnominal domain. One significant consequence of this is that it makes the distinction between weak and strong declensional patterns orthogonal to the phenomenon of stacking. In spite of its liberal aspect, the approach taken here can still be viewed as radical in that it does not allow adjectives to proliferate without constraints. Rather, it imposes a more or less rigid ban which precludes the occurrence of continuous adjectival strings consisting of more than two members. In the data retrieved from the corpus there are 2 isolated examples in which 3 adnominal adjectives are found in a row (see examples 46 and 47). The examples are very similar to each other as they both involve the repetition of the same adjective (i.e. halig) and, as Fischer (2000: 164, Table 2, row 8; cf. 2000: 172) suggests, they are very likely to be direct translations from the Latin sanctus sanctus sanctus. It is hoped that at least some of the observations made in this paper may contribute to an improved understanding of the still poorly understood issue of adjective stacking in the early stages of English. Naturally, there remain many loose ends which await resolution by
32
Agnieszka Pysz
future research in the area. One of the major questions that call for deeper investigation concerns the possible reasons why constraints on stacking are apparently more rigorous in OE than in PDE. The disparity between the theoretically limitless and practically limited number of adjectival modifiers has often been attributed to cognitive factors. Such a suggestion, however, does not explain why the possibility of stacking adjectives should exhibit a variation either cross-linguistically (e.g. Romanian vs other languages) or diachronically (e.g. OE vs PDE). Bibliography Primary sources Taylor, Ann, Anthony Warner, Susan Pintzuk and Frank Beths. 2003. The York – Toronto – Helsinki Parsed Corpus of Old English Prose (YCOE). York: Department of Language and Linguistic Science. Secondary sources Bermúdez-Otero, Ricardo et al. (eds). 2000. Generative theory and corpus studies: A dialogue from 10 ICEHL. Berlin and New York: Mouton de Gruyter. Cinque, Guglielmo. 1995. Italian syntax and Universal Grammar. Cambridge: CUP. Fischer, Olga. 2000. ‘The position of the adjective in Old English’ in Bermúdez-Otero et al. (2000): 153-181. Fischer, Olga. 2001. ‘The position of the adjective in (Old) English from an iconic perspective’ in Fischer and Nänny (eds): 249-276. Fischer, Olga and Max Nänny (eds). 2001. The motivated sign: Iconicity in language and literature 2. Amsterdam: John Benjamins Publishing Company. Haumann, Dagmar. 2003. ‘The postnominal “and adjective” construction in Old English’ in English Language and Linguistics 7(1): 57-83. Herdan, Simona. 2004. ‘Universal quantifiers in the Romanian DP’. Paper given at the Syntax Workshop (University of Connecticut, 6-7 March 2004). Hogg, Richard M. (ed.). 1992. The Cambridge history of the English language. Volume 1: The beginnings to 1066. Cambridge: CUP. Quirk, Randolph and Leslie Wrenn. 1957. An Old English grammar (2nd ed.). London: Methuen. Spamer, James B. 1979. ‘The development of the definite article in English: A case study of syntactic change’ in Glossa 13: 241-250. Traugott, Elizabeth Closs. 1992. ‘Syntax’ in Hogg (ed.): 168-289.
The (im)possibility of stacking adjectives in Early English
33
Appendix 1. List of 100 texts included in The York-Toronto-Helsinki Parsed Corpus of Old English Prose (Taylor, Warner, Pintzuk and Beths 2003). Text Code Adrian and Ritheus coadrian.o34 Ælfric’s Preface to Catholic Homilies I coprefcath1.o3 Ælfric’s Catholic Homilies I cocathom1.o3 Ælfric’s Preface to Catholic Homilies II coprefcath2.o3 Ælfric’s Catholic Homilies II cocathom2.o3 Ælfric’s Homilies Supplemental coaelhom.o3 Ælfric’s Letter to Sigefyrth colsigef.o3 Ælfric’s Letter to Sigeweard – Bodley 343 (B) colsigewB Ælfric’s Letter to Sigeweard - Laud Misc. 509 (Z) colsigewZ.o34 Ælfric’s Letter to Wulfgeat colwgeat Ælfric’s Letter to Wulfsige - Junius 121 (T) colwsigeT Ælfric’s Letter to Wulfsige - CCCC 190 (Xa) colwsigeXa.o34 Ælfric’s First Letter to Wulfstan colwstan1.o3 Ælfric’s Second Letter to Wulfstan colwstan2.o3 Ælfric’s Preface to Lives of Saints copreflives.o3 Ælfric’s Lives of Saints coaelive.o3 Alcuin’s De Virtutibus et Vitiis coalcuin Alexander’s Letter to Aristotle coalex.o23 Anglo-Saxon Chronicle - Ms. A (Parker Chronicle) cochronA.o23 Anglo-Saxon Chronicle - Ms. C cochronC Anglo-Saxon Chronicle - Ms. D cochronD Anglo-Saxon Chronicle - Ms. E (Peterborough Chronicle) cochronE.o34 Apollonius of Tyre coapollo.o3 Augustine coaugust St Augustine’s Soliloquies - Preface coprefsolilo St Augustine’s Soliloquies - Soliloquies cosolilo Bald’s Leechbook colaece.o2 Bede’s History of the English Church cobede.o2 Benedictine Rule cobenrul.o3 Blickling Homilies coblick.o23 Boethius, Consolation of Philosophy coboeth.o2 Byrhtferth’s Manual cobyrhtf.o3 Canons of Edgar - CCCC 201 (D) cocanedgD Canons of Edgar - Junius 121 (X) cocanedgX Charters and Wills - Helsinki period O1 codocu1.o1 Charters and Wills - Helsinki period O1/2 codocu2.o12 Charters and Wills - Helsinki period O2 codocu2.o2 Charters and Wills - Helsinki period O2/3 codocu3.o23 Charters and Wills - Helsinki period O3 codocu3.o3 Charters and Wills - Helsinki period O4 codocu4.o24 Chrodegang of Metz cochdrul Preface to Cura Pastoralis coprefcura.o2 Cura Pastoralis (Hatton 20) cocura.o2
34
Agnieszka Pysz
Cura Pastoralis, (Cotton Tiberius B.XI) Distichs of Cato Gospel of Nichodemus - Ms. A Gospel of Nichodemus - Ms. C Gospel of Nichodemus - Ms. D Gospel of Nichodemus - Ms. E Gregory’s Dialogues - Ms. C Gregory’s Dialogues - Ms. H Heptateuch - Gen, Exod, Lev, Num, Deut, Josh, Judg Heptateuch - Ælfric’s Preface to Genesis Heptateuch - Ælfric’s Epilogue to Genesis Heptateuch - Genesis (CCCC 201) Heptateuch - Exodus (Pierpoint Morgan Library G.63 [P]) Herbarium Honorius of Autun, Elucidarium - Version 1 Honorius of Autun, Elucidarium - Version 2 Institutes of Polity - CCCC 201 (D) Institutes of Polity - Junius 121 (X) Lacnunga Helsinki period O2 - Alfred’s Introduction to Laws Helsinki period O2 - Laws of Alfred Helsinki period O2 - Laws of Ine Helsinki period O3 - Laws of Cnut, prt.1 Helsinki period O3 - Laws of Cnut, prt.2 Helsinki period O3 - Laws of Æthelred, prt.1 Helsinki period O3 - Laws of Æthelred, prt.2 Helsinki period O3 - Northumbra Preosta Lagu Helsinki period O4 - Gerefa Helsinki period O4 - Laws of William Martyrology - Part 1 Martyrology - Part 2 Martyrology - Part 3 Marvels of the East Orosius Invention of the Cross (History of the Holy Rood-Tree) Saint Chad Saint Christopher Saint Euphrosyne Saint Eustace and his Companions James the Greater Saint Margaret - CCCC 303 Saint Margaret - Cotton Tiberius A.III Saint Mary of Egypt Saint Neot Saint Vincent (Bodley 343) Seven Sleepers
The (im)possibility of stacking adjectives in Early English Quadrupedibus Solomon and Saturn - Cotton Vitelius A.XV (I) Solomon and Saturn - CCCC 422 (II) De Temporibus Anni Vercelli Homilies Homily I, Ms. E Homily IX, Ms. L Vindicta Salvatoris Vision of Leofric West-Saxon Gospels Wulfstan’s Homilies
Lists in letters: NP-lists and general extenders in Early English correspondence Ruth Carroll University of Turku
Abstract This paper addresses a hitherto unexplored topic in diachronic English linguistics: the use and form of extenders, phrases such as ‘and other’ and ‘or the like’ (sometimes also called set-marking tags, list completers, and vague category identifiers). The Corpus of Early English Correspondence Sampler is used as the source for a survey of NP lists and the extenders found within them. The result is an inventory of 8-9 EME general extenders and an illustration of the wide variety of possible specific extenders. Jefferson’s claim of a norm of three-partedness for Present Day English lists (1990) is shown to hold true also for Early Modern English, with four-part lists strikingly less common than three-part lists, and longer lists (of 5+ parts) rarer still.
1. Introduction1 This paper presents initial findings from a research project into lists in Middle and early Modern English. The emphasis of the study is not on lists serving as discourse topics, but on elided lists used as syntactic sentential elements such as objects of verbs or prepositions. The results show that, as in Modern English, in earlier English the threepart list was typical. The paper goes on to lay groundwork for a study of general extenders in earlier English by beginning an inventory of the forms that could be used and commenting briefly on those forms. A number of avenues for future research are also introduced. The main 1
I am grateful to students in my Historical Discourse Linguistics seminar at the University of Turku and to participants at the Selim XVII conference for comments on earlier versions of this paper.
38
Ruth Carroll
source of data for this study was the Corpus of Early English Correspondence Sampler (described below), searched with WordSmith concordancing software. This was supplemented by some manual searching of the Cely letters (Hanham 1975). The project began as a desire to study general extenders in the history of English, and especially in early Modern English. General extenders (which have also been called set-marking tags, generalized list completers, extension particles and vague category identifiers) are phrases consisting of a conjunction plus a non-specific noun phrase such as or other thyng or and such lyke.2 Overstreet (1999), in a study of present-day conversational English, considers them to be better analyzed as having an interpersonal function in discourse rather than referential meaning or other discoursal/textual functions. However, before a thorough study of early Modern English general extenders could be made, an inventory of forms was needed. Such an inventory could not be quickly assembled (merely following Overstreet’s inventory, for example) because the relationship between form and function has changed considerably over time. A study of this process of change would be useful, but is beyond the scope of this paper. It became clear that what was also needed was an account of list characteristics in early Modern English. This paper begins that task, focusing on lists composed of NPs.3 The study is thus situated within the field which Brinton (2001: 139) calls ‘historical discourse analysis proper’, the synchronic study of discourse forms (in this case lists and, to a lesser extent, extenders) and their functions in late Middle and early Modern English. It is a micropragmatic study, focusing on the function of small units within the text. 2. Previous Research To my knowledge, there is no previous published discourse analytical research into the historical uses (or forms) of general extenders. 2
A complete history of the use of these terms is beyond the scope of this paper, but these four terms can be found in Winter and Norrby 2000, Jefferson 1990, Dubois 1993 and Jucker et al. 2003. General extenders is used by Overstreet (1999). 3 Schiffrin notes that it may be useful, especially in a macropragmatic analysis, to consider NPs to be “elliptical mentions” of propositions (1994: 299, 310), as mentioned below.
List in letters
39
Overstreet (1999), whose data consist primarily of present-day conversation among familiars, cites one paragraph from Jane Austen (published 1818) and otherwise her oldest material is from J.D. Salinger (published 1945). None of the references in her bibliography appear to be historical studies. Neither the Oxford English Dictionary nor the Middle English Dictionary has a conventional definition or part of speech reserved for general extenders, with the result that even in their electronic forms they are difficult or impossible to use as sources for an inventory of historical general extenders.4 Visser’s Historical Syntax contains little to no treatment of clauses or units other than the verb phrase. Prior work on lists (in present-day English conversation) has been carried out by Schiffrin (1994) and Jefferson (1990). Schiffrin’s work is macropragmatic –she is interested in lists as discourse units. My own tokens are almost all taken from within sentences, thus forming a micropragmatic analysis. Schiffrin’s data include many instances of lists composed of clauses (you got Jersey; Y’got Atlantic City; you got Saratoga (Schiffrin 1994: 294)), although she does note an alternative method of list construction: ellipsis. In the latter type of list, nothing is explicitly predicated of the different items in the list; each item is presented as a bare phrase (Irish, German, and Jewish (1991: 298)). The current study focuses on this kind of list, because tokens can be found by using a concordancing program.5 More specifically, I have concentrated on lists composed of NPs. In these, as Schiffrin (1994: 311) notes, temporal order is less important than it can be in lists composed of clauses or similarly large items. Most of my tokens serve a different function from Schiffrin’s lists, in that they can be seen as elements of grammar. However, a few, like Schiffrin’s, serve as units of discourse, discourse topics in which set membership is enumerated and/or discussed: (1) Neverthelesse, wee goe on with a remonstrance or informacion to his Matie contayning the generall grievance of the realme, which wee have reduced to thease heads, namely, fear of innovation of religion, 4
I was interested to find a token of a general extender used in the definition wording of an OED entry: a statement, enumeration, or the like (s.v. follow, v. 15a). A subsequent search for the phrase the like in OED definitions produced 2388 hits. 5 Another factor making a study of elided-item lists preferable for my purposes is that unlike clause-item lists (Schiffrin 1994: 321-322), their endings are usually clearly identifiable.
40
Ruth Carroll and the ill successes of our late forrain enterprizes, the ill state and decay of our forts and castles, the generall want of powder and all other sorts of ammunition requisite for the defence of the realme, the decay of trade, the great losse and decay of the shipping of the realme, the ill guarding of the narrow seas; and ended in these very termes, that the excessive power of the Duke of Buckm and the abuse of that power is the chief cause of these evils and daungers to the King and kingdome. (CORNWALL: 1627 Thomas Meautys 177-178)6
Here the topic is “concerns” or “grievances”. The list is introduced with namely and consists of “fear of religious innovation”, “recent failures abroad”, “poor condition of forts”, “lack of ammunition”, “loss of trade”, “decrease in shipping” and “lack of security in the narrow seas”. While many of Schiffrin’s data are very different from my own, the majority of Jefferson’s data are, like mine, from within sentences. However, Jefferson’s focus is very much on the effect lists have on conversation (both the restrictions, such as the normative nature of the three-part list, discussed below, and the resource such structures provide to conversational participants, for example in handling potential disagreement). Much of her analysis is therefore not applicable to epistolary data, insofar as the interaction between correspondents is very different from that of face-to-face speakers. Letter-writers generally have time to pause and think while composing their sentences, and to rephrase, in a way that conversationalists do not. Moreover, the recipient of a letter cannot participate in its composition in the way that a listener can participate in a conversation, for example by finishing the speaker’s sentence. 3. Data and Methodology This study is based primarily upon the CEECS, the sampler portion of the Corpus of Early English Correspondence. This is a computerreadable corpus of letters dating from 1418-1680. The bulk of the letters in the corpus are from the later years in that range. The sampler contains data from 194 informants (letter-writers). The letters are 6
References to the CEECS list first the letter collection, then the date and name of writer of the letter, and end with a page reference to the edition from which the letter was taken. The bibliographic details for these editions are available on-line: http://khnt.hit.uib.no/icame/manuals/ceecs/TEXT.HTM
List in letters
41
included in their entirety –the word sampler in the title refers to the fact that this corpus represents only a fraction of a larger CEEC corpus. The CEECS consists of 450,000 words (for discussion of the relevance of data from the sampler in comparison to the full corpus, see Nurmi 2002). The choice of the CEECS as data was based in part on the relative immediacy of letters as a genre. Both Overstreet (1999) and Schiffrin (1994) rely primarily upon spoken, relatively informal, conversation for their data. Indeed, Overstreet finds the proportion of general extenders to be highest in such data (1993: 6). Since analagous data are not available for early stages of English, correspondence is one of the best alternatives. I have not included in my data editorial headings to the letters or external addresses on the letters. For this pilot study I have focused my attention on lists of three or more noun phrases. Using WordSmith software, I searched for phrases collocating with and and or.7 After creating a concordance with a collocational range of approximately twenty-six words (thirteen or so on either side of the coordinator), I then manually discarded all hits other than noun phrases. It is possible that a few NP-coordinations were mis-identified and excluded. For example, I discarded many results in which a parenthetical phrase followed the coordinator; a very few of these may in fact have been sections of lists.8 On the other hand it is also possible that some phrases were included which should not have been, since I included some ambiguous examples such as with hym my lord Darcy and many others (which could be a conjunction of two elements if hym and my lord Darcy are co7 This search methodology will obviously have excluded any lists not containing and or or. However, there are instances in which neither conjunction is present in the letter but one has been added as an editorial emendation. With so few data it is difficult to judge, but the discourse linguist must allow for the possibility that this was not merely a scribal error, but a deliberate authorial omission. for some interloper hath clos’d it with Dr. Sibthorpe, Manwaring, Jackson. Yet this last refused it. (COSIN 1628 Richard Mountague I, 152, emended to and Jackson) For the current paper such tokens were nonetheless counted according to their emended forms. 8 To illustrate, the following list was rejected on the grounds that it coordinates PPs, but it is possible that some NP lists similar to it may have been overlooked: to our queen, our countrye, and, that is of most importance, to the whole cause of Godes chirch (LEYCESTE 1586 William Cecil 420).
42
Ruth Carroll
referential, but was here analyzed as a three-part list). On balance, however, I believe my results to be representative of the data. Conjunctions of two elements alone were not considered, following the common definition of a list as having a minimum of three elements.9 I also excluded lists of titles such as (2) and (3), given the problematic nature of onomastics and my prior decision to exclude material from the address portions of the letters. (2) The duc of Gloucestre, gret Chamberleyn, Constable, and Admirall of England (STONER 1479? Duke of Gloucester II, 81) (3) send the Bukk to Warde, Shreve and Aldyrman of London (STONER 1480 Walter Elmes II, 106)
The list tokens were classified according to the number of items, whether the list consisted of all specific items or contained an extender, and whether such extenders were general or specific. 10
4. General and Specific Extenders This pilot study quickly revealed that the distinction between general and specific extenders is not as clear as might initially be supposed. Like so many linguistic distinctions, it is a fuzzy one. Overstreet’s examples of general extenders include and that kind of nonsense, and business of that sort, or someone like that and or somewhere (1999:4). Phrases she classifies as specific extenders include and all of that 9
For this pilot study I excluded even those two-part lists which were conjoined to another NP if the conjunction differed from the first one used. For example, the counties of Dublin and Kildare or one of them was not included, being analyzed as (NP + NP) or (NP). Similarly, any lists “interrupted” by with or then were excluded: for our organs and our cornets, together with the candlesticks and tapers upon the Communion… (348), and so is your brothers and sisters, with Ned Smith. (538), with the Chif Justise Fortescu, and þen another rule, &c. (642). Some of these decisions were motivated more by the methodology (seaching for lists marked by and or or) than by any theoretical concerns. For bibliography on three-part lists, see Overstreet (1999: 27-28). 10 The list, two and three yeare old heifferss, and runts and lambes could be counted as a three-part list (heifers, runts and lambs) or as a four-part list (two-year-old heifers, three-year-old heifers, and also any runts and all lambs). My tendency was to grammatical counting, based on the number of actual nouns, since this often seemed more objective. However I find it difficult to be certain about my consistency in avoiding semantic interpretations.
List in letters
43
stupid bureaucratic stuff (1999: 12), and any extenders which are further specified by a following relative clause, such as or anything that belongs to your neighbor (1999: 52). Thus, general extenders can consist of (generic or vague) nouns such as business, nonsense or things, as well as such indefinite pronouns and adverbs such as anywhere or sometime. But insofar as any rule can be inferred from Overstreet’s data, it is that any extender qualified by a specific adjective, such as bureaucratic, or by a relative clause, is specific. For the purposes of the first part of this paper a similarly strict definition has been followed. Of the specific extenders found, or oþer about you (see example (20), below) is one of the most general, but it does specify [which are] about you. Even such vague extenders as yours (with no specific noun, but understood to mean kinfolk or household members) have been classified here as specific. The distinction between general and specific should more accurately be seen as a cline rather than as a sharp division. However, that the distinction is made, even by language users, might explain example (4), which contains both a slightly specific extender (man), and a wholly general one (etc.). (4) Germyn, Kyrton, or some other man, &c. (SHILLING 1447 John Shillingford 17)
It is possible in this case that the first extender was felt not to be general enough, and so the writer added the &c. 5. Findings A total of 130 tokens of lists were identified and analyzed. As noted above, the shortest of these were three-item lists; the longest contained eight items. Conjunctive lists (those conjoined with and) were significantly more frequent than disjunctive lists (82 tokens as opposed to 48).
44
Ruth Carroll
Table 1. Elided-NP lists in the CEECS 3-part lists
3 specific items 2 + specific extender 2 + general extender total
AND 50 4 2 56
OR 24 7 7 38
TOTAL 74 11 9 94
4-part lists
4 specific items 3 + specific extender 3 + general extender total
10 1 2 13
2 4 3 9
12 5 5 22
5-part lists
5 specific items 4 + specific extender 4 + general extender total
2 0 1 3
0 1 0 1
2 1 1 4
6-part lists
6 specific items 5 + specific extender 5 + general extender total
1 2 1 4
0 0 0 0
1 2 1 4
7-part lists
7 specific items 6 + specific extender 6 + general extender total
3 1 0 4
0 0 0 0
3 1 0 4
8-part lists
8 specific items 7 + specific extender 7 + general extender total
1 0 1 2
0 0 0 0
1 0 1 2
82
48
130
TOTAL
List in letters
45
5.1. Conjunctive Lists By far the most frequent kind of list was the three-part list composed entirely of specified items (50 tokens). Of those three-part lists in which the third element was an extender, four were specific and two were general (see examples below): No extender: (5) the ffather, the word and Spiritt (JONES 1653 Peter Sterry 250) (6) in Millan, Naples, and Sicilly (LEYCESTE 1585 William Cecil 42)
Specific extender: (7) my horsse and harnes and all my oder goode (STONOR 1474 Thomas Stonor3 I, 148)) (8) yow, my good modyr, and alle yowrs (STONOR 1471? William Stonor I, 121)
General extender: (9) John a Copleston and William Hendeston and other (SHILLING 1447 H. Webber 34) (10) her grete sadnesse and þe vertuous disposicion þat she is of &c. (STONOR 1466? Humphrey Forster I, 93)
Four-part lists were more common than longer lists (13 four-part, 3 five-part, 4 six-part, 4 seven-part, and 2 eight-part), but noticeably less frequent than three-part lists (perhaps providing evidence to support Jefferson’s claim of a norm of three-partedness for lists (1990)). As with three-part lists, the most common four-part lists were those with no extender (10 tokens): No extender: (11) four steeples, three churches, two parishes, and not long since but one preist (WHARTON 1642 Nehemiah Wharton 12)
Specific extender: (12) to my Ladie Carlisle and [?] Bedford, and all my frendes 1642 Thomas Savile 4) 11
11
(SAMPLE
The question mark in brackets indicates editorial uncertainty. I have presumed that what is missing is a title. This particular list is further extended with a prepositional
46
Ruth Carroll
General extender: (13) from Hamborg, and Dansk, Lubeck, &c.(LEYCESTE 1586 William Cecil 200)12
Extenders were also found for all the longer lists (again, a result which mirrors Overstreet’s present-day data), but here the numbers of tokens were so small as to make it difficult to draw any conclusions. Note also that the methodology used to locate examples was skewed toward better recognition of shorter lists than longer. This may mean that some long lists are missing, but my expectation is that the prototypicality of three-parted lists will be borne out. 5.2. Disjunctive Lists As noted above, disjunctive lists were less frequent than conjunctive lists in the CEECS, but the distribution according to length was similar, in that three-part lists were significantly more common. 38 three-part lists were found, against only 9 four-part lists and one fivepart list. No longer disjunctive lists were found. Often in the disjunctive lists, the extender extends not by denotating additional items not yet listed, but by allowing for different combinations of items already listed or parts thereof: (14) I shall give my Lord of Canterbury, or my Lord of London, or both, the contents of your Lordship’s letter. (COSIN 1665 Richard Sterne II, 134) (15) the said yeerely rent or 200 l., or any part thereof (COSIN 1666 John Cosin II, 149)
These were analyzed as specific extenders. The three-part disjunctive lists were primarily (24 tokens out of 38) composed without extenders. Those with extenders were almost phrase: and particularlie to my poore Cozen. In another study, this particular example might be counted as a four-part list, which would then be of interest for having the extender in a position other than final (see also example (24)). 12 For the purposes of this article, I have analyzed this as a four-part list rather than a single PP followed by a three-part NP list. It is possible that some inconsistencies may have occurred in my analyses in this preliminary study, which will be clarified before the next stage of research.
List in letters
47
equally divided between those using specific extenders and those containing general extenders. Some examples are given here: No extender: (16) one, twoe, or three hundreth powndes (HUTTON 1623 Timothy Hutton 245) (17) French, or Duch, or Latyn (LEYCESTE 1585 Robert Dudley 32)
Specific extender: (18) whether he be called Mayor, Alderman, or Bailiff, or by any other title whatsoever (COSIN 1660s? Miles Stapylton II, 385) (19) tax, or talliage, or any manner of charges (LEYCESTE 1586 Robert Dudley 126) (20) eny freind, man or oþer about you (STONOR 1478 Thomas Betson II, 46)
General extender: (21) plant, beaste or the like (JONES 1653 Peter Sterry 250) (22) on Saturday or Sonday &c. (STONOR 1472? Thomas Mull I, 127)
Among the four-part lists were two with no extenders, four with specific extenders, and four with general extenders. One, counted here among the lists with general extenders, actually contained one of each (example (4), repeated as (23)): (23) Germyn, Kyrton, or some other man, &c. (SHILLING 1447 John Shillingford 17)
In the following four-part list, a specific extender is used, but it is not the final item (underlining added): (24) you, my cousin your son, or any other of your name, or servant (PLUMPTON 1503 Robenet Plumpton 173)
This is also found elsewhere in the data, but is not common.13 13
However, the possibility must be allowed for that some such examples were missed in the data collection, being analyzed instead as shorter lists ending with the item
48
Ruth Carroll
Further study into lists in early English is needed to take account of the extent to which lists could be split in ways which are rare or even impossible today. This holds true not only for Old English, but is also still the case in the early Modern period: (25) my brodyr Wyllӻm & my Syst~ ys vyfe & halle the howsold & Margery my Syst~ and her hosbond ver in gowde halle & all her howsold. (MARCHALL 1450s? John Collas F268)
The subject of the verb above appears to be a six-item list, with five of the items occurring before the verb and one following it. Another feature of early Modern English lists which sets them apart from standard written modern English is the repetition of and: (26) the Earle of Northampton, the Lord of Carnarvan, and the Lord Compton and Captn Legge, and other (WHARTON 1642 Nehemiah Wharton 9)
Given the constraints upon us here, these features may be noted, but more in-depth discussion must be left for future study. 6. General Extenders The present study has produced an initial inventory of the general extenders used in late Middle and early Modern English. Those found in this survey of the CEECS are shown in table 2. Table 2. General extenders in lists in the CEECS conjunctive &c and such and other
disjunctive &c or any or more or other or such like or the like14
immediately following the conjunction. This is more likely to be the case with longer/heavier (syntactically) items following the conjunctions. 14 The absence of conjunctive phrases with like might be seen to be confirmed by the OED entry (s.v. otherlike), whose first three citations are all disjunctive.
List in letters
49
In addition, many examples of specific extenders have been identified (and other commaunders, and all other ware perteynyng to the saide crafte, and all heroicall vertues and others cited above). The general extender and such was found only once in the data, and that token was ambiguous as it was found capitalized in a list of names. However, I have analyzed it as an extender rather than as a surname. And other was found, but and others in the plural was not. That general extenders are not required to agree with the other items they conjoin with is a feature noted by Overstreet (1999: 10). In disjunctive phrases, too, other was found only in the singular. However, the word others was found as an element in some specific extenders, and future corpus searches for general extenders should include it. &c was the most frequent of the conjunctive general extenders, and was found elsewhere in the corpus as well, including as a disjunctive extender (for example, in (22)). One of its additional functions is to abbreviate formal set phrases, such as captatio benevolentiae, and extracts or quotations: (27) I recommaunde me unto you &c (STONOR 1462? Hugh Unton I, 61) (28) your later letter, wherein brefely and most truely is recited the effect and meaning of the former letter, viz., that I should use the assistance and advise of Sir Thomas Fairfax, Sir Edward Stanhope, and Mr. Atturney of the Wardes, or any two of them, and all convenient means, to informe my self what place the Vice-President had by the space of 25 or 30 yeres, &c. (HUTTON 1602 Matthew Hutton 164)
This usage is also found earlier in the Middle English period: (29) he seyd thys vers of the psauter, as prophecye, throgh þe holy goste: ‘Hec Requies mea in seculum seculi’ et cetera (English conquest of Ireland as contained in the Corpus of Middle English Prose and Verse)
Whatsoever was not found in the data as an extender, but as an emphatic: (30) if they shall dare to use any other rite, ceremonie, ornament, or order whatsoever (COSIN 1629 John Cosin I, 158)
50
Ruth Carroll
It could be worth searching for this term (and also for howsoever, found as a list-item in these data) in future corpus searches for general extenders, although it would generate many unwanted hits since it is not uncommon elsewhere. Overstreet defines general extenders structurally as a conjunction plus a non-specific noun phrase. This is somewhat problematic historically, as not all of the tokens have this structure etymologically. One common example, though not among the tokens analyzed here, is the adverbial and so forth. The earliest OED citation for and so forth which can be analyzed as having the function of a general extender is from 1574-5 (example 31). It is also found as an abbreviator, analogous to the function of et cetera in (29), as shown in (32): (31) I toy out my time, partly with copying of books..partly in genealogies, and so forth (Abp. Parker Corresp. s.v. OED forth adv., 9b, my emphasis) (32) and so he prophecied seienge: Blessid be oure lord god of Israel for he hath visited thoru3 grace and made redempcioun of his puple. And so forth as it is conteyned in the gospel. (Mirrour of the blessed lyf of Jesu Christ as contained in the Corpus of Middle English Prose and Verse, my emphasis)
The usage and so forth is most likely a case of grammaticalization. It certainly involves semantic shift, and probably grammatical reanalysis. Middle English examples of its adverbial usage are easy to find: (33) he toke his weie for na3areth towarde Jerusalem and so forth til he come to the water Jordane (Mirrour of the blessed lyf of Jesu Christ as found in the Middle English Compendium, my emphasis)
7. Future Research In addition to several topics already proposed above, most especially the study of various individual extenders, future study could encompass lists composed of:
List in letters
51
conjoined clauses: (34) we all desire that ether the Parliament would depose him, or God convert him, or the Devill fetch him away quick (WHARTON 1642 Nehemiah Wharton 6) (35) whether he intends to allow wood or no, or how many half paces, or of what bredth and height the steps shalbe (COSIN 1662 John Cosin II, 90)
verb phrases: (36) to se him, and ride a myle or ij with him, and wellcome him to the country (PLUMPTON 1496 Edward Plumpton 117) (37) to lift us up to Pride, allure us to coveteousnes, or rend us into Factions (jones 1651 John Jones 187)
adjective phrases: (38) Your letter, being a pitthy, sollid, brief, and reall relation,(WHARTON 1642 Nehemiah Wharton16) (39) The Lord give us humble, thankfull, uniting, beleeving, and selfedenying spirits (JONES 1651 John Jones 188)
and other parts of speech such as prepositional phrases: (40) of all of these provinces, of their forces by sea and lande, of their townes and of their treasure (LEYCESTE 1586 Robert Dudley 102) (41) wtynne the jurisdiccion and under the coreccion and punysshment of the cite, &c. (SHILLING 1447 John Shillingford 10)
This paper has not specifically addressed language of vagueness, but this would be another worthwhile topic for the future (see, for example, Jucker et al. 2003). Note, for example, the use of two hedges, or indicators of inexactitude, in the following: (42) extendid half a myle in leyngthe by estimacion or thereaboute (ORIGINA3 1533 Thomas Cranmer 37-38)
A special category of lists could be addresses, lists of multiple titles which all refer to the same person:
52
Ruth Carroll (43) Right reverend worshipful Sir, intirly beloved brother, and singularly my good master, (PLUMPTON 1464 Brian Rocliffe 12)15
Future study would do well to make use of parsed corpora, to enable searches such as “CONJ + NP”. This would also target searches more effectively to exclude hits for the homonym of the conjunction or, that is, the preposition or: (44) that he wol have Wolston or this somer be don (STONOR 1480 Richard Germyn II, 109)
Alternatively, a historical discourse linguistic approach, involving careful reading of entire texts,16 would likely prove very fruitful, and would allow a more thorough better comparison with Modern English analyses such as that of Schiffrin (1994), which involved the close analysis of entire conversations. Texts likely to be worth studying in this way include letters, but also recipes and wills: (45) Also a gowne..in ward, &c. (1418 E.E. Wills, OED s.v. et cetera)
8. Conclusion This paper has highlighted the limitations of corpus research in searching for phrases not distinguished by specific lexemes, especially the limitations of an unparsed corpus, and the limitations for research involving semantic analysis. Nonetheless, it has shown the enormous prevalence of the three-part, non-extended list over other lengths and specificities for the early Modern period. The paper has also made initial steps toward a study of general extenders in earlier English, providing a short inventory of some of the phrases that could be used, and briefly noting some other forms and/or functions for these phrases. 15 This category overlaps somewhat with lists of descriptors such as the following: my lord the Prince your sonne, which was not included in the present study. The decision was based partly on punctuation, although this is likely in some cases to be editorial. The number of such examples in the sample is too few to affect the results greatly. 16 Here I follow the terminology of Hiltunen and Skaffari (2003).
List in letters
53
Bibliography Brinton, Laurel J. 2001. ‘Historical Discourse Analysis’ in Schiffrin, Deborah, Deborah Tannen and Heidi E. Hamilton (eds) The Handbook of Discourse Analysis. Malden (Massachusetts): Blackwell: 138-160. Dubois, Sylvie. 1993. ‘Extension particles, etc.’ in Language Variation and Change 4: 179-203. Hiltunen, Risto and Janne Skaffari (eds). 2003. Discourse Perspectives on English: Medieval to Modern. Amsterdam: John Benjamins. Jefferson, Gail. 1990. ‘List-Construction as a Task and Resource’ in Psathas, George (ed.) Interaction Competence. Washington, DC: The International Institute for Ethnomethodology and Conversation Analysis: 63-92. Jucker, Andreas, Sara W. Smith and Tanja Lüdge. 2003. ‘Interactive aspects of vagueness in conversation’ in Journal of Pragmatics 35: 1737-1769. Nevalainen, Terttu, Jukka Keränen, Minna Nevala, Arja Nurmi, Minna PalanderCollin and Helena Raumolin-Brunberg (compilers). Corpus of Early English Correspondence Sampler. Helsinki: University of Helsinki. Nurmi, Arja. 2002. ‘Does size matter? The Corpus of Early English Correspondence and its sampler’ in Raumolin-Brunberg, Helena, Minna Nevala, Arja Nurmi and Matti Rissanen (eds) Variation Past and Present: VARIENG Studies on English for Terttu Nevalainen. Helsinki: Société Néophilologique: 173-184. Schiffrin, Deborah. 1994. Approaches to Discourse. Malden (Massachusetts): Blackwell. Overstreet, Maryann. 1999. Whales, Candlelight, and Stuff Like That: General Extenders in English Discourse. New York: OUP. Visser, Frederick Theodore. 1963-1973. An historical syntax of the English language (3 parts, 4 vols). Leiden: E. J. Brill Winter, Joanne and Catrin Norrby. 2000. ‘Set Marking Tags – “and stuff”’ in Henderson, John (ed.) Proceedings of the 1999 Conference of the Australia Linguistic Society. On-line at: http://www.linguistics.uwa.edu.au/research/ als99/proceedings (consulted 01.12.2006)
Middle English medical books as examples of discourse colonies: G.U.L Hunter 307 Francisco Alonso-Almeida University of Las Palmas de Gran Canaria
Abstract This paper investigates the case of manuscript G.U.L. Hunter 307 as an example of a ‘discourse colony’. This concept refers to medieval books containing items which are apparently unconnected. By considering volumes as colonies, we can offer fresh theories as to how books were compiled and what their intended audience might have been. Texts analysed from the ‘colony’ perspective also help us to detect textual affiliations within a compendium, allowing us to identify more effectively scribal activities of copying, excerption and rephrasing.
1. Introduction The history of Middle English medical books has been the focus of a number of studies from different perspectives, including cultural studies. However, until Carroll (2003) drew our attention to the need for a new approach to medieval texts and books, manuscript items were edited and analysed with no regard to their textual context. Halliday and Hasan (1989) have pointed out that those textual cues are just as important to our understanding of a discourse as any other type of context. For this reason, I intend to study Glasgow University Library, Hunter 307 (U.7.1) (henceforward H) following Carroll’s methodology (2003) and Hoey’s definition of discourse colony. An earlier paper by me (Alonso-Almeida 2005) undertakes a similar survey of some twenty fifteenth-century manuscripts, arguing that a
56
Francisco Alonso-Almeida
medieval volume should be studied as a single unit rather than as a collection of discrete items. The term discourse colony was firstly coined by Hoey (1986, 2001) to refer to “cinderella discourses that get neglected in most discourse theories”, as “these discourses form a relatively homogeneous class and can be described in terms that allow integration with conventional descriptions of mainstream discourses. They are homogeneous in respect of their discourse characteristics but highly heterogeneous in respect of their appearance and use” (1986: 12). I should point out that the words text and discourse in the present paper will be used indistinctly, as traditionally these two terms have met with fuzzy definitions; so much so that Bonilla (1997: 9-10) adverts that even van Dijk shifts ambiguously from the term text to discourse in his studies from 1981 onwards. Beaugrande, meanwhile, opts for the expression text and discourse systematically (Bonilla 1997: 10). The structure of the paper is as follows: Section 2 describes the classifying and physical features of H, together with its contents and eVK numbers. In the next section I define the concepts of discourse colony as given by Hoey (1986), followed by an analysis of H according to the framework established by that author, and finally a summary of the conclusions drawn. 2. The manuscript H, housed in Glasgow University Library, was written in the early fifteenth century using a Texture script, specifically a combination of Semiquadrata and Rotunda. It contains 174 folios and the text is chiefly in English, although some Latin passages are also present. The table of contents (transcribed in section 4.2) is of a later date, probably the sixteenth century, but it does not offer information for the entire volume, only for Items 2 and 3 (see (1) below). We can only guess, therefore, at the intention of the scribe in annotating certain remedies and disease descriptions in the margins of the text. The manuscript is in generally good condition, although a few parts are almost unintelligible owing to the degradation of the ink. In this article, all revisions are my own, unless otherwise stated. In all the examples of Middle English texts, I substitute thorn for present-day
and yogh is given as <>ޚ.
Middle English medical books as examples of discourse colonies
57
The contents of H together with the eVK numbers, i.e. the cataloguing numbers used in Voigt and Kurtz’s electronic catalogue Scientific and Medical Writings in Old and Middle English, are summarised in table (1), below; TK numbers, i.e. the cataloguing numbers in A Catalogue of Incipits of Mediaeval Scientific Writings in Latin, are also provided where applicable. Items 2-3 in the table belong to the Compendium medicinae by Gilbertus Anglicus (Gilbert of England), as suggested by Voigts and Kurtz (2000). Green (1992) has identified item 3 as Version 1 of Anglicus’s Sicknesses of women within the Compendium. The author offers a classification of gynaecological texts in Middle English, mainly after Trotula and Gilbertus Anglicus.1 In section 4, I provide a description of each version of the Sickness of women, as explained by Green (1992), because we will need this information to understand H properly as an example of a discourse colony. (1) Incipit
It is to understand that a man is made of four elements and every man hath four
Elements; humours; ages of man; urine and uroscopy
2
A man that will help men in their sickness him behooveth to know the encheasons
Disease; etiology; medical recipes; herbs and herbal medicine; epilepsy
3
As we shall understand that women have less heat in their body than men and more moistness
Gynecology and obstetrics
1
See Green (1992) for a list of the texts surveyed for her study.
eVK number 3206.00
640.00
1162.00
58
4
5
Francisco Alonso-Almeida
A physician behooveth to know three manner inspections in bloodletting that is to say Aloe is hot and dry in the second degree and it
Hematoscopy; blood; bloodletting Platearius, Circa instans (excerpt) Herbs and herbal medicine
BL, Royal 18.A.6, ff. 54-55 BL, Sloane 965, ff. 161v-63 BL, Sloane 3486, f. 147v
659.00
860.00 TK 211C
The Trotula-based manuscripts are divided into five main groups: (1) Trotula translation A, (2) Trotula translation B, the Liber Trotuli, (3) Trotula translation C, the “Boke Mad [by] a Woman Named Rota”, (4) Trotula translation D, Secreta mulierum, and (5) Trotula translation E, chapters from the Trotula major. In the first group, texts either begin with the incipit Oure Lord God when he stored the world or feature the phrase elsewhere within the text. The main sources for this group are the Trotula major and the Trotula minor, and women are addressed directly. Group 2 also deals with cosmetics, and there is no direct address to women. Treatments are drawn from Hippocrates and Galen. The third group comprises free translations of chapters from the Trotula major and Trotula minor, in addition to other extra material. The fourth group contains literal translations from the Trotula major and the Trotula minor, with no direct address to women. Finally, group 5 presents literal translations from the Trotula major. For gynaecological texts based on the Compendium medicinae by Gilbertus Anglicus, Green divides the surviving copies into two main groups: (1) Version 1, and (2) Version 2. In the first version, some sections of Anglicus’s Compendium are absent; the text deals mainly with the womb. The second enlarges upon the material contained in version 1 and includes images of foetus-in-uterus. The use of Latin is widespread. The text is usually introduced by a prologue that explains the need for a manuscript of this type: (2)
For as moche as ther ben manye women that hauen many diuers maladies and sekenesses nygh to the deth and thei also ben shamefull to schewen and to tellen her greauaunces unto eny wyght, therefore I schal sumdele wright to herre maladies remedye… And thowgh women have diuers evelles & many greet greuaunces mo than men knowen of, as I seyd, hem schamen for drede of repreving in tymes coming & of discuryng off vncurteys men that loue women but for
Middle English medical books as examples of discourse colonies
59
her lustes and for her foule lykyng… And therefore, in helping of women I wyl wright of women prevy sekenes the helping, and that oon woman may helpe another in her sykenesse & noght diskuren her previtees to suche vncurteys men (London, British Library, Sloane 2463, f. 194r; Rowland 1981: 58).2
This does not happen in the case of H whose introductory lines read (3)
Also we schulen vndirstonde that wymmen han lesse hete in her body than men & more moistnesse for defaute of hete that schulde drye her humours and her moistnesse/ But netheles kynde hath ordeyned wymmen a purgacioun at certeyn tymes of bledyng to make her bodyes cleene and hool fro siknesse (H, f. 149r).
The lines are copied directly from Anglicus’s Compendium medicinae, although in reverse order, i.e. the information on the nature of women is given first in the case of H: (4)
Retentionem menstruorum & alias matrices passiones paciuntur mulieres quibus <...> temporibus accidentaliter et quarter quibus naturalister . Naturaliter retinentur menstrual infra .xij. annos & vltra .l. accidentaliter retinentur intra xij & .l. Menstrua aunt naturalitater mulieribus adueniunt propter frigiditatem & humiditatem sue complexionis. naturaliter eium frigidiores sunt & humidiores viris (Anglicus 1510: ccxcrb).
This will be discussed in detail in section 4, below. What follows is a description of the theoretical framework used in the analysis of H. 3. Text and Hoey’s discourse colony My definition of text departs from the traditional concept in the sense that the semiotic component incorporates what is inherent in medieval manuscripts, i.e. the codicological and the palaeographic information, since these are essential to understanding the text elaboration process fully, along with the text’s history and readership. It also illustrates the dialogue between writer and reader through the numerous additions and corrections. All this said, I consider text as “an act of linguistic expression that involves semiotic, social, cultural and cognitive 2
See Bellwort (2002) for more information on early English midwifery manuals.
60
Francisco Alonso-Almeida
elements, but also morphological, syntactic, semantic and pragmatic elements that altogether provide sense and unity to this linguistic expression” (Alonso Almeida 2005). In this study I do not intend to focus on text from a genre and register perspective. Studies of historical discourse analysis aimed at describing texts as genres/text types include Alonso-Almeida (2002), Carroll (1999), Görlach (1992), Taavitsainen (1997 and 2001), and Pahta (1998 and 2001). Note the use of the oblique, however, to indicate that these concepts are not used clearly by scholars. As Carroll et al. (2003: 8) argue, genre and text type “are structured as fuzzy categories”. However, I think that the difference between the two concepts has been clarified by Taavitsainen (2001: 89) –following Halliday and Hasan (1989), and Halliday (1985), among others: “genres are groupings made on the basis of external features and functions of texts, whereas text types are defined according to the internal linguistic features of texts”. Moessner (2001) in his introduction to the special issue of the European Journal of English Studies on genre and register theory elegantly describes the history of the use of these terms.3 Hoey (1986) develops a new approach to the study of genres such as shopping lists and encyclopaedias that are often very complicated to analyse because they do not conform to standard text conventions from the perspective of coherence or cohesiveness. He claims that these texts are heterogeneous in form and use, but manifest homogeneity with respect to their discourse characteristics. A visual metaphor which may explain how these texts function is the beehive or the anthill metaphor: “the beehive and ant hill are made up of many independent units, which are not interconnected in a physical sense, and the loss of one or more of them will not affect the viability of the colony” (1986: 3). Obviously, this analogy cannot be applied to narrative texts in which the loss of a paragraph can complicate the interpretation of the text or render it meaningless; but it works for the types of discourse mentioned above, as I shall demonstrate in my analysis in section 4. The notions of utility and meaning are key concepts in this view of discourses as colonies. As Hoey (1986: 4) puts it, “a colony is a discourse whose component parts do not derive their meaning from 3 For more on genre and text types, see Eggins and Martin (1997) and Biber (1988, 1995).
Middle English medical books as examples of discourse colonies
61
the sequence in which they are placed. If the parts are jumbled, the utility may be affected but the meaning remains the same.” This implies that the order of the constituents in a given colony can be altered, or components removed, without altering the meaning and function. Utility is affected slightly in the case of the removal of components, since part of the information originally given is absent. Carroll (2003) has shown that recipes in a collection can be modified, excerpted and relocated without having an effect on meaning. They can also be included within another colony. Hoey’s newspaper example (1986: 6) is instructive: a daily publication contains sport pages, social pages and news pages and each section is a subcolony within the general text. The colony presents several of the nine properties listed in (5), below (Hoey 1986: 20); I include a description of each feature: (5) Property 1. Meaning not deriving from sequence 2. Adjacent units do not form continuous prose 3. There is a framing context 4. No single author and/or anon 5. One component may be used without referring to the others 6. Components may be reprinted or reused in subsequent works 7. Components may be added, removed or altered 8. Many of the components serve the same function
Description The order of constituents does not incur a change of meaning. The elements do not follow on one from the other. This has obvious implications for cohesion strategies in the texts. This refers to the presence of titles or introductory prefaces which contextualise the text that follows. Some colonies do not need this cue, as in the case of the shopping lists. This refers to authorship; including whether there is one or several authors, or an editor, a queen bee in Hoey’s words (1986: 12). Components are used independently of one another. This does not undermine the possibility of cross-reference within the colony. Items may be used in a new colony.
Items in a given colony are liable to be removed, added or relocated to tailor to the user’s needs. This is important for the study of a book’s history through the detection of corrections and additions in subsequent editions of a text. Items of a colony serve the same function within the colony. This is termed matching relations by Hoey (1986: 14); in this type of relationship, “statements are considered in respect of what they share or where they differ”. This stands in contrast with another type of relationship, i.e. sequence relations, in which “statements are seen as
62
9. Alphabetic, numeric or temporal sequencing
Francisco Alonso-Almeida following one from the other temporally or logically” (Hoey 1986: 14). This property relates to sequence relations as a cohesive device. This is construed by means of a system that is either arbitrary (alphabetic, numerical) or non-arbitrary (time and date).
In the next section, I apply the notion of discourse colony to H to demonstrate the presence of these properties in the medical manuscript. 4. Glasgow University Library, Hunter 307 as an example of discourse colony The analysis of H following Hoey (1986) has produced the results listed in (6), below. The remainder of the section is devoted to analysing each parameter of properties used to classify a colony. In the interests of clarity, parameters 1 and 2, and 6 and 7 will be grouped together in single sections. In discussing the data, I will use additional texts to support my arguments. Where applicable, these texts will carry their corresponding eVK numbers so that the reader may consult the database for extra information on the manuscript under survey. (6) Property
Score
1. Meaning not deriving from sequence
+
2. Adjacent units do not form continuous prose
+
3. There is a framing context
-
4. No single author and/or anon
+
5. One component may be used without referring to the others
?
6. Components may be reprinted or reused in subsequent works
+
7. Components may be added, removed or altered
+
8. Many of the components serve the same function
+
9. Alphabetic, numeric or temporal sequencing
-
Key: + affirmative condition, - negative condition, ? unclear or both conditions apply.
Middle English medical books as examples of discourse colonies
63
4.1. Meaning not deriving from sequence and continuous prose The first two properties, meaning not deriving from sequence and continuous prose, apply in H. Table (1) above gave a summary of all the contents pertaining to the manuscript. The items showed little connection among them from a topic perspective, a fact which is further illustrated by the incipits, listed below in (7). Observe, however, that a deeper pragmatic analysis of the text beginners is necessary in order to interpret correctly the role and behaviour of linguistic strings like Also, in (7c), within the framework of the book. (7) a. b.
c.
d.
e.
f.
It is to vndirstonde that a man is maid of four elements. And euery man hath iiij humours (f. 1r). A man that wole helpe men in her syknesse hym bihoueth to knowe the enchesions and the kyndes of the syknesse yat is to seie whey yei be hoote eiyer coold eiyer drye eiyer moiste. and this is by mony dyuerse toknes (f. 13r). Also we schulen vndirstonde that wymmen han lesse hete in her body than men & more moistnesse for defaute of hete that schulde drye her humours and her moistnesse (f. 149r). A phicisian behoueth to knowe three manere inspecciouns in blod letyng that is to seie whether it be thicke or thinne or meene whil a man bledith (f. 165v). Aloen Aloe is hote & druye in the secounde degree & it is maad of the iuys of an herbe that is clepid aloe & that herbe is founden in ynde in perse & in grece & in apuleie also (f. 167ra). Ache of the head…foll 16 (iiir).
As we can tell simply by looking at the incipits, there are no markers to structure the text hierarchically, save for the word also in (7c) to introduce the gynaecological tract. This discourse marker shows the connection between the treatise on female diseases and the preceding pharmaceutical tract. Both pieces of writing share the same author, i.e. Gilbertus Anglicus. In the table of contents given on folio 173r (modern foliation), they are recorded as if the sixteenth-century scribe identified the two as a single unit. From a codicological perspective, however, the tracts are differentiated by means of a blank page left between the two. The decoration of the text beginning on f. 149r is also richer in detail, featuring a flourish motif that occupies a large portion of the margins. Another piece of evidence that leads me to think that these treatises were considered as separate units is the fact
64
Francisco Alonso-Almeida
that the treatise on female disorders very frequently appeared on its own –which is not to say that it was not known to appear jointly with a version of the Compendium medicinae, as in the present example. The contents also mark a difference so that, even when the two are taken together as a single tract, the resulting text is a colony. From here, we can deduce the second feature, i.e. adjacent units do not form continuous prose; the absence of textual linking particles confirms that items may be relocated in the book without producing a change in meaning. 4.2. Framing contexts H does not comprise a title or a general preface for the whole book. There is, nonetheless, a paragraph written in a later hand which serves as a contextualising cue to what is included in the volume: (8)
And old System of Physic in English; of which I have another Copy. This is more compleat by all that is said on female disorders in the end. Tis is to undirstonde that a man is maid of four Elements. And evry man hath iiij humors &c (f. iiv).
Dr William Hunter wrote these lines in the sixteenth century, and the copy he refers to is Glasgow University Library, Hunter 509. The items contain their own prefatory material that introduces the topic of each treatise. Sometimes these prefaces contain programmatic information that lends coherence to the treatise by indicating the order of the constituents, as in the following example: (9)
Also we schulen vndirstonde that wymmen hau lesse hete in her body than men & more moistnesse for defaute of hete that schulde drye her humours and her moistnesse/ But netheless kynde hath ordeyned wymmen a purgacioun at certeyn tymes of bldeyng to make her bodyes cleene and hool fro siknesse and they hay suche purgaciouns fro the tyme off xij winter age into the age of fufty winter netheles sum wymmen hau it lenger as thilke that ben of hi ޚcomplexioun /& ben norischid with hote metis & drinkis /&lyuen in moche reste & thei hau this purgacioun euery monethe ones but if it be wymmen that ben with childe or wymmen that ben of drie complexioun /& trauelen moche /for wymmen aftir that thei ben with childe til thei ben delyuered thei hau not this purgacioun for the childe in her wombe is norischid with that blod /& if thei hau purgaciouns in this tyme it is a
Middle English medical books as examples of discourse colonies
65
tokne that the childe refusith the blod ./& than the child is fallen into sum sikenesse . or it wole die in his modir wombe wymmen than ben of an hi ޚcomplexioun /& faren wel /& lyuen in moche ese hau this purgacioun ofter than ones in amonethe //& this blood that passith fro wymmen in the tyme of her purgacioun cometh out of the veynes of the modir/ [paraph mark] the modir is a skyn that the child is closed inne in his modir wombe/ and many of the greuaunces that wymmen hau comen of the modir /oon is stopping of the blood that thei schulden be purgid of as y seide/ [paraph mark] An othere is to moche flowing of this blood/& in vntyme /& this siknesse feblith wymmen moche an othere is precipitacioun of the modir/ an othere is whan the modir is flawe with inne an othere is a postum of the modir/ an othere is swellyng of the modir/ an othere is greuaunce in bering child & hardenesse or than thei be delyuered/ an othere is going out of the modir bynetheforth/ an othere is with holding of the secoundyne/ an other is ache of the modir [paraph mark] (in VegaDéniz 2002: 100-101).
Unlike what happens in the other items in H, the prologue gives information on the contents of the treatise, and also the structure according to which these are organised. The two linguistic referential devices are oon and an othere, used nine times. The latter is interpreted by referring back to the first element in the listing, that is, oon. However, the meaning of oon cannot be retrieved by any linguistic cue, and hence interpretation relies on other textual clues relating to the external context of culture. Only if the reader knows the genre can he or she understand when the writer is referring either to sections or to chapters of the tract. The history of the titles often used for this tract on the diseases of women is uncertain, as no one title has been found for either version of the manuscripts. Green (1992: 77) comments on this peculiarity in the following terms: (10)
No single title of either version is consistently found in the manuscripts. Two copies of Version1 (Wellcome 5650, Yale 47) refer to it as “The Sekenesse of wymmen” (which I have adopted as the title here), one “Sores and greuaunce that women have (Trinity 0.9.37), another “The secreate diseasys of women” (Bodley 178), one even by the peculiar title “Liber perucreseos Galieni” (Sloane 5). Version 2 is untitled except for the misleading “Liber Trotularis” in Royal College of Surgeons 129 a.i.5.
Apart from prologues and titles, another contextualising element is the later inclusion, probably from the sixteenth century, of
66
Francisco Alonso-Almeida
the table of contents, which I present in (11). Readers may like to note that the recipe on page 124 has been marked by the scribe:4 (11) Ache of the head—foll 16 Anatomy of the braine—16 fransy—17 litargie—20 falling euyll—21 apoplexy—24 of ye eyes—27 obtalmye—32 webbe of the eye—34 canker of the eye—37 <…> of sighte—38 pimples —39-40 <…> —40 wormes in a man—44 rinsing in the care—45 scafenes—46 paines of the nose—47 suma rosacea infection of ye nose— 50 of the mouth and stinking thereof— 51 of the teeth & ye rotten nesse—53 of the tounge & greifes—55 Squinaury—57 for snesse of the throat—59 strangiusse of the breast—61 plurasie—67 Spitting of blood—69 spitting of the glas & quicre—71 Tissike—72 lemes that be broken—73 default in breathing—75 disease of the hart—76 of the greifes of the stomake—81
4
<...> indicates unreadable text.
(col. b) defaulte of appetite foll. 82 default of thurst—83 chocking—86 casting & speweing—87 swelling of the stomake—93 apostem of the stomake—94 a hote stomake—97 fluxe—98 purging ye iiij humors—99 blody fluxe—102 colico pasio of a gutt colon collica passio of another gutt—104 of wormes in the guts—107 a man that canot goe to stale—109 the greifes of the lyuer—111 the spleene—118 greifes of the head—124 X for swelling of sinewes—124 of the reynes & humours—125 a postem of ye reines—128 of the blader—130 rising of blood—131 of Diabetes—133 a pricking of ye bladders—135 <...iphacke>—136 of ye yeard & flowing of food—139 swelling of the codd—141 of the <…are>—143 bleine in the fundament—145 of bubo—146 a postum of the thighes—147 of the toes—147 the nature of women 150 dropsie of the mother 154 sleyng of ye mother 154 floweing of blood —155 Suffocacion of the mother—157 precipitacion of the mother 159 wind of ye mother —162 a postem of ye mother 162 wounds & canckering of the moder 163 ache of the moder—163 greuaunce in beareing of childe—164
Middle English medical books as examples of discourse colonies
67
the Secondine—165 of blood letting—166 of ventoseing—166
This table of contents serves the purpose of helping the reader to find the information needed in a hurry, but it also helps to trace the history of the book. The fact that the treatise contents are only given in part in this table suggests either that the reader found the selection relevant, or reliable, for his purposes, or that the volume took its final shape after this time. Observe, for instance, that we counted up to ten chapters in the treatise on the diseases of women, but here the number rises to fourteen because the author has subdivided some chapters into more than one. 4.3. Authorship The criterion of anonymous authorship is fully met in H as no reference is made to the author. However, this does not mean that the sources of the items in the text are not identifiable at least by those in academic circles. Items 1 and 4 in (1) have not been identified, but 2/3 and 5 are excerpts from Gilbertus Anglicus’s Compendium medicinae and Platearius’s Circa instans, respectively. Notwithstanding, based on the textual evidence –which is none– we have to regard H as anonymous, since we cannot actually say who exactly was responsible for translating and adapting the excerpts. 4.4. One component not referring to others This property is observed in most of the items in H, except for the gynaecological treatise which features Also as a conjunctive (this point is exemplified and discussed in section 4.1, above). Apart from this linking particle, components in H are not related to each other by means of grammatical devices, though topics can be similar: for example, humours and blood description are referred to in item 1, while item 4 deals with the inspection of blood. However, the connection is not a very strong one, at least not one that would make either item depend on the other.
68
Francisco Alonso-Almeida
4.5. Components are liable to be reprinted or reused and components may be added, removed or altered These two criteria are strongly related. It was customary during the medieval period to form new manuscripts by selecting material from different sources so that volumes matched the needs of their readers. Taavitsainen explains this as follows: (12)
Both the manuscripts as a whole and the immediate contexts of specific writings are important. Shorter tracts often make up constellations within a manuscript (Barratt 427-8). A particular text may have been intended to relate in a specific way to the other texts in the MS, or it may form a part of a longer thematic unit. Relationships of this kind become evident from an examination of the context (Marx 70). Even when texts of apparently different kinds are written one after another, they reflect the tastes and needs of the reader and the compiler (1988: 133).
Her view is also shared by Keiser (1998: 3597) who claims that “scribes frequently modified what they found in their exemplars, deleting apparently irrelevant or confusing material, or at least abbreviating it, interpolating material from elsewhere to expand undeveloped or more relevant subjects and to add new information, and revising and rearranging material to serve a perceived need.” H presents excellent examples for these two criteria. The remedies in the Compendium medicinae have been widely reproduced, either in collections or as independent units, and, although the sources for medical recipes are generally difficult to trace owing to extensive copying, some connections can still be observed between anonymous receptaria and the Compendium. Two such examples are found in Glasgow University Library, Hunter 185 (T.8.17), which I here contrast with H and another version of the Compendium: (13)
For the colica passio. Tak paritore & seeth it wel. As hoot as the seke may suffre it, ley it to his wombe & bynde it faste to. Hit schal cese. Also, take comyn & poudere of pepur & seeth it in stale ale. Soupe it hoot & it schal cese (GUL, Hunter 185, in Alonso-Almeida 2000: 267). And if this sikenesse come of wynd it is heelid in the maner as thouޚ it come of cold humours/ netheles ther ben special medecynes therfor oon is mak a plastre of comyn sode in wyn & lei to his wombe but
Middle English medical books as examples of discourse colonies
69
first mak him a suppositorie ¶ Also make him a fumygacioun of peritorie sode in wyn & let him drinke of that wyn and plastre that herbis on his wombe for peritorie is a special medecyn for siche wyndis bothe withinforth or withoutforth (H, ff. 106r-107v). And if this sekenes cometh of wynde, it is y-helid in the same maner as thou ޚit com of colde humours. Netheles ther be special medicines therfor. Oon is, make a plastir of comyn y-soden in wiyn and ley to his wombe. But first make him a suppositorie. Also make him a fumigacion of peritorie y-soden in wyin. And let him drinke of that wiyn and plastir the erbe on him. Peritorie is a special medicin for such wyndis, bothe withinforthe and withouten (London, Wellcome MS 537, in Getz 1991: 206).
The therapeutic functions of the three recipes are the same, but there are obvious differences in terms of structure, lexicon and, of course, spelling, although this last aspect has little relevance to our purpose. The most outstanding feature of this particular set of samples is the coincidence of the description of the remedy application in Hunter 185 and Hunter 307, while Wellcome 537 is less specific in regard to where the preparation should be applied. A second difference involves the remedy with cumin, which is given earlier in Hunter 307 and Wellcome 537 than in Hunter 185. In the first two, the cumin and the pellitory are specifically indicated to be sodden in wine; however, in Hunter 185, only the cumin is instructed to be soaked, and this time in ale. The treatise on the diseases of women has also been copied many times and disseminated throughout England. Green (1992: 7475) has pointed out how Yale University, Medical Library 47, also from Version 1, compares with the original by Gilbertus Anglicus. In the following table, I offer a summary of the correspondence of chapters in H with analogous chapters in Y and in A(nglicus): (14) H Preface: Also we schulen vndirstonde With holding of this blod Dropesie of the modir Sleyng of the modir To moche flowing of
Y Preface: Ye schal vnderstonde that Wyth haldynge of here blode Dropcy of the modere Fleynge of the modere To myche floynge of
A De retentione menstruorum De retentione menstruorum De ydropisi matricis De excoriatione matricis De fluxu nimio
70
Francisco Alonso-Almeida
blood Suffocacioun of the modir Precipitacioun of the modir Moche wynd ther is also in the modir A postum of the modir
blode Suffocacio matrices Precipitacioun of the moder Wynde yn the moder
menstruorum De suffocatione matricis De precipitatione matricis + De exitu matricis De ventositate matricis
For aposteme yn the moder Woundys yn the moder Cancur of the moder Ache of the moder
De apostemate matricis
Woundis of the modir Cancryng of the modir Ache of the modir Greuaunce that woman hau in bei3ring child The secoundyne
Greuaunce of the moder
Also women bleden otherwhil to moche aftir that thei hau bore child
Bledyng of wymmen after the birthe
Secundina matrice
De vulneribus matricis De cancro matricis De dolore matricis post partum De difficultate pariendi De secundina intus remanente post partum De fluxu sanguinis + De cancro matricis
This table shows that, apart from some variations in spelling and lexis, the same chapters are kept in the two versions of the Gilbertus Anglicus treatise. In fact, Green (1989) has found that another copy in British Library, Sloane 2463 of Version 2 in the Anglicus tradition presents material that is not given in H or Y. These components absent in Y include (Green 1989: 455-456): For as moche as ther ben many women (Preface), Also a worschipfull serip, Yff a woman be with childe, Greuaunce that women haue in bering (plus picture of foetus in uterus), Mola matricis is in two manners, Fyrst, yf she be repleted, Woman whan they ben with childe. In addition, Y features two further components that are shorter in that text than in British Library, Sloane 2463: Woundes of the marice and Cancring and festers of the marice; the first accounts for a tenth of the original, the second some thirty percent. The Circa instans was another well-known tract during the medieval period, and a number of versions have survived through time. Voigts and Kurtz (2000) have identified the following variants:
Middle English medical books as examples of discourse colonies
71
(15) Circa instans BL Sloane 105 BL Sloane 635 BL Sloane 297 Wellcome Library, London Med. Soc. 131 Wellcome Library 397 Bodleian, Bodley 178 Bodleian, Ashmole 1477 Bodleian, Ashmole 1481 BL Egerton 2433 CUL Ee.1.13 CUL Kk.6.33.III Gonville and Caius College 609/340 NY Academy of Medicine 13
Circa instans (excerpts) BL Sloane 297 GUL Hunter 307 TCC R.14.32
Garrido-Anes (2005) has edited the Circa instans text in Cambridge, Cambridge University Library, Ee.1.13, ff. 1r-91v. A comparison of the text in her edition with the text in H reveals that the entries in H are very much shorter than in CUL Ee.1.13. The Cambridge manuscript also contains many more herbal entries than H, and the scribe uses a great deal of shorthand. One illustrative sample is given in (16), below: (16)
Affodillus Affodille . it is an herbe & his leues ben liche the leues of lijke . h . d . in .2. [de]gree (H, f. 167r). Affodil. The rote is of more vertu to medicinis than the leues. And the herbe is better gren than druy. Hit hath vertu to consumen, and to druy, and to drawyn. For the morfew and allopycy take the powdre of been ibrant, and medul hhit with the juse of affadil, and hit is a good oynement for the forsaid thinges. Ffor strangory and dissure, in .iii. ownces of affadil, resolue .i. ownce of saxefrage, and an .i. ownce of gromel, and let hem boyle for to .ii. partis ben consumed. And seth, klanse hit and yef hit to the paciant with sugre. Ffor the dropsy take of the middil rynd of walwort and of elarun, and of ffilipendula, and of euerych of hem, .iii. dragmes. And seth hit in ½ .iii. ownces of the juse of affodil, and yef him to drinke (CUL Ee,1,13, ff. 6v-7r, in Garrido-Anes 2005: 254).
72
Francisco Alonso-Almeida
4.6. Many of the components serve the same function Property eight, i.e. many of the components serve the same function, indicates the utility of the items included in the manuscripts. In H, though the components are not tightly connected by linguistic cues, they do in fact complement each other in various ways. Their function is to give therapeutic information. In a more concrete sense, item 1 serves the function of introducing the theory of humours so that the reader has a better understanding of how a disease is generated. The Compendium offers solutions for humoral disorders. The same happens in the tract on the diseases of women, although the solutions in this case are given separately for practical reasons. Finally, the Circa instans provides information on a set of herbs, most of which are used in the pharmaceutical preparations. The information is varied, encompassing not only details to help readers identify the plant indicated, but also recommendations to supplement the medical advice suggested in the previous items. 4.7. Alphabetic, numeric or temporal sequencing The last property refers to the order placed on the constituents of a given volume. This can be arbitrary or non-arbitrary. In the case of H, the arrangement of items is completely arbitrary; there is nothing to suggest that any temporal sequencing is intended. Save for the case of also in the tract on the diseases of women, no other linking particle of this nature is used. Even in this instance, we cannot assume that the scribe really intended to imply any logical sequencing. Something different occurs with the organisation into chapters of the Compendium. Most of the chapters follow one another without any clear indication of sequencing. Thus, as we see in the Appendix, the last line of the chapter on the tongue does not indicate in any way what comes next, and neither does the first line of the new chapter on the throat make any reference to what came before. This facilitates not only the excerption of chapters to be inserted in other compendia, but the internal rearrangement of the material without any alteration of meaning. Utility, however, will be affected, although a comprehensive marginal apparatus could remedy that, whereby the chapters would function in exactly the same way as encyclopaedic entries. On this basis, the Compendium can be said to represent another example of a
Middle English medical books as examples of discourse colonies
73
colony in which the main purpose is to offer the solution to a problem. McDermott and Walsh (1991: 51) summarises the utilitarian function of colonies in the following way: “Hoey argues that discourse types have arisen to correspond to different reading strategies, and that the colony corresponds to the reading strategy of ‘scan[ning] a discourse with a view to finding the answer to a particular question (22-3)’”.5 The internal distribution of the information in the chapters follows a very simple arrangement into sections that can be diagrammatically represented, as follows: (17)
* (MN) ^ (DO) ^ DD ^ ¬ R
This schematic representation reads: the sections M(arginal) N(ote) and D(escription) of O(rgan) are optional and normally follow one another when they appear together; the MN section can also appear as an optional stage at other points in the discourse. The D(escription) of D(isease), which offers information regarding the causes, symptoms and associated affective ailments, is an obligatory stage that precedes the second obligatory stage R(emedy). This last stage is dependent (hence the symbol ¬) on the limit imposed by the number of diseases associated with an organ. The DD and the R sections are sometimes so strongly connected that it becomes very difficult to separate them; in such cases, the schematic representation in (17) should be modified to signal this closeness: * (MN) ^ (DO) ^ ¬{DD ^ R}. (A good example of this is found in the first part of the chapter on the mouth, given in the Appendix.) All the recipes in the chapters of the Compendium and the tract on female diseases are connected thematically, so much so that if an excerpt is copied to another compendium, the recipes often carry what they are good for in their titles, since this information cannot be retrieved later from the context. Taking one example from the Appendix, this would resemble something like the text in (18); the scribe may specify the remedy use or efficacy further by the addition of relativizers to the title: (18)
5
An eynement [for the tunge (+relativizer)] take of the iuys of sauge .j.3. of the iuys of wormod of maiorane & of march .j.3. of the baies of lorer umiperi castore .j.3. of marciaton & of oile of lorer .ij.3. of
The authors are referring here to Hoey (1986).
74
Francisco Alonso-Almeida reed wax .quarter.j. & medle hem to gidere and therwith anoynte the tunge.
The R(emedy) section in itself is also structured in a set of stages which can be outlined as follows: (5) (6)
(TT) ^ I ^ P ^ (A) ^ (E) Key: Title (TT), I(ngredients), E(valution)
P(reparation),
A(plication),
The TT is an optional stage that normally precedes the recipe and is highlighted by the use of certain typographical conventions, including punctuation. It is connected in purpose to the MN stage in the sense that it does not only show what the recipe is intended for, but also indicates the starting point of this new recipe. The I section offers the material needed to prepare the remedy and the quantities recommended. P indicates how the ingredients should be combined in order to get the finished therapeutic product. A provides information on how, when and where the product should be applied. The E stage informs about the efficacy of the preparation. All these stages in (17) and (19) are supported by a set of lexicogrammatical choices that are constrained by the variables of register and genre. However, this is beyond the scope of the present article and I will not even attempt a tentative analysis here owing to limitations of space. In the Appendix, I have tagged the last recipe as an illustration. The text types used in a chapter are varied, but the most common by far are exposition and instruction, according to Werlich’s (1976) and Biber’s (1989) classification. The pattern used throughout the chapter as a macrostructure is the PROBLEM-SOLUTION structure identified by Hoey (1994). The same pattern is also applied to the R section where the problem equates to the TT stage and the solution to the remaining stages. This PROBLEM-SOLUTION pattern has a sound didactic purpose, both in the Compendium and in the tract on the diseases of women (assuming that we consider this second tract as an independent item). As I mentioned previously, this type of organisation makes accretion and excerption easier, as sequencing is only strongly marked in the R section to make the manuscript function as a unit. One can successfully excerpt the DD or the R sections to reproduce them elsewhere without quoting the whole chapter, as we saw earlier in section 4.5.
Middle English medical books as examples of discourse colonies
75
In the sample chapter included in the Appendix, sequencing is linguistically expressed by oon and anothere in the string: A mannis tonge hath seruyses . oon is to taaste with a mannis metis & drynkis . anothere is to speke & thoru ޚdyuerse greuaunces that come of corrupt humours. Otherwise, I have not found any other similar linguistic cue. Thus, organisation relies on the disposition of contents described in the DD section where the diseases are attributed to certain causes, which are subsequently used to introduce the R sections. The sequencing in the other tracts operates in the same way as described here for the specific case of the Compendium. Item 4, the inspection of blood, presents a different organisation, though. Its contents are developed in a series of logical arguments but the series is not arranged according to any obvious criterion of order. The parameters used in the description of blood are its colour and consistency. Each type of blood is further explained in relation to the disease with which it is associated. For this the scribe uses cues such as also and now to structure the text. However, these terms should not be understood as we use them nowadays, as sometimes they express new independent information, rather than information to support the previous idea. The particles in question are marked in bold: (7)
Also in the bledyng me schal byholde whether it be hoot or cold . for hoot blood bitokneth hote humours to haue the maistrye in the body/ & cold/ ¶ Whan it engelith me schal biholde whether it be harsch or crassyng or cherkyng/ for if it be so it bitokneth moche corrupcioun in the body or disposyng to lepre/ ¶ also if it be fatty it bitokneth ouer moche fatnesse¶ Also if the smel of the blod be stynkyng . it bitokneth corrupcion of humour or rotnesse in the body (ff. 165v166r, in Vega-Déniz 2002: 135).
5. Conclusions In this article I have applied the notion of discourse colony to the text of H. I have found a variety of reasons why H may be considered an example of a discourse colony. The first of these is that the items are anonymous, although their readers may identify the sources of the manuscript components on the basis of their popularity within a particular discourse community (Swales 1990). Even if they do, however, that identification could be mistaken. Secondly, there is no
76
Francisco Alonso-Almeida
framing context to indicate clearly what the items are about, save for the presence of a tabula added a century later that does not include the full contents of the volume. A third characteristic is that meaning does not derive from sequence. This usually means that adjacent units do not form continuous prose, which is an additional feature of colonies. There is no linkage among the items in the volume in terms of a clear set of linguistic cues, except in the case of the tract on the diseases of women, which begins with Also as if expressing a connection with what has been said previously. A further feature of H as a colony is the possibility of excerpting and modifying the components therein, as well as that of adding new items. The sequencing of the components is arbitrary, and good evidence of this is the lack of grammatical markers to signal the structure. This study has also produced one unexpected conclusion. I have demonstrated that the text of H is an example of what Hoey has termed a discourse colony because it is made up of several components with little or no connection among them. However, I have also discovered that some of the items, such as the Compendium or the Circa instans, are examples of colonies within the volume. Within the framework of H, the different chapters and herbal entries may be modified, excerpted and used elsewhere in other manuscripts without affecting the meaning and the utility of the volume. Finally, the evidence would seem to indicate that H was created to meet the needs of his reader, probably a practitioner or a practitioner-to-be. The fact that the manuscript presents similar typographical conventions and style proves that the manuscript was planned beforehand with a definite scribal intention, most likely to fill an order. Bibliography Alonso-Almeida, Francisco. 2000. Edition and Study of a Late Medieval English Medical Receptarium. G.U.L. MS Hunter 185 (T.8.17). PhD thesis, University of Las Palmas de Gran Canaria. ___. 2002. ‘Punctuation Practice in a Late Medieval English Medical Remedybook’ in Folia Linguistica Historica 21(1-2): 207-232. ___. 2005. ‘All Gathered Together: On the Construction of Scientific and Technical Books in 15th-Century England’ in International Journal of English Studies 5(2): 1-25. Anglicus, Gilbertus. 1510. Compendium medicinae Gilberti anglici tam morboru[m] vniversali[u]m qua[m] particularium nondum medicis sed [et] cyrurgicis
Middle English medical books as examples of discourse colonies
77
vtilissimum. Iacobum Saccon [expensis Vicentii de Portonariis]. Lyon: Lugduni. Biber, Douglas. 1988. Variation Across Speech and Writing. Cambridge: CUP. ___. 1989. ‘A Typology of English Texts’ in Linguistics 27: 3-43. ___. 1995. Dimensions of Register Variation: A Cross-cultural Comparison. Cambridge: CUP. Bonilla, Salvador. 1997. ‘Estudio preliminar’ in de Beaugrande, Robert and Wolfgang Dressler Introducción a la lingüística del texto. Barcelona: Ariel: 7-25. Carroll, Ruth. 1999. ‘The Middle English Recipe as a Text-Type’ in Neuphilologische Mitteilungen 100.1: 27-42. ___. 2003. ‘Recipes for Laces: An Example of a Middle English Discourse Colony’ in Hiltunen, Risto and Janne Skaffari (eds) Discourse Perspectives on English. Amsterdam and Philadelphia: John Benjamins: 137-165. Eggins, Suzanne and James Martin. 1997. ‘Genres and Registers of Discourse’ in Van Dijk, Teun (ed.) Discourse as Structure and Process. London: Sage: 230256. Garrido-Anes, Edurne. 2005. De simplici medicina (Circa instans) en inglés medio: vernacularización del tratado salernitano de Mateo Plateario. PhD thesis, University of Huelva. Getz, Mary Faye. 1991. Healing and Society in Medieval England: A Middle Enlish Tranlation of The Pharmaceutical Writings of Gilbertus Anglicus. Wisconsin: University of Winsconsin Press. Görlach, Manfred. 1992. ‘Text-types and Language History: The Cookery Recipe’ in Rissanen, Matti et al. (eds) History of Englishes: New Methods and Interpretations in Historical Linguistics. Berlin and New York: John Benjamins: 736-761. Green, Manfred. 1989. ‘Women’s Medical Practice and Health Care in Medieval Europe’ in Signs 14(2): 434-473. ___. 1992. ‘Obstetrical and Gynaecological Texts in Middle English’ in Studies in the Age of Chaucer 14: 53-88. Halliday, Michael. 1985. Introduction to Functional Grammar. London: Arnold. Halliday, Michael and Ruqaya Hasan. 1989. Language, Context and Text: a social semiotic perspective. Oxford: OUP. Hellwarth, Jennifer Wynne. 2002. ‘“I Wyl Wright of Women Prevy Sekenes”: Imagining Female Literacy and Textual Communities in Medieval and Early Modern Midwifery Manuals’ in Critical Survey 14(1): 14-63. Hoey, Michael. 1986. ‘The Discourse Colony: A Preliminary Study of a Neglected Discourse Type’ in Coulthard, Malcolm (ed.) Talking about Text: Studies Presented to David Brazil on his Retirement. Birmingham: English Language Research, University of Birmingham: 1-26. ___. 1994. ‘Signalling in Discourse: A Functional Analysis of a Common Discourse Pattern in Written and Spoken English’ in Coulthard, Malcolm (ed.) Advances in Written Text Analysis. London: Routledge: 26-45. ___. 2001. Textual Interaction: An Introduction to Written Discourse Analysis. London and New York: Routledge. Keiser, George R. 1998. Works of Science and Information. Volume 10 of A Manual of the Writings in Middle English, 1050-1500. A. E. Hartnung (ed.). New Haven (Connecticut): The Connecticut Academy of Arts and Sciences.
78
Francisco Alonso-Almeida
McDermott, Anne and Marcus Walsh. 1991. ‘Editing Johnson’s Dictionary: Some Editorial and Textual Considerations’ in Small, Ian and Marcus Walsh (eds) The Theory and Practice of Text Editing. Cambridge: CUP: 35-61. Moessner, Lilo. 2001. ‘Genre, Text Type, Style, Register: A Terminological Maze?’ in International Journal of English Studies 5(2): 131-138. Pahta, Päivi. 1998. Medieval Embryology in the Vernacular: The Case of De spermate. (Mémoires de la Société Néophilologique de Helsinki 53) Helsinki: Société Néophilologique. ___. 2001. ‘Changing Conventions of Writing: The Dynamics of Genres, Text Types and Text Traditions’ in International Journal of English Studies 5(2): 139150. Rowland, Beryl. 1981. Medieval Woman’s Guide to Health: The First English Gynaecological Handbook. Kent (Ohio): Kent State University Press. Swales, John. 1990. Genre Analysis: English in Academic and Research Settings. Cambridge: CUP. Taavitsainen, Irma. 1988. Middle English lunaries: A study of the genre. (Mémoires de la Société Néophilologique de Helsinki 47). Helsinki: Société Néophilologique. ___. 1997. ‘Genres and Text Types in Medieval and Renaissance English’ in Poetica 47: 49-62. ___. 2001. ‘Middle English Recipes: Genre Characteristics, Text Type Features and Underlying Traditions of Writing’ in Journal of Historical Pragmatics 2(1): 85-113. Vega Déniz, Alicia. 2002. A Middle English Translation of Gilbertus Anglicus’s ‘De morbis mulierum’ G.U.L. MS Hunter 307 (ff. 149v-166v). Undergraduate thesis, University of Las Palmas de Gran Canaria. Voigts, Linda E. and Patricia D. Kurtz. 2000. Scientific and Medical Writings in Old and Middle English: An electronic Reference. [CDROM] Ann Arbor: Michigan University Press. Werlich, Egon. 1976. A Text Grammar of English. Heidelberg. Appendix: Text from H, ff. 54v-56v [marginal note: Tonge] [DO] A mannis tonge hath seruyses . oon is to taaste with a mannis metis & drynkis . anothere is to speke & thoru ޚdyuerse greuaunces that come of corrupt humours . a mannes tonge is otherwhiles vndisposid that a man may neither speke ne taaste ¶ [DD + R] Greuaunces of the tonge ben postemes & bleddris other stoppyng of his senewis thorou corrupt humours that maken the tonge laxe & falle into a palesye & if ther be apostem other bleyne of the tonge if it be of blod the tonge is to swolle & reed & hoot & neisch & akyng & if it is of colre the tonge is ful hoot & citryn that is to seie a derk reed the ache is more scharp than of blood/ but it is of flewme the tonge is coold & swellyng & neische & of white colour ¶ Of malencolie cometh cooldnesse & swart ޚelewe colour .& tunge ne swellith not but is sumwhat drawe togidere /and these ben the tokenes bi whom me may knowe what humours is the enchesoun of the palesie of the tounge and if the bleynes of the tonge ben blake other greene it is a tokene that the tonge is cancrid ¶ ffor postemes & bleynes that comen of [marginal note: ¶ Postemes & bleynes] coold let him vsen hoot gargarysmis to chewe in his mouth .& to make hym fnese & hoot fumygaciouns for the heued as
Middle English medical books as examples of discourse colonies
79
with anet mellilot camomylle & sausoke soden & let anoynte the heed with hoot oynementis that moun dissolue the materis ¶ But if it be of heete let hym vsen fumygaciouns gargarysmis & oynementis & emplastris on his heed as coold thingis & purgen his heed as we telden in the sikenesse of the heed . and let hym blood in bothe the veynes vndir the tonge .& sith let hym waische the mouth with coold thingis & anoynte the tunge as it was teld in the cleenyng of the lippis . hou me schulde heele bleynes of the mouth / and if the complexioun of the tonge be coold vse siche thingis as we schulen telle in the palesie of the tonge /and if the posteme be hard . make him a gargarysme of mylk & of watir that drie figis & fenel ben soden inne /and if it be of coold waische it with watir of camomylle & make a gargarysme of seueney & hony ¶ ffor the pymplees that ben of heete let hym hau the iuys of [marginal note: ¶ Pymplees] sicomor other of wey<...> ode vndur his tonge . other the iuys of solatre /& medle therwith red saudres & water sumac other balaustia other myerre other rooses ben soden inne ¶ A good poudre for siche postemes of the tonge take of dispumed hony .vj. ounces of the iuys of the erthe appel & of the melk of a fige tree .ij3. sethe hem togidre on an esy fyre & than medle therwith this poudre . Take tartarum of wyn & seueneth of ech .3.j.of white pepir & alitel sugir .3. l. of borac .3 .ij. of olibanum & of os sepie .3.ij. poudre alle these & medle hem with the othere forseid thingis .& ther with anoynte the sore withinne the mouth ¶ [marginal note: oynement] A white oynement that is good for the same greuaunce . take of mastik & of olibanum half an ounce . of litarge & of brend leed .j.3. of ceruse . iij.3. & do alle these in a morter & medle hem with a litil vynegre & sithe with a litil oile & eftsoones with vynegre & medle hem with oon aftir an othere til it be neither to thicke ne to thinne ¶ But if the tunge haue the palesie thorou sum corrupt humour that cumberith the senewis of what humours that it be: thou maist knowe bi the tokenes that we telden bifore ¶ And if it be of flood let him blood at the veynes that ben vndur the tonge and sith let holde castore vndur his tonge other the iuys of welde caul /& vndur stonde that caul seed comfortith moche the tonge and the throte/ the tokne therof is for the nyޚtyngale aftur that he hath ete siche seedis he syngith faste afturward/ let waischen jis mouth with oxymel & pepir & seueneth & ysope & origane/ Also let waischen his mouth with oxymel that maythen ben medlid with & let hym holden his mouth opene that the humours flowen out at the mouth / and let hym chewe drie piretre other piretre of mastik other staphie & mastik to gidere/ and let him froten his tonge with the poudre of piretre of pepir of staphine of seueney of castore of gynger of sal gemme either of sal nitre either of sal armoniac and summe castith ther to clowes notemuge galyngale cardomonim spyk & siche other thingis a gargarisme may be maid of the same thingis other thus take origane piretre gynger pepir white & blac & long canel ysope nygille sausuke cost of euerych ylych moche . tempre hem with oxymel . & here with thu myޚtist also frote the gummes on bothe sidis and the tunge aboue & [marginal note: sirip] binethe ¶ A sirip therfore take an ounce of the flouris of borage antheos drie sauge calamynte origane baies of lorer pyonthe lauandre the erbe of the palesie and of carses .j.3. of quybibis reed saundris rooses melisse amonium half an ounce poune hem & make up thi sirip with .ij.li. of sugir ¶ An eynement take of the iuys of sauge .j.3. of the iuys of wormod of maiorane & of march .j.3. of the baies of lorer umiperi castore .j.3. of marciaton & of oile of lorer .ij.3. of reed wax .quarter.j. & medle hem to gidere and therwith anoynte the tunge ¶ A stufe for hem schal be maad of mugwed mynte horehowne origane calamynte carses & of the erbe of the palesie and if the palesie be of coold other of plente of humours aboute the tunge /& thoo that hau this
80
Francisco Alonso-Almeida
sikenesse moten leeue al maner fleumatik meetis than ben yuele to defien & let hem vsen sauge in her meetis & leek & oynouns & leeue they strong wyn & drynke smal wyn & ne bathen they hem not moche & namely whanne thei ben repleet ¶ [R] A good gargarisme[TT] take origane piretre pepir blac & white and longe ginger canel ysope nygalle sausuke rost of euerych ylych[I] moche poudre hem & tempre hem with sape that is swete vynegre[P] & make vp thi gargarisme with oxymel[A]// (new chapter) [marginal note: Squinacie] Squinecye is a postum...
The second-person pronoun in late medieval English drama: The York Cycle (c. 1440)1 Rosa Eva Fernández-Conde University of Oviedo
Abstract PDE lacks number distinction for the second-person pronouns. This distinction was lost as a result of extending the use of the plural form to singular designation, and the consequent disappearance of the singular form. The first examples of this custom date back to the second half of the thirteenth century. This article presents the results of a qualitative and quantitative analysis of the uses of the second-person pronouns during this initial stage (i.e. Late Medieval English), in an attempt to widen the scope of previous studies (e.g. Nathan 1959, Evans 1967, Dillon 1969, Burnley 2003), dealing specifically with the dramatic texts of The York Cycle (c.1440).
1. Introduction The paradigm for the English second-person pronoun underwent a process of change that ended with the loss of number distinction. This loss was the result of extending the plural forms to singular reference and their final predominance over the singular forms around the eighteenth century.2 This extension of the plural seems to have come 1
This research has been funded by the Spanish Ministry of Education and Science (FPU program # AP2003-4815). I would like to acknowledge the editors and the anonymous reviewers of this paper for their suggestions. I am especially indebted to S. G. Fernández-Corugedo, for his help and observations. 2 The old singular forms of the second-person pronoun disappeared at least from the spoken standard (Baugh and Cable 1993: 237, Blake 1996: 260, Lass 1999: 153) but some exceptional examples can still be found. During the eighteenth and nineteenth
82
Rosa Eva Fernández-Conde
about under the influence or imitation of the French courtly practice by which plural forms of the second-person pronoun were used as a form of polite or courtly address (Barber 1976: 208, Burrow and Turville-Petre 1992: 41, Blake 1996: 219, 292, Lass 1999a: 148).3 Although the first examples of this practice can be found as early as the second half of the thirteenth century (Mossé 1991: 94, Blake 1992: 536, Burnley 2003: 28), according to Evans (1967: 34) its use was still mostly restricted to the upper classes around the midfourteenth century. The aim of this paper is to take a brief look at this issue and see to what extent speakers used plural forms for singular reference during the Late Middle English period (LME c.1350-1476). Similar studies in LME have focused mainly on the works of Chaucer (c.1343-1400, London English dialect: cf. Butler 2003, Honegger 2003 and Nathan 1959), the Gawain-poet (c.1375-1400, North-West Midland dialect(?): cf. Evans 1967) or Thomas Malory (c.1405-1471, Southern Midlands dialect: cf. Dillon 1969), but none to our knowledge has approached the issue from the perspective of a northern variety of English. It seems appropriate then that some attention be devoted to examining the impact of this LME linguistic trend on the Yorkshire dialect. 2. Textual sources: drama For the present study we chose as our corpus The York Cycle, using the electronic edition of Richard Beadle’s text (1982) that is available on the website of the University of Michigan (see references). We opted for a corpus of drama texts since, being in the form of dialogue, it presents a greater number of pronominal forms than other genres. Moreover, the York Plays are one of the most complete cycles of mystery plays. This meant we had not only a corpus of medieval century, for instance, singular forms are still common in Biblical texts and in literary works (Yaswen 2003: 1). And even today old singular forms are still found in some dialects of Northern England, such as West Yorkshire (Hogg 1992: 144), and in Scottish. 3 In fact, this courtly practice is also found in many other European languages (Baugh and Cable 1993: 237). During the Middle Ages, French was regarded in Europe as the language “representing chivalrous society in its most polished form, and the French language was an object of cultivation at most of the other Courts in Europe” (Baugh and Cable 1993: 131). This status may have helped to spread to other European languages certain French linguistic practices, like, for example, these polite usages.
The second-person pronoun in late medieval English drama
83
drama written in a northern dialect but a corpus of considerable size: the corpus contains a total of forty-seven plays plus a number of variant versions (see appendix for a complete list of the plays). The corpus contains 99,241 words and the approximate date of creation given by Beadle is 1440. However, the earliest reference to the cycle goes back to “as early as 1378” (Campbell 2005), so we can presume that the original fourteenth-century texts must have undergone revision up to the present edition (and up to the last known performance of the cycle in 1569). 3. The uses of the second-person pronoun in The York Cycle Examining the works of Brown and Gilman (1960) and Burnley (2003), we find that The York Cycle is characterised by a very deliberate use of the second-person pronoun. According to Brown and Gilman (1960), modern languages present a T/V system, a dual system of pronouns where V pronouns (usually the plural ones) are used as the polite, formal or honorific forms of singular address while T pronouns (usually the singular ones) are used as the informal or familiar forms of address. As we mentioned before, the first traces of this kind of system in Middle English began to appear round the mid-thirteenth century. However, this early system of address was more complex than that discussed by Brown and Gilman (Burnley 2003: 30). This polite system of pronominal address is, according to Brown and Gilman, closely related to the relationship established between speakers and, more specifically, to questions of power and solidarity. Power refers to the “degree that [a person] is able to control the behaviour of the other” (Brown and Gilman 1960: 255); while solidarity is seen as a relationship established between equals, between people who feel they are close to each other in some way. Therefore, whenever the relationship is based on terms of power,4 speakers will present a non-reciprocal use of the pronouns: i.e. the person with superior power uses the singular form of the pronoun 4
Brown and Gilman (1960: 255) suggest that a person might hold power over another as a result of the following social variables: “physical strength, wealth, age, sex, institutionalized role in the church, the state, the army, or within the family”.
84
Rosa Eva Fernández-Conde
(henceforth T-forms) and receives plural forms, or Y-forms, in return.5 Similarly, the degree (or lack) of solidarity between people is also reflected by a reciprocal use of the same pronoun: either T-forms (to reflect familiarity, intimacy, like-mindedness) or Y-forms (to express formality, reverence or deference). By studying these factors, we were able to establish a series of dyads that reflected the most common relationships between interlocutors. We will begin by looking at those dyads based on power relations, which consequently present a non-reciprocal pattern of pronominal exchange. 3.1. Power-based relationships: common dyads and special uses 3.1.1. Social superior vs social inferior: Master-Servant, KingSubjects (Giver-Petitioner) The epitome of power semantics is the relationship between master and servant. Between them there is always an explicit contract of service on the part of one person (the servant) to another (the master) by which the latter has power over the former. Most of the time this superiority in power is matched by a superior social status. In The York Cycle, we identified many examples of this relationship consistent with Brown and Gilman’s power semantics classification. For instance: Abraham and his servants (in play 10), Herod and his messenger (in play 19) or Pilate, Annas, Cayphas and their soldiers or messengers (in plays 28, 29, 32, 33, 38). However, this is not always the case. For instance, Cain’s servant, Brewbarret, seems to enjoy a certain privileged status in relation to his master. In an incomplete scene from the seventh play, we see how Brewbarret repeatedly offers excuses (e.g. a sore back, pain in the shoulder, a broken toe…) to avoid doing what his master asks him to do. Far from scolding Brewbarret, Cain ends up entreating his servant to help him, i.e. to do 5 In the present paper, we will use the label “T-forms” to refer to all the LME forms of the second-person singular pronoun (e.g. thou, thy, thee, thyself and any variant spelling that may appear in the texts). “Y-forms” will refer to all the LME plural forms (e.g.: ye/yoe, your(s), you, yourself, spelling variants included). We must not forget that until the seventeenth century we cannot talk about a regularised spelling system so the same word could have several allographs. For example, the oblique case you was found written as: yoou, yow, yoow, yowe, among other variants.
The second-person pronoun in late medieval English drama
85
his duty.6 The servant’s incongruous status is clearly reflected in the way Cain addresses him and switches from T-forms to more honorific terms like sir and Y-forms: (1) Brewbarret: Lo, maister Cayme, what shaves bryng I, Evyn of the best for to bere seyd, And to the feylde I wyll me hye To fetch you moo, if ye haue neyd. Cain: Cume vp, sir knave, the devyll the speyd, Ye will not come but ye be prayd. Brewbarret: O, maister Caym, I haue broken my to! Cain: Come vp syr, for by my thryft, Ye shall drynke or ye goo. (Play 7, 73-82) (BREWBARRET: Lo, master Cain, what sheaves bring I!/Even of the best to bear seed./ And to the field, I will me hie,/ To fetch you more, if you have need. CAIN: Come up, sir knave! The devil speed thee!/ You will not come unless you are prayed! BREWBARRET: Oh! Master Cain, I broke my toe! CAIN: Come up now, sir! for my thrift, / You shall drink before you go!)
The reason for this switch is Cain’s anxiety to hide his brother’s corpse. Hence, in spite of being the master, he chooses to honour his servant by employing the Y-form. By doing this, Cain’s address appears less an order than a request. And, as with any requester, in order to achieve his object he has to adopt the most appropriate communication strategy, in this case dignifying his servant by means of Y-forms. Similar deliberate switches of pronoun are also found in other requests: in play 17, for instance, when Joseph speaks to the Presbyter; in play 26 when Judas addresses Pilate, Anna, Cayphas, and so on. In all these examples, the speakers address their hearers using Y-forms and receive T-forms in return. We may say, therefore, that in these examples, Y-forms are not used to mark the 6
Although the T-form in the fragment is used in an ironic context, the fragment illustrates how Cain switches from one form to the other as he changes from “ironic” order to genuine entreaty.
86
Rosa Eva Fernández-Conde
social superiority of the hearers but to flatter and persuade them in favour of the speaker’s request. This is shown clearly in play 30 when the Beadle addresses Pilate using the expected Y-form and adorns his address with the following words: (2) Bedellus: My liberall lorde, o leder of lawis, O schynyng schawe that all schames escheues, I beseke you my souerayne, assente to my sawes, As ye are gentill juger and justice of Jewes. (Play 30, 55-59) (BEADLE: My liberal lord, oh, leader of laws,/ Oh, shining show, who all shame eschews,/I beseech you, my sovereign, assent to my words /As you are gentle judge, and justice of Jews.).
As we can see, the Beadle’s communication strategy is obviously that of flattery. He tries to ingratiate himself with Pilate and incline him towards accepting his request. This is why he uses Yforms and, in spite of his high social status, receives T-forms in return. In this fragment, the Beadle is clearly depicting himself as inferior in power by means of this gratuitous flattery. This is shown more clearly in the following fragment, where Judas addresses a janitor (a lowclass servant) using a Y-form: (3) Judas: [after a lengthy speech] Do open, porter, the porte of this prowde place That I may passe to youre princes to proue for youre prowe. Janitor: Go hense thou glorand gedlyng, God geue the ille grace, Thy glyfftyng is so grymly thou gars my harte growe. Judas: Goode sir, be toward this tyme and tarie noght my trace, For I haue tythandis to telle. (Play 26, 155-160)
The second-person pronoun in late medieval English drama
87
(JUDAS: open, porter, the door of this proud place, / So that I may pass to your princes, to prove your good profit. PORTER: Get hence, thou glowering rascal! God give thee ill grace;/ Thy glaring is so grim thou give my heart pain. JUDAS: Good sir, be toward this time, and tarry not my trace, / For I have tidings to tell.)
The social status of the janitor is not very different from that of Judas, yet he uses Y-forms to address him. According to their similar low social status, the expected form should have been a reciprocal use of solidarity T-forms, however, this is not the case.7 Again, Judas is trying to ingratiate himself with the janitor who overtly despises him as a “glorand gedlyng”. Nevertheless, in another scene (see example 4), Peter and Philip (two other disciples) use TForms to request a donkey from a different janitor. (4) Peter: Sir, with thi leue, hartely we praye THis beste that we myght haue. Janitor: To what intente firste shall yoe [plural] saye, And than I graunte what yoe [plural] will crave Be gode resoune. (Play 25, 71-75) (PETER: Sir, with your leave, we heartily pray / That we this beast might have. PORTER: To what intent first you shall say,/ And then I may grant what you crave/ In good season.).
Although in both these scenes (examples 3 and 4) we find Christ’s disciples making a request to a janitor, they do not use the same pronominal form. In fact, the different choice of pronoun points to two different strategies to incline the interlocutor in their favour. We can conclude, therefore, that in the relationship between giver and petitioner the choice of pronoun is not pre-determined as in the other 7
Judas’ request is to enter the palace where he will betray Jesus. Once he betrays Christ, he becomes one of the lowest characters in moral terms. Hence, from that moment onwards, he always receives T-forms and uses Y-forms in return. However, at this precise moment, he has not yet been disloyal and his use of Y-forms is clearly aimed at flattering the janitor. There is no clue in the text to indicate any other reading, as, for example, that Judas’ use of Y-forms reflects his low moral status resulting from the base motive of his visit.
88
Rosa Eva Fernández-Conde
relationships. Pronoun selection is really based on the context and on the combination of the speakers’ status, the power of the interlocutor and the goal of their speech act. Another interesting case, comparable to that of master-servant, is the relationship between the king and his subjects, or more precisely between the king and his “personal court”. Even though they occupy a high social status and are employed as counsellors to the monarch, the members of the king’s private court behave as common servants and, as such, always address the king using Y-forms and receive T-forms in return: (5) Dux 1: My lorde, vnlase you to lye, Here schall none come for to crye. Rex Herod: Nowe spedely loke that thou spie That no noyse be neghand this none. Dux 1: My lorde, youre bedde is new made, you nedis noyot for to bide it. Rex Herod: Ya, but as thou luffes me hartely, laye me doune softely, For thou wotte full wele that I am full tendirly hydid. (Play 31, 34-48) (DUKE 1: My lord, unlace you to lie;/ Here none shall come to cry KING: Now speedily look that thou spy,/ that no noise be near here this midnight. DUKE 1: My lord, your bed is new made, you need not abide. KING: Yeah, but if heartily thou love me, lay me down softly;/ For thou know very well I am tenderly hided.)
Similar situations are found in play 11 between the Pharaoh and the Egyptians and counsellors around him. However, there are many other examples where this distinction is not observed: e.g. between Herod and his counsellors (play 19), or between Pilate and Anna, Cayphas or the Doctors (plays 26, 30, 32, 33, 36, 38). All represent the same type of power relationship, i.e. a king and his personal court, a ruler and his subjects, but in the latter examples Yforms are used reciprocally between speakers. We will explain the reasons for this later on.
The second-person pronoun in late medieval English drama
89
3.1.2. Genealogy and family relations: Parents and Children, Husband and Wife Another example of social superiority and power semantics in The York Cycle is found within family relations. According to Brown and Gilman (1960), the power that parents exert upon their children is reflected in a non-reciprocal use of the pronouns: the parents address their children with T-forms and receive Y-forms in return. This pattern is followed with no exception throughout the cycle. It does not matter whether the family has a low social status (like the family of Noah, Abraham or even Christ, for example) or a high one (like that of Pilate or Herod). In all these families, the children address both their parents with Y-forms and receive T-forms in return.8 (6) Filius 1: Fadir, we shal nouyot fyne To youre biddyng be done. […] Filius 1: Where are ye, modir myne? Come to my fadir sone. Uxor:
What sais thou sone? (Play 9, 51-57)
(SON 1: Father, we shall hear / Till your bidding is done. […] SON1: Where are you, mother dear? / Come to my father, soon! WIFE: What sayest thou, son?) (7) Filius: Fadir, if thai like noght to listyn youre lawes, As traytoures ontrewe ye sall teche them a trace, For fadir, vnkyndnes yoe kythe them no cause. Herod:
Faire falle the my faire sone, so fettis of face. (Play 16, 46-49)
8
It is very interesting to observe that when Jesus is dying on the cross and tells Mary to take John as her son, even in this forced mother-son relationship, John addresses Mary with Y-forms and in return receives T-forms.
90
Rosa Eva Fernández-Conde (SON: Father, if they like not to listen to your laws, / As traitors untrue you must teach them their place; / For, father, for unkindness you give them no cause./ HEROD: Fairness befall you, my fair son, so handsome of face.)
The relationship between husband and wife is more complex. Taking into account that the plays were written in the Middle Ages, we would expect women to present a lower status than men. During the Middle Ages, the position of women in society was secondary, even in the domestic sphere since the man was considered the head of the family. Hence, we expected women to use Y-forms consistently towards their husbands and in return receive T-forms. However, this non-reciprocal pattern of pronominal exchange between husband and wife was only found between Joseph and Mary. The reason for this may be, as suggested to me by Cristina Mourón Figueroa, that Joseph is much older than Mary.9 Indeed, certain fragments show him treating Mary as his daughter and, as we have noted, parents and children present a non-reciprocal pattern of pronominal exchange. Nevertheless, some exceptional and temporary switches, like the example below, were also found:10 (8) Joseph: A, Mary, blyssed be thowe ay, Thowe thynkes to do after Goddes wyll, As thowe haist said Mary, I say, I will hartely consent theretyll […]. Mary: Therto am I full redy dight, But one thyng Joseph I wolde you meyve. Joseph: Mary my spouse and madyn bright, Tell on hartely, what is your greyf? Mary: […] Both beest and fewll hus muste neydes haue, As a lambe and ij dove-byrdes also. 9 Although we have not looked at this social variable in the present paper, age is also a determining factor in the choice of T- or Y-forms. The older the speaker, the greater deference he or she must receive. Hence, in this case, Mary could also be addressing her husband with the polite and honorific Y-forms used to address hearers who are much older than the speaker. 10 In this example, Joseph switches to Y-forms as a way of honouring and approaching his wife so as not to make her feel afraid to speak her thoughts.
The second-person pronoun in late medieval English drama
91
Lame haue we none nor none we crave, Therefore Joseph what shall we do, What is your read? (Play 17, 223-240) (JOSEPH: Ah, Mary, blessed will thou be ay; / Thou thinkest to do after God’s. / As thou have said, Mary, so I say; / I will gladly consent theretill / […] MARY: To do so, I am ready here. / But one thing, Joseph, I would you move. JOSEPH: Mary, my spouse and maiden clear. / What is your trouble? Tell me this. MARY: […] Both beast and birds we must needs have: / A lamb and two dove-birds also. / But we have no lamb, nor may not crave. / Therefore, Joseph, what is your advice?)
Unlike Joseph and Mary, the other couples mainly present a reciprocal use of pronouns. In the cases of Adam and Eve and Noah and his wife, we find that they usually address each other using Tforms. Pilate and his wife also present a reciprocal use of the pronouns, but they use Y-forms. (9) Uxor:
Wher arte thou Noye?
Noah:
Loo, here at hande,
Come hedir faste dame, I the praye. Uxor:
Trowes thou that I wol leue the harde lande
And tourne vp here on toure deraye? (Play 9, 75-79) (WIFE: Where are thou, Noah? NOAH: Right here at hand. / Come here, and quickly, I thee pray. WIFE: Thinkest thou that I’ll leave the land/ And turn up here in all this fray?)
92
Rosa Eva Fernández-Conde (10) Pilate: the same.
Nowe saye itt may ye saffely, for I will certefie
Uxor: gayne.
Gracious lorde, gramercye, youre gode worde is
Pilate: Yhitt for to comforte my corse me muste kisse you madame. Uxor: fayne.
To fulfille youre forward my fayre lorde I am (Play 30, 46-49)
(PILATE: Now, you may say it safely, for I will certify the same. PROCULA: Gracious lord, great thanks now; your good word is my gain. PILATE: Yet, to comfort my flesh, I must kiss you, madam. PROCULA: To fulfil your command, my fair lord, I am fain.)
It would be nonsense to claim that Adam and Eve or Noah and his wife share a more intimate relationship than that of Pilate and his wife on the basis of their use of the reciprocal T-form of solidarity or the Y-forms of relations of lesser solidarity. Although possible, there is no clue in the texts to lead us to this conclusion. Besides, a marriage with a low level of solidarity would use a non-reciprocal pattern of pronominal exchange as this would represent the greater distance between them. Hence, the most obvious reason is that, in this case, pronouns are used as a social index, as an indicator of the couples’ “social identity” (Calvo 1992: 23) and not of their relationship. As we noted before, the use of Y-forms for singular address began as a refined courtly practice associated with manners. During the LME period, there was a considerable “hunger for reverence” among the upper classes of society (Burnley 2003: 33) which favoured the use of Y-forms for singular address among members of the highest social status to differentiate them from the lower classes who used T-forms. The practice became so widespread that it soon came to denote the higher social status of the hearer or both speaker and hearer, whereas the use of T-forms denoted the lower social standing of the speakers (see Evans 1967, Mossé 1991, Baugh and Cable 1993: 237, Dyer 1994, Blake 1996: 260, Lass 1999). Pilate and his wife, as members of the upper social classes, present this reciprocal use of Y-forms (plus other courteous love addresses) to denote their superior social status
The second-person pronoun in late medieval English drama
93
and the same “hunger for reverence” is reflected in other situations where most of the characters are of high status and speakers present a reciprocal pronominal exchange (e.g. Herod and his counsellors, Pilate and his personal court or the high priests Annas and Cayphas). 3.2. Status-marked situations One of the most interesting features uncovered by our analysis of the pronominal forms is, actually, the high percentage of T-forms (67%) compared with the percentage of Y-forms (31%) used for singular address:
Singular address pronominal forms
2% 31%
Y-forms with singular reference T-forms Dubious Y-forms
67%
Figure 1: Percentages for the total number of second-person pronouns used for singular address in The York Cycle
Our quantitative analysis would seem to support the idea that this courtly practice was still not very common during the LME period. However, a qualitative analysis shows that The York Cycle does present a very deliberate pattern of pronoun use. The explanation for the great difference between the percentages may have to be sought elsewhere. Burnley (2003), who centred his work on LME literature, concluded that pronoun use depends on the genre of the text. For example, the more courtly a text is, the more common it is to find Yforms used, hence courtly genres would favour the use of Y-forms, such as courtly romances (Blake 1992: 538) or the most refined and courtly of Chaucer’s The Canterbury Tales, “The Knight’s Tale”
94
Rosa Eva Fernández-Conde
(Honegger 2003: 65). Non-courtly genres, like drama, would favour the use of T-forms. In our case, not only the genre but the fact that most of the characters belonged to a low social status seemed also to favour the use of T-forms since, as we established above, Y-forms were used as a social index of high social status. However, some upper social class characters, like Nicodemus and Joseph of Arimetea (two important and wealthy Jews), always talk in reciprocal T-forms. Even Pontius Pilate, a Roman official, sometimes exchanges T-forms with high priests Annas and Cayphas. Our feeling is that in these examples Y-forms are not so much a reflection of the speakers’ status or relationship but of the relation of these factors to the situation. If we put together all the scenes where Y-forms were found, we discover that Y-forms are mainly used not on the basis of the speakers’ status but because they are in formal situations. Formal situations are situations where upper-class people meet in order to deal with the subject of status or public affairs. Formal situations usually take place in an official building (e.g. Herod’s palace, the Pharaoh’s Court) or in a public space where some figure of authority is present (e.g. Herod, Pilate, the Pharaoh, even the Roman centurions). Therefore, the presence of a high governmental or religious authority or any Roman authority –what Butler calls “tyrants” (2003: 148)– can turn the scene into a formal situation, a “status-marked situation” (Ervin-Tripp 1972: 227); a situation where “status is clearly specified, speech style is rigidly prescribed, and the form of address of each person is derived from his social identity”. Hence, social superiors are addressed with Y-forms from both upper and lower social classes as a result of being in a status-marked situation and having a high social status. This is corroborated by the fact that almost 83% of the Y-forms of The York Cycle appear in just 15 plays where formal and statusmarked situations of this kind take place.
The second-person pronoun in late medieval English drama
Figure 2: Ratio of T- and Y- forms in the whole cycle
3.3. Addressing God: Solidarity-based relationships To complete this overview of the different uses of pronouns in The York Cycle, my last (but not least) aspect examines how God is addressed. We cannot make an analysis of the pronouns in a cycle of mystery plays without dealing with what is for some the controversial issue of addressing God (see Honegger 2003: 70). According to Brown and Gilman (1960) we could have expected for God, as the highest power in the world, to be addressed with a Y-form. However, this is not the case. In The York Cycle, as in other written records (Blake 1992: 539, Burnley 2003: 28, Butler 2003), God addresses and is addressed using T-forms. According to Burnley (2003: 36), this is because “religious discourse is presented to dissent from worldly values” and its “hunger for reverence”. The reason behind this choice of T-forms would then be similar to those furnished later on by Quaker: to erase all traces of social differences. In our opinion, however, people used T-forms with God because of a growing feeling of closeness to Him. Although The York Cycle features scenes from both the Old and the New Testament, the God that appears in the plays is closer to that of the New Testament: our Lord, father and friend. In the case of Adam and Eve, God can be considered almost a third partner in Paradise, and in the others God was also seen as a close friend, lord or father (e.g. mercyfull maker 1.41, lufly lorde 1.43, blyssid lorde 3.61, worthy lorde 8.49, father of heven 17.345). In our opinion, this feeling of friendship, love and closeness grew out of regular prayer. God is
96
Rosa Eva Fernández-Conde
usually addressed by people praying in private contexts and, usually, asking for help. Once again, the best strategy would seem to be to appeal to His mercy and compassion through the use of the familiar Tform. People usually talk to God about personal affairs and they generally do it in the privacy of their own company. The situation is not marked and so they do not have to follow any social convention and are free to use the more familiar T-form. Unlike the shifts found in The Canterbury Tales (Honegger 2003: 70), in The York Cycle all the conversational exchanges with God (in His presence or otherwise) are performed using T-forms. His angelical messengers and His son Jesus Christ also present this same pattern almost without exception since they are seen as an extension of God. Only in some exceptional instances is Jesus addressed with the more honorific Y-form, by Peter and Mark, for instance, who seem to do so to reinforce his role as their master, teacher and prophet. We see this, for example, in The Last Supper: (11) Marcellus: Maistir, we haue arayed full right Seruise that semes for youre sopere. Oure lambe is roste and redy dight, As Moyses lawe will lely lere. (Play 27. 5-9) (MARK: Master, we have well arrayed / A seemly service for your supper. / Our lamb is roasted, and ready made; / To Moses’ law we did adhere )
Another example can be found in play 22 where Jesus is tempted by the devil. Here, the angel addresses Jesus with the Y-form to remind him that he is the son of God and as such his master and that he should avoid the devil’s temptation. (12) Angel 1: A, mercy lorde, what may this mene? Me merueyles that yoe thole this tene Of this foule fende cant and kene Carpand yoou till, And yoe his wickidnesse, I wene, May waste at will.
The second-person pronoun in late medieval English drama
97
Methynke that yoe ware straytely stedde Lorde, with this fende that nowe is fledde. Jesus: Myn aungell dere, be noyot adred, He may not greue; The haly goste me has ledde, THus schal thow leue. For whan the fende schall folke see […] (Play 22, 181-193) (ANGEL: Ah, mercy, Lord, what may this mean? / I marvel you endure this pain; / of this foul fiend, artful and keen, / Talking to you still-- / And you his wickedness, I ween, / May waste at will! / Methinks that you were strictly stead, / Lord, with this fiend that now is fled. JESUS: My angel dear, be not afraid; / He may not grieve! / The Holy Ghost, has me led; / Thus shall thou believe.)
This careful use of T- and Y-forms does not only reflect the role and relationship of the speakers but is also the means by which “the York cycle thus strategically constructs … rhetorically a Christian understanding of the universe, particularly incarnational theology” (Butler 2003: 154). 4. Conclusions We have seen that, by the mid-fifteenth century, the use of both plural and singular forms for singular address had already extended into the Yorkshire dialect. Yet, the different uses second-person pronouns presented remains a complex issue. The fact that pronouns reflect social relations is clear from what we have seen. However, it is not so much a reflection of the interlocutor’s power, solidarity, status or relationship, as a combination of all of these elements. In the choice of one or other form, there are many factors at play: closeness of the interlocutors, deference, social variables (eg. status, genealogy, power, solidarity), the object of the communication (request, dealing with personal or public affairs), and context (e.g. public/private place, presence of “tyrants”). The pronoun use of certain characters in The York Cycle is conditioned, to a considerable extent, by these variables (e.g. greater or lesser degree of closeness, solidarity or deference; higher or lower
98
Rosa Eva Fernández-Conde
status, more or less important nature of the affairs) and the interrelation among them. Depending on the speakers’ prioritisation of social status, context or communicative goal, they choose one or other form. For example, Pilate gives more importance to deference and to social status so he uses Y-forms; Noah and his wife prioritise their relationship and use T-forms; Judas and Cain place their objective first over the rest of the variables and switch to Y-forms to flatter their interlocutors. In short, the choice of pronominal forms is a contextconditioned choice by the speaker. To understand the reasons why a person uses one or other form requires a careful analysis of the situation, these variables as well as other contextual elements –e.g. vocatives, verb forms, tense and agreement, other forms of address. It is only then that we can understand the meaning speakers wished to convey through the pronoun used (e.g. flattery as in the Beadle’s example, reverence as in Pilate’s, irony, mockery, disrespect and even rage). Unfortunately, this kind of analysis is beyond the scope of this paper. Nonetheless, we have been able to establish that the use of the pronouns in the LME period is not a simple question. In fact, we must admit that, despite the efforts of present-day researchers, there are still a great number of switches and uses that cannot be explained by existing theories. What we can safely conclude, though, is that pronouns in the LME period were not yet used in the same systematic way as they are today. Bibliography Primary Sources Beadle, Richard. 1982. The York Plays. Arnold, London. E-Text on-line at http:// www.hti.umich.edu/cgi/c/cme/cme-idx?type=header&idno=York (consulted 30.07.2005). Secondary sources Barber, Charles. 1976. Early Modern English. London: André Deutsch. Baugh, Albert C. and Thomas Cable. 1993 (1951). A History of the English Language. London: Routledge. Blake, Norman F. (ed.). 1992. The Cambridge History of the English Language, Vol. II, 1066-1476. London: CUP. ___. 1996. A History of the English Language. London: Macmillan. Brown, Roger and Albert Gilman. 1960. ‘The Pronouns of Power and Solidarity’ in Thomas A. Sebeok (ed.) 1971 (1960), Style in language. Massachusetts: MIT Press Cambridge. 253-276.
The second-person pronoun in late medieval English drama
99
Burnley, David. 2003. ‘The T/V distinction in late Middle English’ in Taavitsainen, Irma and Andreas H. Jucker (eds): 27-46. Burrow, John A. and Thorlac Turville-Petre. 1992. A Book of Middle English. Oxford: Blackwell. Butler, Michelle M. 2003. ‘All hayll, all hayll, both blithe and glad’: Direct Address in Early English Drama, 1400-1585. Duquesne University, PhD Thesis. Campbell, Kathleen. 2005. Notes on the English mystery plays. The English Mystery Plays. On-line at: http://artemis.austincollege.edu/acad/hwc22/Medieval/ english_mystery_plays.html (consulted 20.07.2005) Calvo, Clara. 1992. ‘Pronouns of address and social negotiation in As You Like It’ in Language and Literature 1: 5-27. ___. 1996. ‘The Tragicall Historie of Hamlet, Prince of Denmarke and the Pronouns of Address: Q1 (1603) vs Q2 (1604/5)’ in Sederi VII: 17-21. Dobson, Eric J. 1997. ‘Early Modern Standard English’ in Lass, Roger, Historical Linguistics and Language Change. Cambridge: CUP: 419-439. Dyer, Christopher. 1994. Everyday Life in Medieval England. London: The Hambledon Press. Evans, William W. 1967. ‘Dramatic use of the second-person singular pronoun in Sir Gawain and The Green Knight’ in Studia Neophilologica 39: 38-45. Ervin-Tripp, Susan M. 1972. ‘Sociolinguistics Rules of Address’ in Pride, J.B. and Janet Holmes (eds) Sociolinguistics: selected readings. Penguin: Harmondsworth: 225-240. Görlach, Manfred. 1991. Introduction to Early Modern English. Cambridge: CUP. ___. 1999. ‘Regional and social variation’ in Lass, Roger (ed). The Cambridge History of the English Language, Vol. III, 1476-1776. Cambridge: CUP: 459538. Hogg, Richard M. 1992. ‘Phonology and morphology’ in Hogg, Richard (ed.) The Cambridge History of the English Language. Vol. 1. The Beginnings to 1066. Cambridge: CUP: 67-164. Honegger, Thomas. 2003. ‘“And if ye wol nat so, my lady sweete, thanne preye I thee, [...]”: Forms of address in Chaucer’s Knight’s Tale’ in Taavitsainen, Irma. and Andreas H. Jucker (eds): 61-84. Jucker, Andreas H. and Irma Taavitsainen. 2003. ‘Introduction’ in Taavitsainen, Irma and Andreas H. Jucker (eds): 1-25. Lass, Roger. 1999a. ‘Phonology and morphology’ in Lass, Roger (ed.): 56-186. ___. (ed). 1999b. The Cambridge History of the English Language, Vol. III, 14761776. Cambridge: CUP. Mossé, Fernand. 1991 (1952). A Handbook of Middle English. Baltimore: The John Hopkins University Press. Nathan, Norman. 1959. ‘Pronouns of Address in the Canterbury Tales’ in Mediaeval Studies 21: 193-201 Nevalainen, Terttu and Helena Raumolin-Brunberg. 1993. ‘Early Modern British English’ in Rissanen, Matti, Merja Kytö and Minna Palander-Collin (eds) Early English in the Computer Age: Explorations through the Helsinki Corpus. Berlin and New York: Mouton de Gruyter: 53-73. Taavitsainen, Irma and Andreas H. Jucker (eds). 2003. Diachronic Perspectives on Address Term Systems. Amsterdam: John Benjamins.
100
Rosa Eva Fernández-Conde
Walker, Terry. 2000. ‘The Choice of second person Singular Pronouns in Authentic and constructed dialogue in late 16th century’ in Mair, Christian and Marianne Hundt (eds) Corpus linguistics and linguistic theory: papers from de Twentieth International Conference on English Language Research on Computerized Corpora (ICAME 20). Amsterdam: Rodopi: 375-384.
Different paths for words and money: The semantic field of “Commerce and Finance” in Middle English Isabel Moskowich and Begoña Crespo University of A Coruña
Abstract Our aim in this paper is to examine the relationship between society and language change. Developments such as the discovery of new territories and the introduction of new and exotic products and practices are known to have caused a significant increase in the number of terms with economic nuances from the sixteenth century onwards. We believe that the economy is a good indicator of relations among different groups of people and of the way in which those relations contributed to the evolution of language even from the early stages, a conclusion we base on the etymological origin of specialised terms introduced during Middle English.
1. Introduction The main purpose of this paper is to investigate the relationship, if any, between the frequency of terms relating to trade and finance and their etymological origin over a period in the history of the English language prior to what is referred to as the Early Modern English period, throughout which commercial relations between England and other countries were intense. The first part of this study will be concerned with the historical events and other external factors which may have brought about the expansion of vocabulary relating to commerce. Our aim is to examine the relationship between society and language change. Developments such as the discovery of new territories and the introduction of new and exotic products are known to have caused a significant increase in the number of terms with economic nuances from the sixteenth century onwards. Less attention,
102
Isabel Moskowich and Begoña Crespo
however, has been paid to what may have happened immediately before this period, that is to say, in Middle English. The survey proposed here will be carried out within the theoretical framework of Historical Sociolinguistics. As Nevalainen and Raumolin-Brunberg (2005: 34) have put it: “…the social factors affecting linguistic variation have also become the focus of interest in their own right, despite the problems researchers encounter in reconstructing the social realities their informants lived in”. They insist that the external circumstances restricting linguistic variation “will have to be reconstructed on the basis of what we know about the past societies themselves”. In this respect, it is most important to bear in mind that results obtained in this type of analysis are always biased because we can only measure the written evidence of the speech community, not the actual performance itself. Results, therefore, will be partial only. Having looked at the historical background, section 3 will be devoted to describing our corpus of data. We will refer primarily to the samples of ME contained in the Helsinki Corpus of English Texts. Diachronic and Dialectal (HC henceforth). As we will see, the criteria followed include the need to deal with all text-types1 as well as all varieties of English in a period characterised by its high dialectal diversity. Once the data have been established, the next section (4) will focus on analysing the terms selected using various computer resources. The above-mentioned analysis revolves around two main axes, etymological and chronological. Finally, section 5 will deal with the conclusions obtained from our examination of the data. The relationship between etymological origin and frequency of occurrence will demonstrate that the history of the vocabulary of a language is inextricably related to the historical heterogeneity of the different peoples implicated in the life of that language. From a social point of view, the field of “commerce and finance” represents an activity which called for linguistic interaction and mutual intelligibility since the exchange of products among speakers of different languages could not be effective without mutual understanding. In addition, exchange with foreign countries was necessary more often than not in a society which was increasingly 1
What was referred to as “text-types” at the moment the HC was published would now probably be described as “genres”.
Different paths for words and money
103
urbanised and money-based. Therefore, economy can be regarded as a good indicator of relations among groups of people and the way in which their relations contribute to the evolution of language. 2. The socio-historical background of Middle English The fact that “language does not exist in a vacuum; [but] it is used in concrete cultural and historical circumstances” (Anttila 1989: 377) opens the way for a simultaneous study of culture, historical events and language changes in terms of the established relationship among them. External factors give rise to internal language changes (Fernández 1993; Millward 1996; Thomason 1997; Hughes 2000; Nevalainen and Raumolin-Brunberg 2003, 2005), especially in certain fields which are quicker to feel the influence of these exogenous circumstances. As some scholars have already pointed out, the vocabulary of a language is the subsystem which best mirrors the relationship between language and society (Leith 1983; Su Jinzhi 1992, Tejada Caller 1999). In Knapp’s words (2000: 8), “… close attention to the range and tone of words is important evidence of how social change is being naturalized into the life-worlds of men and women, and therefore how tightly linguistic and social histories are woven together”. Social, economic, political and demographic events in the period concerning us (the second migration of the Vikings, the Norman Conquest, the religious movement of Wycliffe and the Lollards, The Black Death, The Peasants’ Revolt, The Hundred Years War, taxation and commercial transactions with other countries) took on a new significance in the development of the English lexicon from the beginning to the end of the Middle Ages. As we will see, each event can be linked to developments in separate linguistic strata. The successive waves of Scandinavian migrations to England began in the eighth century. However, the massive arrival of Old Norse speakers to the Isles took place from the tenth century onwards and began on the eastern coast. Though the first of these speakers had arrived some three centuries earlier, the real linguistic effects were not appreciated until the eleventh century (Moskowich 1995). In fact, from the eighth up to the beginning of the eleventh century very few loans are apparent. Most of them are of a miscellaneous character to designate types of ships or to refer to people, along with a number of
104
Isabel Moskowich and Begoña Crespo
other social and legal terms. From the beginning of the eleventh century until the middle of the twelfth century we encounter Scandinavian terms relating to the sea, social life, law, commonplace objects and even trade (toft “plot of land”, bond “tie”, knife, call, mal “payment”, coup “buy”, scogh “wood”, marc “money”, thrift “material wealth”). During the ME period, dialectal variation also accounts for the greater or lesser presence of Scandinavian borrowings. Naturally, documents and literature from the northern region contain many more Norse and Danish elements than those in the south. The Latin substratum, meanwhile, can be accounted for by the Norman Conquest on the one hand, and by the pervasive effect of Christianisation in the spheres of art and learning, on the other.2 Some examples of this contact include: intellect, mechanical, secular, submit, testify, zenith. After the Norman Conquest (1066) another foreign language left its mark on the vernacular. The language of prestige and power in England from the twelfth to the fourteenth century was to be Norman French. Contacts between the two peoples resulted in a linguistic hybrid known as Anglo-French or Anglo-Norman, which became the official language of justice, politics and the court, not the language of everyday speech among the conquered. This peculiar relationship between the two peoples is reflected in the different semantic fields to which loanwords belong. Written records from the eleventh century provide few examples of loans; they become more numerous from the twelfth century onwards. According to Serjeantson (1935: 216-239) they cover the following fields: person-rank, finance, buildings, law, social relations, religion, military, nature, clothes, household objects, physical action, appearance, faculty, moral and intellectual and miscellaneous. In texts from the centuries that followed, the presence of technical, religious, abstract and learned terms is more pronounced. French loanwords in English convey so many different shades of meaning that some scholars (Serjeantson 1935) have found it 2
We will not look into the earlier contact between the inhabitants of England and Latin speakers from before Christianisation because the inhabitants of England at that stage did not speak English.
Different paths for words and money
105
necessary to include a further subdivision of the moral and intellectual field.3 Though the ascription to different semantic fields of Scandinavian and French loans may appear unbalanced, an explanation can be sought in the socio-external context. The Scandinavian population intermingled with the natives thereby precluding many terms from being considered specialised. When Scandinavians finally decided to settle in England, they did so in places already inhabited by an English-language speech community (Moskowich 2002). The coexistence of both peoples led to the assimilation of linguistic habits because of the similarity between the two languages and so the native speech community adopted Scandinavian terms without feeling they were foreign (Moskowich 1995). This was the case for very basic terms such as take, both and are. The Normans’ attitude towards the native population was completely different. The relationship between the two peoples was not established on equal terms. For at least a century the linguistic effect was that French terms were regarded as exclusive to certain spheres of life and semantic fields because far from integrating into ordinary English life, the Normans kept linguistically aloof, thereby compounding the low degree of genetic relation between the NormanFrench and the English. Historical events also indicate close relations between English and other languages. The lexicon of Middle English adopted examples of Flemish and Dutch (mart “market”, pickle, guilder). Common interests were mainly commercial (handicraft, merchants, sheep, wool, fishermen, profit, ryche, pore, stor, wage), sometimes military (in times of war), and later, in the sixteenth century, religion and art served as contact-points. Occasionally, language can function as a tool to reveal the way in which society developed. We know, for example, that commercial contact with the Germans was based on the exchange of minerals and money because terms originating in High German abound in the field of mineralogy (zinc, nickel). As Crespo (2000: 31) points out, other languages such as Arabic (maravedi, amber, cotton), Celtic (quay, vassal), and even Indian (pepper, ginger) and Semitic 3
Baugh and Cable (1993) provide extensive lists of vocabulary items which were borrowed from different linguistic strata, especially from Latin and French.
106
Isabel Moskowich and Begoña Crespo
dialects (emerald, coral) can be said to have left an imprint on the vocabulary of trade affairs in English during the fourteenth and fifteenth centuries as a result of cultural and commercial contacts. 3. Corpus material Here we present the empirical research data that have led us to some general conclusions about the semantic field of commerce and finance during the Middle Ages. The data we will be using for this analysis are taken from the Helsinki Corpus of English Texts (1991). Since the HC is mainly a diachronic corpus, all periods of the history of the English language are represented there. However, we will focus on Middle English, the period the HC compilers situate between 1150 and 1500. All the texts contained in the above-mentioned corpus are classified by means of COCOA headers according to a number of different parameters of information. For our analysis we will draw on two of these parameters: date and text-type. With respect to date, the HC makes an additional division within each of the traditionally accepted periods of the history of the English language, establishing a total of four subperiods for the Middle English period: M1: 1150-1250; M2: 1250-1350; M3: 1350-1420; M4: 1420-1500. Another decision adopted by the compilers of the HC was to add information about the type of text each of the samples recorded is taken from. In the interests of uniformity, we have selected samples containing between 2000 and 2500 words from all the text-types included in each of the four sub-periods. This makes a total of 102,182. Table 1 below illustrates the different types of text represented in each of the sub-periods since not all of them appear all of the time.
Different paths for words and money
107
Table 1. Number of words from each sample according to period and text-type
PERIOD ME I 1150 – 1250
ME II 1250 – 1350
TEXT-TYPE Handbooks Philosophy Homilies Religious treatises History Biography, lives Documents Homilies Religious treatises History Biography, lives Fiction Romances Bible Undefined
ME III 1350 – 1420
Documents Handbook, astronomy Handbooks, medicine Science, medicine Philosophy Homilies Sermons Rules
TITLE PERI DIDAXEON VESPASIAN HOMILIES, NO. III <sample 2> ORM, THE ORMULUM HISTORY OF THE HOLY ROODTREE LAYAMON <sample 1> KATHERINE <sample 1>
WORDS 2121 1949
THE PROCLAMATION OF HENRY III KENTISH SERMONS
362
DAN MICHEL, AYENBITE OF INWYT ROBERT OF GLOUCESTER THE LIFE OF ST. EDMUND (THE EARLY SOUTH – ENGLISH LEGENDARY) DAME SIRITH; INTERLUDE THE ROMANCE OF SIR BEUES OF HAMTOUN THE EARLIEST COMPLETE ENGLISH PROSE PSALTER SONG OF THE HUSBANDMAN SATIRE ON THE CONSISTORY COURTS SATIRE ON THE RETINUES USK, APPEAL(S) CHAUCER, A TREATISE ON THE ASTROLABE A LATE MIDDLE ENGLISH TREATISE ON HORSES A LATIN TECHNICAL PHLEBOTOMY CHAUCER, BOETHIUS THE NORTHERN HOMILY CYCLE ENGLISH WYCLIFFITE SERMONS THE BENEDICTINE RULE
TEXT-TYPE Religious treatises History Travelogue Fiction
Letters, nonprivate Bible ME IV 1420 – 1500
Law Documents Handbooks, medicine Handbooks, other Handbook, astronomy Science, medicine Sermons Rules Religious treatises Prefaces Proceed, depositions History
Biography, lives Fiction
TITLE THE PRICKE OF CONSCIENCE
WORDS 2343
CURSOR MUNDI <sample 1>, <sample 2> MANDEVILLE’S TRAVELS CHAUCER, THE GENERAL PROLOGUE TO THE CANTERBURY TALES HENRY V, LETTER LETTER(S), LONDON THE OLD TESTAMENT (WYCLIFFE) STATUTES (II) INDENTURE, PETITIONS (M4) SHILLINGFORD (DOCUMENTS) THE LIBER DE DIVERSIS MEDICINIS’ IN THE THORNTON MS REYNES, THE COMMONPLACE BOOK METHAM, DAYS OF THE MOON
2315
THE CYRURGIE OF GUY DE CHAULIAC MIDDLE ENGLISH SERMONS... MS. ROYAL AELRED OF RIEVAULX’S DE INSTITUTIONE INCLUSARUM KEMPE, THE BOOK OF MARGERY KEMPE CAXTON, THE PROLOGUES AND EPILOGUES DEPOSITIONS GREGORY, THE HISTORICAL COLLECTIONS OF A CITIZEN OF LONDON THE LIFE OF ST. EDMUND CAXTON, THE HISTORY OF REYNARD THE FOX
2141 2150
2433 2080 2374 2298 2230
2302 2210 2193 2264 1846 2331 2029 1975 2136
2078 2098
Different paths for words and money
PERIOD
TEXT-TYPE Romances Drama, mystery plays Letters, private Letters, nonprivate Bible
109
TITLE THE SIEGE OF JERUSALEM IN PROSE LUDUS COVENTRIAE
WORDS 2218 2337
SHILLINGFORD (LETTERS)
2135
PASTON, WILLIAM
2198
ROLLE, THE PSALTER OF PSALMS 2177 OF DAVID
As the table illustrates, the earlier sub-periods exhibit fewer types than the more recent ones which is consistent with the general scarcity of written texts immediately after the Conquest. Since our main interest is to examine the evolution of terms relating to the sphere of commerce and finance, we have decided to limit the scope of our analysis to nouns. In his doctoral dissertation about the diachronic semantic classification of religious vocabulary in English, Chase (1988) affirms that, for religious vocabulary at least, nouns are the most important part of speech semantically. Since language uses verbs to represent states and events, and nouns to embody things, and most vocabulary items denote things, nouns are the part of speech with semantic primacy. Room (1991) argues the same principle.4 4. Data retrieval and analysis From the 102,182 words that form our corpus we have selected those corresponding to the nominal category (17,945 or 17.56%) of which 234 (1.3%) belong to the semantic field of “commerce and finance”. Almost half of these occurrences (140) appear in the last sub-period between 1420 and 1500. This gives us an idea of the use of financial terms increasing in parallel with the rate of social evolution. In search of some explanation for the scarcity of financial vocabulary in the earlier sub-periods, we might reflect on all the historical events mentioned in section 1. The persistence of the feudal system, for 4
The same theory is also subscribed to by Halliday, Sager, Dungworth and MacDonald (1986) and Gotti (1992), among others, for the scientific register.
110
Isabel Moskowich and Begoña Crespo
example, restrained the possibility of economic expansion, which only began after feudalism had faded away and the urbanisation process gradually taken root. As we see in Graph 1 below, the semantic field of “commerce and finance” is represented in all the selected sub-periods: 140 140 120 100 80 60 40 20 0
38
43
13 1150-1250 1250-1350 1350-1420 1420-1500
Graph 1. Commerce and finance nouns for each sub-period
Notwithstanding, its presence is not uniform throughout the Middle Ages: as might have been expected, only 13 occurrences (5.5% of all the nouns examined) were identified in ME1 (feor, mede, maðmes, æht), probably owing, on the one hand, to the fact that it is too early to observe the effects of the reforms introduced by the newcomers (the Normans); changes take time to settle in. On the other hand, the low frequency of terms may due to the fact that most of the texts recorded in this period belong to the religious sphere. The middle centuries of the period (between 1250 and 1420) show a greater balance in terms of the number of occurrences: ME2 contains 38 occurrences, which constitutes 16.24% of the total number of finance nouns (chaffare, shilling, coustage, gauelinge,tax, pris, arrerage, peni, gold); ME3 contains 43 occurrences or 18.38% (marchande, rente, cost, welles, eschaunge, purchas, ferthyng, bargaynes). Although at this stage the Normans with their institutions and bureaucratic system were already well established, they were not completely integrated into ordinary Anglo-Saxon life. If the number of terms relating to commerce and finance is not greater, this is because the period was one of social and, therefore, economic instability caused by the constant revolts and
Different paths for words and money
111
political struggles taking place, such as the Barons’ War, The Peasants’ Revolt, The Black Death and The Hundred Years War. It was the beginning of the end of feudalism. Most remarkable is the number of these nouns contained in the last period: 140 or 59.8% of the total (mynte, rate, somme, a-countes, payement, tidynges, levy, talliages, taxacions, profit, brybe, purchasour, spense, dokatys, debyteyys, fees). This reveals that once the most urgent problems had been solved, there was an increasing concern for social welfare and national prosperity. In a study of socio-economic terms from the twelfth to the eighteenth century Crespo (2002) concludes that the semantic sub-fields of “mercantile finance” and “money” prevailed over others such as “property”, “taxation” or “profit”. The basis for this conclusion lies, once again, in the interaction between sociohistorical events and language evolution: the period under survey seem[ed] to constitute a time in which it was necessary to show evidence of the commercial deals of the royal institutions through written records at a social or general level, whereas money and property or possessions appear to be the most relevant economic matters at an individual level. All in all, they seem to alter the meaning of some vocabulary items so as to adopt economic overtones or, even, new terms are created to cover the linguistic vacuum of economic development. (Crespo 2002: 265)
For each of the tokens recorded in our database we have considered different types of information which may reveal something about the history of some of these words. As stated earlier, we examine date and etymological origin in relation to the semantic field of finance and commerce. The information concerning this last aspect has been obtained from the Middle English Dictionary.5 All the terms examined have their origins in one of the following source languages: Old English (OE), Old French (OF), Anglo-French (AF), Latin (L), Old Norse (ON), Middle Dutch (MDu) and Middle Latin (ML); and they can all be grouped in two families: Germanic (OE, ON and MDu) and Romance (OF, AF, L and ML). Within these two groups the predominant source languages are OE and OF but, with the passage of time, a wider range of origins for these financial loanwords is observed, as table 2 below illustrates. 5
Access to the MED on line has been funded by the Xunta de Galicia, grant number PGIDT 99 PXI 10401 A. This grant is hereby gratefully acknowledged.
112
Isabel Moskowich and Begoña Crespo
Table 2. Etymological origins for ME nouns of finance and commerce SUB-PERIOD 1150-1250 1250-1350
1350-1420
1420-1500
ETYMOLOGY OE OF OE OF AF OE OF L AF ON OE OF L AF MDu ML
No. 12 1 24 12 2 21 15 1 5 1 54 74 1 6 3 2
The same information is displayed in graph 2: 80 70 60 50 40 30 20 10 0
1150-1250 1250-1350 1350-1420 1420-1500
OE OF
AF
L
ON Mdu ML
Graph 2. Etymological origins for ME nouns of finance and commerce
Old English is found to be the predominant source language for financial terms at the beginning of the Middle English period (gold, penne, æht) whereas Old French stands out between 1420 and 1500,
Different paths for words and money
113
which is consistent with the provenance of English vocabulary in general (vsure, dette, coin, prouffyte). Once more, the middle centuries exhibit a more balanced range of origins; the socio-external explanations adduced for the balanced number of financial nouns during these centuries can also be applied here. The same factors can be argued to account for the presence of one single non-Old English noun between 1150 and 1250 (coste) as opposed to the abundance of non-Old English nouns (83 nouns) between 1420 and 1500. It is also significant that of these 83 loans, 74 belong to OF. 5. Conclusions Our intention in this paper has been to study the extent to which financial terms are present in medieval English texts as well as to identify the etymological origins of those nouns, while referring to the possible socio-external reasons for them. We know that the Anglo-Saxons maintained commercial relations with neighbouring peoples. However, the evidence from our data suggests that until a certain level of stability was acquired in late Middle English, that intercourse was not manifested linguistically in texts. A closer inspection of these terms also shows that the same historical events may account for the presence of loans belonging to different source languages. The low presence of different languages at the beginning of the period is in marked contrast to their high presence at the end of the period under survey. It is surprising though, that languages such as ON and MDu are only rarely found as sources for loans in the semantic field of our analysis since both the Scandinavian and the Dutch maintained intense commercial relations with the Anglo-Saxons. Two main reasons can be identified for this: firstly, since OE and ON share a common Germanic origin it can be difficult to distinguish the actual origin of a particular item (it is easier when there is a divergent phonological evolution in the languages involved). Secondly, some of the terms taken from other Germanic languages probably belong to a nonformal register and would therefore not have been considered as appropriate as Old French ones for use in writing. We may conclude that, in this particular case, written language does not accurately reflect the social reality. Though we know for certain that commercial exchanges took place primarily
114
Isabel Moskowich and Begoña Crespo
between the English and other Germanic-speaking peoples, the vocabulary used in common transactions was not fixed in writing. It was the language of those in power that was to be recorded as the vehicle of expression of economic and social growth. Bibliography Anttila, Raimo (ed). 1989. Historical and Comparative Linguistics. Amsterdam and Philadelphia: John Benjamins Publishing Company. Bammesberger, Alfred. 1984. English Etymology. Heidelberg: Carl Winter Universitätsverlag. Baugh, Albert. C. and Thomas Cable. 1993. A History of the English Language. London: Routledge and Kegan Paul. Chase, John P. 1988. A Diachronic Semantic Classification of the English Religious Lexis. PhD thesis. Universidad de Glasgow. Crespo García, Begoña. 1996. ‘English and French as L1 and L2 in Renaissance England: A Consequence of Medieval Nationalism’ in SEDERI VII: 107114. ___. 2000. ‘Historical Background of Multilingualism and its Impact on English’ in Trotter, David A. (ed.) Multilingualism in Later Medieval Britain. Cambridge: D. S. Brewer. ___. 2002. ‘A Preliminary Approach to the Semantic Analysis of Socio-economic Terms in the History of English’ in Quaderni di Semantica XXIII(2): 257272. De la Cruz Cabanillas, Isabel and Francisco Javier Martín Arista (eds). 2001. Lingüística histórica inglesa. Barcelona: Ariel. Fennell, Barbara A. 2001. A History of English. A Sociolinguistic Approach. Oxford: Blackwell Publishers. Gotti, Maurizio. 1992. ‘The Development of a Scientific Language in the Seventeenth Century’ in Nocera Avila, Carmela, Nicola Pantaleo and Domenico Pezzini (eds) Early Modern English: Trends, Forms and Texts. Fasano: Schena editore: 319-343. Hughes, Geoffrey. 1988. Words in Time. A Social History of the English Vocabulary. Oxford: Basil Blackwell. ___. 2000. A History of English Words. Oxford: Blackwell Publishers. Knapp, Peggy A. 2000. Time-bound Words. Semantic and Social Economies from Chaucer’s England to Shakespeare’s. Basingstoke, Hampshire: Macmillan. Knowles, Gerry. 1997. A Cultural History of the English language. London: Arnold. Kurath, Hans et al. 1959-. Middle English Dictionary. On-line at: http://ets.umdl. umich.edu/m/med/ (consulted 28.06.2001). Machan, Tim W. 2003. English in the Middle Ages. Oxford: OUP. Millward, Celia M. 1996. A biography of the English language. Fort Worth, Texas: Harcourt Brace College Publishers. Moskowich-Spiegel Fandiño, Isabel. 1995. Los escandinavos en Inglaterra y el cambio léxico en inglés medieval. A Coruña: UDC Servicio de Publicacións.
Different paths for words and money
115
___. 2002. ‘El mito vikingo: el escandinavo como el “otro” en la Inglaterra medieval’ in Cuadernos del CEMYR (Centro de Estudios Medievales y Renacentistas, Universidad de la Laguna) 10: 55- 79 Nevalainen, Terttu and Helena Raumolin-Brunberg. 2003. Historical Sociolinguistics. London: Longman. ___. 2005. ‘Sociolinguistics and the History of English: A Survey’ in International Journal of English Studies 5(1): 33-58. Rissanen, Matti et al. 1991. The Helsinki Corpus of English Texts. The Diachronic Part. Helsinki: Helsinki University Press. Room, Adrian. 1991. NTC’s Dictionary of Changes in Meanings. A Comprehensive Reference to the Major Changes in Meanings in English Words. Lincolwood, Illinois: NTC Publishing Group. Serjeantson, Mary S. 1935. A History of Foreign Words in English. London: Routledge and Kegan Paul. Singh, Ishtla. 2005. The History of English. London: Hodder Arnold.
Contents Introduction Isabel Moskowich Spiegel-Fandiño and Begoña Crespo García Part 1. Linguistic aspects of early English The (im)possibility of stacking adjectives in early English Agnieszka Pysz Lists in letters: NP-lists and general extenders in early English correspondence Ruth Carroll
Part 2 Language and culture
How might Everyman have been performed? John McKinnell University of Durham
Abstract This paper describes the rationale and methods of a 1999 production by Durham Medieval Theatre Company of the English morality play The Summoning of Everyman, which is a translation of the Dutch Elckerlyc, a rhetoricans’ play from Antwerp which won a competition there in 1485. Most modern editions and productions of the English play have been based on the Britwell Court copy of an edition printed by John Skot c.1530, but the earliest text is actually the British Library fragment printed by Richard Pynson in or before 1528. Our performance text was therefore based mainly on this authority except for the first 305 lines, which have been lost and for which the Britwell Court copy represents the earlier of two surviving editions, both printed by Skot. Despite the efforts of some commentators to date Everyman to the last years of the fifteenth century, there is in fact no evidence that it existed before the 1520s, and it seems best regarded as an anti-reformation play designed for performance in a secular setting. Our production therefore used costumes and music that might have been seen and heard in 1528, the year of Pynson’s retirement. The prologue to The Summoning of Everyman, which has no counterpart in the Dutch play, suggests that the English translator saw a distinction between the play’s moral ‘entent’ and its literal ‘mater’ or story material, and our production set out to investigate this distinction. Perhaps surprisingly, the story material includes a good deal of comedy, and it also draws heavily on an allegorical romance about an unjust steward, which the Dutch author may have derived from the Gesta Romanorum. The paper goes on to consider how the structure and meaning of the play might have been made clear in performance, particularly through the interpretation of the character of Death and the strategic doubling of parts.
120
John McKinnell
1. Introduction There is of course no one ‘right’ way of producing any play, and I shall not try to argue for one in this paper. What I do want to do is to present some thoughts about producing The Summoning of Everyman which result from the production of it by Durham Medieval Theatre Company in June 1999. Our aim in this production was to recreate as much as possible of the effect that the play might have had for its first audiences, whether in actual performance or in the contemporary imagination. Of course we had to use modern pronunciation to enable the sense of the words to reach our audiences, and for reasons of cost we could not avoid using some modern materials for the costumes. We must also accept that we are performing before modern audiences, who will inevitably have many attitudes and reactions that are different from those of the sixteenth century. Nonetheless, this still seems a valuable way of performing any medieval play, not least because it respects the integrity of the original and does not try to impose modern meanings or opinions on it. Let me give an example: in the very lively RSC production of 1997, the speech of Five Wits in praise of the sacramental powers of priesthood was distorted into a satire on the arrogance and fraudulence of all priests of all religions, including not only a Roman Catholic, but also a Scottish Presbyterian, a ranting rabbi and a fundamentalist Muslim imam. Effects like that may be entertaining, especially to modern atheists, but they misrepresent the play and fail to respect it as a work of its own time. But if one sets out to perform Everyman in the context of its own genre and era, one must begin by deciding what those are. It is now widely accepted that Everyman is an adapted translation of the Dutch play Elckerlyc by Peter van Diest, which won first prize in a rhetoricians’ contest in Antwerp in 1485 (Mills 1995: 128). The ‘chambers’ of rhetoricians in the cities of the Low Countries were in effect amateur dramatic societies, which competed each year in contests organised and judged by their city councils. Despite the heavily religious content of the play, therefore, the Dutch original was certainly written as civic drama, to be performed in a secular setting, most likely a civic hall; and unless there is evidence to the contrary, we must imagine a similar setting for Everyman.
How might Everyman have been performed?
121
2. Base Text and Date The date of Everyman is a matter of some debate. The text survives in single copies of four different printed editions, all produced in London. These are: 1. Richard Pynson, ‘Imprynted at London in Fletestrete by me Rycharde Pynson, prynter to the kynges moost noble grace’ (15081528), British Library fragment (l. 305 to the end). Title page lost, no closing woodcut. Greg calls this D, but I shall refer to it as BL.1 2. Richard Pynson, ‘Imprynted at London in Fletestrete at the Sygne of the George by Rycharde Pynson prynter vnto the Kynges noble grace’ (1508-1528), Bodleian Library Douce fragment (l. 683 to the end, with some lines lost on damaged pages). Title page lost, no closing woodcut. Greg calls this C.2 3. John Skot (1521-1537): ‘Jmprynted at London in Poules chyrche yarde by me John Skot +’ - the ‘Britwell Court’ complete copy, now in the Huntington Library, San Marino, California. Title page with composite woodcut: left, an elegant youth, with a blank scroll above him; right, death in a graveyard, holding a coffin-lid; packing to left (including a scroll with the irrelevant ‘B.M. aue’) and below. Closing woodcut device of John Skot, with French motto and name in mirror-writing. Greg calls this A, but I shall refer to it as BC.3 4. John Skot (1521-1537), no closing colophon: ‘Huth’ complete copy, now in the British Library. Title page with woodcut as in 3, but the scroll above the youth is labelled ‘Eueryman.’; left-hand packing simpler, with no inscription, and no packing below. Title page verso, composite woodcut with six characters from the play, labelled ‘Felawshyp’, ‘Eueryman.’, ‘Beauty’, ‘Dyscrecyon.’, ‘strengthe.’, ‘kynne.’ Closing woodcut in baroque style with arms/device of IOHN SCOTT. Greg calls this B.4 The relationship between the four editions has also been a matter of dispute. Pynson was active 1493-1528, but only became printer to the king in 1508; Skot was active 1521-1537. Greg seems to have labelled the editions from A to D in his order of preference; he 1
For a facsimile of this edition, see Everyman (1912). Ed. Greg (1910), with a facsimile of the last side. 3 Ed. Greg (1904), with facsimiles of woodcuts. 4 Ed. Greg (1909), with facsimiles of woodcuts; facsimile of the whole edition in Everyman (1912). 2
122
John McKinnell
cannot arrive at a traditional stemma, but thinks that the Britwell Court copy is the most accurately representative of the work of the translator.5 Relying on his work, most subsequent editions of the play (Adams, Bevington, Cawley) have taken the Britwell Court copy as their base text. All four copies have errors and variants of their own, and the relationship between them can only be studied fully for that part of the text for which all four survive (l. 683 to the end). Such a comparison suggests to me a continuing process of modernisation combined with some correction and some corruption, and that the chronological order of the editions is probably: BL - Douce - Britwell Court - Huth. For this section of the play, the most significant variants fall into four categories:6 2.1. Modernisations a)
b)
c)
5 6
In four cases, Douce modernises BL and is followed by BC and Huth, e.g.: 699: BL almesse; Douce, BC, Huth almes; 825: BL Go trusse the into the grounde; Douce Go thirste the in to the grounde; BC and Huth Go thryst the into the grounde). In nine cases, both Skot’s editions modernise both Pynson’s, e.g.: 752: BL, Douce gaue he vs; BC, Huth he gaue; 797: BL, Douce by my fay; BC, Huth by my fayth; 813: BL, Douce fro the; BC, Huth from the. There are five cases where only Huth modernises, and two where it regularises the grammar, e.g.: 684: BL, Douce, BC by you stande; Huth stonde by you; 687: BL, Douce, for swete ne for soure; BC for swete ne soure; Huth for swete nor soure; 734: BL, Douce, BC fro synne; Huth from synne; 870: BL, Douce, BC All erthely thinges is but vanite; Huth all ertly thynge is but vanyte.
For his detailed argument, see Greg (1910: 60-69). In the following lists, insignificant variations of spelling are ignored.
How might Everyman have been performed?
123
2.2. Corrections a)
b)
c)
d)
7
In six cases, Douce corrects BL’s errors and is followed by Skot’s two editions, e.g.: 763: BL These be without synne made blynde; Douce, BC, Huth These be with synne made blynde; 827-8: BL But I se well he that trusteth in his strength / Is greatly disceyued at the length; Douce, BC, Huth He that trusteth in his strength / She hym deceyuth at the length (compare Elckerlijc 797-798: Wie wil hem verlaten op zijn cracht / Si vliet alst mist doet wter gracht, which also has the feminine personification of Strength).7 838: BL Loke in my graue and thou shalt se; Douce, BC, Huth Loke in my graue ones petyously (BL’s text here is dramatically effective, but Elckerlyc 809 dat bid ick u ootmoedich shows that the compositor of BL has made a mistake); In six cases, obvious errors introduced in Douce are recorrected in both Skot’s editions, e.g.: 691: Douce aduysemente and lyberacion; other three editions aduysement and delyberacion; 848: Douce I take you for my best frende; other three editions I toke you.... In six cases, both Skot’s editions correct errors in both Pynson’s, e.g.: 754: BL, Douce He helde them nat to that lorde omnipotent; BC, Huth He solde them not to vs that lorde omnypotent (referring to the sacraments); 880: BL, Douce Into thy handes lordes; BC, Huth In to thy handes lorde. In six cases, Huth re-corrects obvious errors introduced in BC, e.g.: 789: BC My lymmes vnder me doth folde; other three editions do folde; 702: BL, Douce I it bequethe to be retourned (reflecting Elckerlijc 671: Ghenick daer si schuldich is te gaen ‘I bequeath it to where it ought to go’); BC In queth (‘in
For a text of Elckerlyc and commentary on it, see Willems (1934).
124
John McKinnell
bequest’) to be retourned; Huth misunderstands this and ‘corrects’ to In quyet to be returned. 2.3. Retention of errors or variants a)
b)
In seven cases, both Skot’s editions retain errors introduced by Douce, and in a further twelve the same is true of casual variants, e.g.: 705: BL Euer after this day; Douce, BC, Huth share the error Euer after and this daye; 821: BL yet promise is dett / this ye well wot; Douce, BC, Huth share the less effective variant Wyll ye breke promyse that is dette. In four cases, both Skot’s editions share casual variants, e.g.: 745: BL, Douce No remedy may we fynde; BC, Huth No remedy we fynde; 778: BL, Douce his hande; BC, Huth youre hande.
2.4. Deliberate doctrinal corrections At three points in Five Wits’s speech on the sacramental power of priests (730-749), Skot seems to have introduced variants in order to emphasise Catholic beliefs about the sacrament: a)
b)
c)
738: BL, Douce Goddes body in flesshe and blode to take; BC, Huth ...to make. Skot’s text is closer to Elckerlyc 706 maecken, and may come from a copy MS that was still available to him, but it also emphasises the effective change wrought by the priest according to the doctrine of the real presence. Pynson’s text merely reflects the medieval practice of the priest receiving the sacrament on behalf of the people. 748: BL, Douce And letteth them (i.e. priests) in his stede amonge vs be; BC, Huth And setteth them in his stede amonge vs to be. Skot’s alteration emphasises that priests are appointed by God, and by implication stresses the doctrine of the apostolic succession. 732: BL, Douce, BC For preesthode excedeth all other thynge; Huth For good pryesthod excedeth.... Huth adds that
How might Everyman have been performed?
125
the priest must be good, perhaps to counter Protestant denigration of priests as corrupt. The overall conclusion to be drawn from this evidence can be summarised as follows: 1.
2.
3.
Pynson’s editions probably preceded Skot’s, in the sequence: BL - Douce - BC - Huth. Any modern edition which seeks to represent the translator’s work as closely as possible should therefore be based mainly on BC until line 305, and thereafter mainly on BL, though we must also be aware that the compositor of BL made many errors. It follows that we now need a new critical edition of the play. Each edition relies mainly on its immediate predecessor, but both printers must also have had access to a copy manuscript (see for example the readings of Douce at 827-828 and BC at 738). However, they did not always consult this manuscript, and both of them were inclined to go astray when the preceding edition had a corruption whose correction was not obvious. There is a gradual process of modernisation and correction throughout the sequence; this suggests co-operation between Pynson and Skot rather than rivalry.
The likeliest scenario is probably that Skot took over the title when Pynson retired, and if this is in fact what happened, Skot’s two editions must be dated to 1528 or later. Pynson’s editions may date from any time between 1508 and 1528, but the title seems to have sold vigorously and new editions may have been produced quite frequently. In that case, all four editions may have appeared within a few years of each other. There is certainly no evidence that the English Everyman existed before the 1520s, when the groundswell of Protestantism was already beginning; read in that context, this is an anti-reformation play. We therefore decided to set the notional date of our production in 1528, the year of Pynson’s retirement.8 8
For this argument, see Mills (1995) and Proudfoot (1982). The earlier dating adopted by Adams, Bevington, Cawley and Wickham, at least a decade earlier than the earliest possible date of Pynson’s editions, sits oddly with their preference for
126
John McKinnell
3. Performance and Theme There is, however, some doubt about whether Everyman was ever physically performed in English in the early sixteenth century. The translator adds a fifteen-line prologue which has no counterpart in the Dutch original and which describes The Summoning of Everyman as: “By fygure a morall playe” (3). It is sometimes argued, despite the fact that we are urged in line 2 to “here this mater with reuerence” (my italics), that this means that the translator was presenting his work as only ‘figuratively’ a play, but actually as a moral tract designed for private devotional reading. This is possible, though one might equally understand by fygure to mean ‘in the appearance or form of (a moral play)’. However, there is no need to agonise over this question. The evident popularity of the printed text does suggest that the majority of those who bought it did so in order to read it privately –but then, the same must have been true of those who bought the quarto editions of Shakespeare’s plays. Whether Everyman was actually performed in English or not, the writer of the prologue invites readers to visualise a production, and does not make actual production seem at all unlikely. Perhaps more interesting is the question of what the prologue means by the distinction drawn in ll. 7-8: ‘This mater is wonders precyous But the entent of it is more gracyous.’
The entent is fairly clearly the moral meaning, that is the need for every human being to repent and amend their life before having to answer for their deeds before God. But how, then, does this differ from the mater –the content or story material– which seems at a first glance to be exactly the same thing?9 Bearing in mind the fact that the only context we have for the play is a secular one, our production set out to test the idea that the mater is a superficially secular story. This story includes not only the motif of the summoning of a steward to render his accounts,10 but also that of the constant girl (Good Deeds) Skot’s editions and seems to be based only on the assumption that Everyman must date from approximately the same period as Elckerlyc. 9 Mills (1995: 130) suggests that the mater is the play’s visual impact as opposed to its moral import, but that would involve taking the word in a very unusual sense. 10 For an analysis of this, see Kolve (1972).
How might Everyman have been performed?
127
who eventually gains the love of a heedless young man (Everyman), of which the best-known example today is the plot of All’s Well that Ends Well.11 And if it is a secular story, the unremitting atmosphere of religious solemnity which has surrounded most modern productions of Everyman needs at least to be questioned. It was not until after our production was over that I discovered that these two story motifs appear together, along with a moralisation which makes the protagonist into ‘every man’, in the mid-fifteenthcentury English version of Gesta Romanorum story 36, whose Latin original probably comes from early fourteenth-century England. This story runs as follows: An Emperor (who is moralised as God) goes on a pilgrimage, leaving his daughter (the Soul), five knights (the Five Wits), their teacher (Reason) and his greyhound (the Flesh) in the charge of his steward (Every Man). The steward rapes the daughter and then drives her out, so that she is forced to become a beggar; then he expels the knights and confiscates their property, kills the teacher and fails to prevent the greyhound from escaping. Hearing that the Emperor is about to return, the steward is struck with remorse and goes to meet him, barefoot, wearing only shirt and breeches, and carrying three cords with him. He tells the Emperor that he deserves to be bound with the first cord, dragged behind a horse with the second, and hanged with the third. The Emperor promises to spare his life if he will confess what he has done, and he does so. He then finds and secures the greyhound, restores the knights to their property, “and with grete worshipe weddid the mayde... and endid fair” (ed. Herrtage 1962: 139-147). The resemblances between this story and Everyman are obvious: Everyman is certainly portrayed as a steward who has reason to be afraid when asked to present his accounts, and he has wronged a virtuous female character (Good Deeds) whom he must restore to health. The character of Five Wits recalls the five knights; Reason in the Gesta Romanorum fills roughly the same slot as Knowledge in Everyman; and the penitential use of three cords in the story could have suggested the scourge wielded by Everyman to punish his own flesh. It seems at least possible that some version of the Gesta 11
This itself is ultimately derived from Boccaccio, Decameron, Day 3, story 9, ed. Marti (I, 247-255), trans. Rigg (I, 241-249).
128
John McKinnell
Romanorum story might have provided Peter van Diest with the idea for his play. But even if this was not the case, the popularity of the Gesta Romanorum in England suggests that the idea of the surface story of Everyman as the romance of the unjust steward would not have been too unfamiliar to early audiences. 4. God and Death If Peter van Diest did start from a story like the one in the Gesta Romanorum it is hardly surprising that both Elckerlyc and the main action of Everyman begin with a furious diatribe by God against the sinfulness of man. In representing this on the stage, there is a danger that God may appear to be a cantankerous old tyrant, and we, like the RSC production, felt that some sort of mime of Everyman in sin was needed at the opening of the play. There is nothing unlikely about mimed prologues in early drama, as we can see from plays like Gorboduc12 and the play within a play in Hamlet (Act III Sc. II, ed. Munro 1958: V, 471) –but there is little point in showing an irrelevant group of attractive young people simply having a slightly silly party, as the RSC did. We therefore decided: 1. That the mime has to introduce Everyman in the company of the people he naturally turns to first –Fellowship, Kindred, Cousin and Goods (to whom we added Beauty, in the role of Kindred’s wellmeaning but intellectually challenged serving maid). This is very likely to have been done in the sixteenth century, when performance in a secular hall would often make it difficult to conceal the rest of the cast at the opening of the play. 2. That both he and they must alienate the audience with some convincingly nasty behaviour. We made them burst in during the prologue, interrupting it with a crude rendering of Skelton’s anti-Dutch song Jolly Rutterkin (c.1495; ed. Henderson 1959: 25) in the early-sixteenth-century setting by William Cornish. God’s speech gives an indication of the sins they are guilty of: treating pride, covetousness, anger and lechery as if they were virtues, and piling up wealth without regard to the need for charity towards others. We were particularly conscious of the last 12
For the text of Gorboduc, which has a mimed prologue to each act, see Adams (1952: 503-535).
How might Everyman have been performed?
129
of these, since our heads were still full of the words of our own 1998 production of the York Doomsday, where the bad souls are told, in accordance with standard Christian teaching, that when they failed to show charity to the needy: ‘To me ye did þe selue and þe same’ (Doomsday 364; The York Plays, ed. Beadle 1982: 415)
This also provides a neat answer to the question of how to introduce God: he effectively entered by throwing off the cloak of a beggar, in which guise Everyman had rejected him as he came in. An old female beggar also made a plea for alms from Everyman, who lost his temper at the repeated request and kicked her to the ground –thus placing Good Deeds, who when she first speaks is lying on the ground unable to walk. If something like this was not done, her later entry would pose a real problem, since she could not walk on without contradicting her own statements that she cannot move or stand (see ll. 488, 498). But if the audience have already seen Everyman kick her to the ground, they know that she is telling the truth when she says that it is his sins that have incapacitated her (486-488). Attention then shifted away from her to the revellers, and the beggar who was to become Good Deeds seemed like no more than a pile of rags in a corner. The audience became more interested in the attempts of Everyman and Fellowship to buy the sexual services of Goods, and in Kindred’s comic bullying of Cousin –and in being so easily diverted from Everyman’s brutality, they became accomplices in his heedlessness. This, too, is essential to the play’s meaning –after all Everyman must include every member of the audience, and his sins are theirs. God’s summons is delivered to Everyman by Death, and it gave our audience a suitably uncomfortable feeling that his first words in response to God’s summons came from behind them. The main reason for bringing Death in through the audience is to bring out the fact that we are invited to see the sinful Everyman through his eyes: ‘Loo yonder I se Euery man walkynge Full lytell he thynketh on my comynge His mynde is on flesshely lustes and his treasure.’ (Everyman 74-82, Britwell Court copy)
130
John McKinnell
The same technique can also be seen as Death points out Herod in the N-Town Death of Herod: ‘I am Deth, Goddys masangere. Allmyghty God hath sent me here Yon lordeyn to sle, withoutyn dwere, For his wykkyd werkynge...’ (177-180; ed. Spector 1991: I, 193)
We found it an effective way of making mankind look very small and vulnerable. However it does not follow that Death in this play should be presented as a bogeyman. True, many medieval sources do present him as the tyrant with the spear, as we can see from the image of him menacing great lords in Dunbar’s Timor Mortis Conturbat Me (c.1507; ed. Kinsley 1979: 178-181): He sparis no lord for his piscence, Na clerk for his intelligence; His awfull strak may no man fle: Timor mortis conturbat me. (Timor Mortis Conturbat Me 33-36)
Death needs a spear in this play as well (though in practice it often seemed more like a staff of office than a weapon). But Death can equally be the confidential advisor who comes to whisper a message in your ear or embraces you like an old friend, as he does to the poets Robert Henryson and Sir John Ross in Dunbar’s poem: In Dunfermelyne he has done roune With Maister Robert Henrisoun; Schir Johne the Ros enbrast has he: Timor mortis conturbat me. (Timor Mortis Conturbat Me 81-84)
A third approach was used by the RSC: Death appeared as an attractive young woman whom Everyman tried to seduce until he suddenly learned her name. This was quite effective in its own terms (perhaps because of the highly charged performance of Josette Simon as Death), but it seems to me to involve images drawn from
How might Everyman have been performed?
131
nineteenth-century romanticism, such as Keats’s La Belle Dame Sans Merci. The sources of that are probably also medieval, but they come from the tradition of the fairy ballad, and have nothing to do with Christian personifications of death. The scene between Death and Everyman can also show how eloquent a small number of significant props can be in a morality play. In our production Everyman had two books of accounts –a large, splendidly bound volume detailing his worldly wealth, with part of which he tries to bribe death to go away (Everyman 121-123), and a small shabby black one of his blotted and neglected spiritual accounts. On Everyman’s table we placed a graduated candle, to reflect his constant awareness that time is slipping away. After Death’s exit he says: “The day passeth / and is almost ago” (194). Before his appeal to Goods he reflects: “I lose my tyme here longer to abyde” (386). Confession reminds him that “your tyme draweth fast” (569); and shortly before his death, he says: “For I se my tyme is nye spente awaye” (866). This became a recognised device in later sixteenthcentury drama, as we can see from Faustus’s: “Ah Faustus / Now hast thou but one bare hower to liue” (Dr. Faustus 1419-1420; ed. Brooke 1969: 192), and Macbeth’s “Out, out brief candle” (Macbeth Act V, Sc. VI, 23; ed. Munro 1958: VI, 1194). Here, it is a recurring theme, and it seemed helpful to have a visual reminder of it; we originally thought of an hourglass, but couldn’t find a suitable one –and in fact it was just as well, for in the end the candle was a better symbol. Just as Death should take the audience by surprise when he enters, we felt that he should also exit in a way unlike anyone else – after all, Everyman’s worldly associates are all typical, semi-literal characters, while Death is an allegorical, universal figure. So we made him prepare to leave by turning his back on Everyman with his arms outstretched. Desperate to delay him, Everyman ran and seized him by the cloak, which enabled Death to slip out of the cloak and exit under the stage without being seen by the audience, so that Everyman was left holding only the empty cloak. Observant members of any audience could of course see how this trick was done, but it still worked, because the sudden disappearance of Death left attention focused on the empty cloak. This effectively symbolises the fact that Death is an intangible absence, a fact which it is absurd to try to bribe, placate or negotiate with. The allegorical character of Death has
132
John McKinnell
disappeared, leaving Everyman to face the literal prospect of his own death. 5. Structure and Doubling The structure of Everyman falls into three balanced parts: 1. God’s summons to Everyman is delivered by Death; he appeals to the worldly figures to go with him, but is rejected by all of them. This section is summarised in Everyman’s soliloquy at the end of it (463-478). 2. Everyman repents, on the advice of Good Deeds and Knowledge, and his penance restores the disabled Good Deeds to health. 3. Everyman summons his bodily qualities; they support him at first but then abandon him; only Good Deeds will go with him, while Knowledge will accompany him as far as the edge of the grave. The beginning of this section is marked by Everyman’s formal summons to the bodily qualities (668-669), the only point in the play at which several characters enter at once. There seems to be a deliberate patterning between the worldly figures in part 1 and the bodily qualities in part 3, except that they abandon him in reverse order: Part 1: 1. 2, 3. 4.
Fellowship Kindred and Cousin Goods
Part 3: 4. 2, 3. 1.
Five Wits Strength and Discretion Beauty
Fellowship is another sinful man like Everyman himself, and Five Wits represents his own bodily consciousness. Both of them begin by assuring him that they will not forsake him (213-214, 686687), and when they do abandon him, they do so in closely similar words, both using the rhyme forsake : take, both associating farewell and end in the same line, and with Everyman unsuccessfully invoking a special relationship with both of them (cf. 285: “We haue loued longe and now I need”, with 848: “For I toke you for my best frende”). The case for bringing out this symmetry by doubling the parts of Fellowship and Five Wits seems very strong. Fellowship is followed by the comic double-act of Kindred and Cousin. When the play is liberated from the usual modern
How might Everyman have been performed?
133
atmosphere of churchy seriousness, this can be played for music-hall laughs, with its absurd excuses (Cousin suddenly develops a cramp in his toe, and Kindred offers to lend Everyman her flighty maidservant). The RSC also adopted this approach, and it was for me one of the most successful parts of their production. Kindred is the obviously dominant member of this couple, but the reference to my mayde (360) suggests that this may be a female character, whereas Cousin’s opening promise that they will live and die together (324) suggests pretensions to military valour. The bodily qualities in Part 3 are not kept apart to the same extent, but there is again a suggestion that Strength and Discretion are a couple. Discretion always follows Strength (835-836), who makes references to military valour (e.g. 685: “Though you wolde in batayle fight on the grounde”) and seems obviously male, while the text gives no indication of the gender of Discretion. Both couples begin by assuring Everyman that they will go with him (cf. 322-324 with 675-677), but their desertions of him are more interesting. Kindred and Strength employ similar evasive strategies, effectively accusing Everyman of emotional blackmail: Kindred says that it’s no good trying to entice her (359) and Strength gets angry and declares that he will leave “Though thou wepe tyll thy hert brast” (814: notice his switch from the respectful you to the dismissive thou). Cousin and Discretion have better reasons for leaving Everyman: Cousin has his own unprepared reckoning to think of (375-376), while Discretion must dutifully follow Strength (a detail which suggests an obedient wife, 835-836). This suggests that the actors who played Kindred and Cousin may also have represented Strength and Discretion, but that there could be a gender reversal when they did so. We chose to represent the two couples with a very large man playing Kindred (as a woman) and Strength (as a man) and a very small boyish woman playing Cousin (as a man) and Discretion (as a woman). The folk-play figure of the ‘Betty’, the earth-mother played by a large male actor, certainly existed as early as 1433, when the Bursar’s account for the Priory of Durham includes a payment to a male minstrel called ‘Moder Nakett’. There are also several records of minstrels who performed with boy assistants (who were probably used for most female roles), so there is nothing historically unlikely about this pairing.
134
John McKinnell
If these doublings are accepted, the question must arise whether one should also double the roles of Goods and Beauty, the remaining figures in each category. Both begin by asking Everyman what he wants (398, 671), and interestingly, each of their scenes contains suggestions that we would expect to find associated with the other. Everyman addresses Goods in terms that suggest the protestations of a courtly lover (457-458), while Beauty says that she will not go with Everyman: “Nat & thou wolde gyue me all the golde in thy chest” (804). At the same time, one must recognise that not all textual echoing can be a signal for doubling: there is a very clear echo in the scornful rejection of Everyman by Goods of Death’s earlier warning that life and property are merely lent, not given: Goods:
What, wenest thou that I am thyne?
Everyman: I had went so. Goods: Nay Euery man I say no. As for a while I was lent the, ......... For whan thou art deed this is my gyse Another to desceyue in the same wyse
.
As I haue do the / and all to his soules reprefe. (Everyman 437-440, 448-450, BL text) Death: What, wenest thou thy lyue is gyuen the And thy worldely gooddes also? Everyman: I had wende so veryle Death: Nay nay it was but lende the, for as soone as thou arte go Another a whyle shall haue it and than go ther fro Euen as thou hast done. (Everyman 161-167, BC text)
In this case the effect of Goods’ echo of the words we have already heard from Death must be to suggest that devotion to material property brings spiritual death with it. Beauty must be female (she ‘takes her tappe –her spinning flax– in her lap’ as she goes: 802), but Goods is once referred to as he (475). Nonetheless, we decided to play Goods as female, for a number of reasons. Firstly, while references to characters as female in morality
How might Everyman have been performed?
135
plays have to be accepted, male pronouns need not be taken so seriously, since they sometimes break through the allegorical character and refer to the male actor who played the role. An example can be seen in Henry Medwall’s Nature (c.1500), where the nurse Innocency is dismissed by Sensuality with the words “He ys but a boy” (Nature I, 657) although the character is certainly female and is referred to elsewhere as fayre woman and syster (Nature I, 442, 505; ed. Nelson 1980: 107, 102, 104). There are a number of reasons why it is effective to portray Goods as female. It enables the audience to see Everyman as Death sees him, with his mind on fleshly lusts and his treasure (both at once, if he is talking to an expensive prostitute). It also avoids the problems that are usually caused by the awkwardly literal opening of Goods’ speech, which refers to being tied up in bags and locked in chests. Any attempt to portray this literally is likely to make Goods resemble an absurdly large and ill-shaped parcel, and to remove any sense that we too might be tempted by Goods as Everyman has been. We were able to play this opening speech as provocatively sensual laziness, with undertones of bondage. Later in the scene between Goods and Everyman, having a female Goods enabled us to bring out faint echoes of the language of fin amour, and then to provide an interesting contrast between her and a virtuous woman –Good Deeds– whose first scene immediately follows. It contributes to the ironic ambiguity of what ‘good’ is in the play if these two characters are seen as morally opposite partners for Everyman. In the end, we did not double Goods with Beauty, but this was for a purely practical modern reason –since our director was male, we needed to have an understudy for the female roles in case any of the women in the cast fell ill. This problem would not, of course, have affected an all-male sixteenth-century cast. 6. The Romance of Good Deeds One aspect of the second part of the play would probably have worried early audiences far less than it does modern ones, namely Everyman’s penitential scourging of himself on the instructions of Confession (611-628). Bodily penance was a normal part of medieval religious devotion, but to a modern audience it is liable to seem cruel of Confession, and either pointless or masochistic for Everyman.
136
John McKinnell
There was certainly an element of shock for our audiences when Everyman stripped his back in preparation for it. It is, however, central to the play’s meaning, besides providing the practical context in which Everyman takes off his fine doublet in preparation for putting on the garment of contrition twenty lines later (648-650). There was no way of dodging it, and it wouldn’t do, either, for Everyman to whip himself without taking his shirt off. For a sixteenthcentury audience, this would probably not have been a problem. In any case, the narrative point of the penance is to be found in the romance element of the story –it magically restores Good Deeds to health, and redeems Everyman’s past cruelty by showing that he is now prepared to endure pain for her sake. At this point she can also be the old woman made young again, renewed like the heroine of the Wife of Bath’s Tale (Canterbury Tales Fragment D, ll. 1236-1256; ed. Benson et al. 1988: 121), and able to throw off the rags of the old beggar woman and become Everyman’s love. Naturally, she rushes to prevent him from suffering further (and unnecessary) pain. While the bodily qualities are engaging Everyman’s attention in the first half of Part 3, Good Deeds tends to slip into the background again, and it is natural for them to push her aside as they crowd round him (and for her not to resist this). It is not until they have abandoned him that Everyman turns back to her. But if this has become the romance of Everyman and Good Deeds, as I have suggested and as the Gesta Romanorum story seems to confirm, it makes sense that it should end with a wedding. It follows naturally that Everyman is finally ‘married’ to his Good Deeds at the moment of death, and by Death as priest. Many modern productions of Everyman are extremely solemn, even depressing, at the end, but I think this is a misreading. It is Knowledge, representing authoritative truth, who hears the angels singing for the reception of Everyman’s soul, and Everyman himself must be physically resurrected in order to be addressed in the Angel’s first line: “Cume excellent electe spouse to Iesu” (894). It seems odd, in the last thirty lines of the play, to introduce a new character who has no previous connection with Everyman –so who is this angel? In the almost contemporary Digby Mary Magdalen the heroine has both a good and a bad angel who seem to be personal to her, and the Good Angel rejoices when Mary’s repentance is accepted by Jesus (Mary
How might Everyman have been performed?
137
Magdalen 705-721).13 The role of this Good Angel seems very similar to that of Good Deeds in Everyman, and we decided to treat the Angel as the continuing effect of Good Deeds beyond the grave. This introduced a delightful ambiguity into the Angel’s first line. Its probable primary meaning is: ‘Come, you excellent, chosen bride of Jesus’ (referring to the Soul). But since Everyman’s soul is standing before us in the male form of Everyman himself, may there not also be a fulfilment of the romance here, in the sense ‘Come to Jesus, (my) excellent chosen husband’? At all events, on stage that was the way in which it seemed to work, underlining the fact that the end of this play, though sober in its warning, is predominantly triumphant. This is the story of a man whose Lord and spouse are in the end both well pleased with him. Bibliography Primary sources Everyman. 1910. (repr. Walter W. Greg). Materialen zur Kunde des älteren Englischen Dramas 28. Louvain: Uystpruyst. (Editions of Pynson’s two fragmentary eds., with brief introduction, facsimile of the last side of the Douce fragment, and critical apparatus including lists of variants.) Everyman. 1912. Tudor Facsimile Texts 28. London: Tudor Facsimile Texts (facsimiles of the Huth copy and the BL fragment). Everyman. 1932. (repr. W. W. Greg). Materialen zur Kunde des älteren Englischen Dramas 24. Louvain: Uystpruyst. (Original ed. 1909, corrected repr. 1932.) (Edition of the Huth copy, with brief introduction and facsimiles of woodcuts.) Everyman. 1963. (repr. W. W. Greg). Materialen zur Kunde des älteren Englischen Dramas 4. (Original ed. Louvain: Uystpruyst, 1904, repr. Vaduz, 1963.) (Edition of the Britwell Court copy, with brief introduction and facsimiles of woodcuts.) Secondary sources Adams, Joseph Q. (ed.). 1952. Chief Pre-Shakespearean Dramas. Cambridge (Massachusetts): Riverside. (Original ed. 1924. Boston: Houghton Mifflin.) (For Everyman, see pp. 288-303; Adams dates the play to ‘before the close of the fifteenth century’.) Baker, Donald C., John L. Murphy and Louis B. Hall (eds). 1982. The Late Medieval Religious Plays of Bodleian MSS Digby 133 and E Museo 160. Early English Text Society 283. Oxford: OUP. 13
Ed. Baker, Murphy and Hall (1982: 47-48). At the end of Mary Magdalen (ll. 2115-2126, ed. p. 94), Mary’s soul is received into heaven in a way that seems very similar to the end of Everyman.
138
John McKinnell
Beadle, Richard (ed.). 1982. The York Plays. York Medieval Texts, 2nd Series. London: Edward Arnold. Bevington, David (ed.). 1975. Medieval Drama. Boston: Houghton Mifflin. (For Everyman, see pp. 939-963; Bevington dates the play to c.1495.) Boccaccio, Giovanni. 1990. Il Decamerone (ed. Mario Marti). 8th ed. (2 vols). Milano: Rizzoli. ___. s.d. The Decameron of Giovanni Boccaccio (trans. J. M. Rigg; 2 vols). London: Navarre Society. Cawley, Arthur C. (ed.). 1993. Everyman and Medieval Miracle Plays. London: Dent. (Original ed. 1956; repr. with preface and bibliography by Anne Rooney.) (For Everyman, see pp. 195-225; estimates that ‘Everyman was probably written before the end of the fifteenth century’.) Chaucer, Geoffrey. 1988. The Riverside Chaucer (ed. Larry D. Benson et al.). 3rd ed. Oxford: OUP. Dunbar, William. 1979. The Poems of William Dunbar (ed. James Kinsley). Oxford: OUP. Herrtage, Sidney J. H. (ed.). 1962. The Early English Versions of the Gesta Romanorum. Early English Text Society, Extra Series 33. London: OUP. (Original ed. 1879.) Kolve, V. A. 1972. ‘Everyman and the Parable of the Talents’ in Taylor, Jerome and Alan H. Nelson (eds) Medieval English Drama, Essays Critical and Contextual. Chicago: Chicago University Press: 316-340 (on Everyman as a steward who must present his accounts). Marlowe, Christopher. 1969. The Works of Christopher Marlowe (ed. C.F. Tucker Brooke). Oxford: OUP. (Original ed. 1910, reprinted 1969.) Medwall, Henry. 1980. The Plays of Henry Medwall (ed. Alan H. Nelson). Tudor Interludes Series. Cambridge: D. S. Brewer. Mills, David. 1995. ‘The Theaters of Everyman’ in Alford, John A. (ed.) From Page to Performance. Essays in Early English Drama. East Lansing: Michigan State University Press: 127-149 (sets Everyman in the context of the 1520s). Proudfoot, Richard. 1983. ‘The Virtue of Perseverance’ in Neuss, Paula (ed.) Aspects of Early English Drama. Cambridge: D. S. Brewer: 92-109 (places Everyman in the 1520s). Shakespeare, William. 1958. The London Shakespeare (ed. John Munro; 6 vols). London: Eyre and Spottiswoode. Skelton, John. 1959. The Complete Poems of John Skelton, Laureate (ed. Philip Henderson). London: Dent. Spector, Stephen (ed.). 1991. The N-Town Play. Early English Text Society, Special Series 11-12 (2 vols). Oxford: OUP. Wickham, Glynne. 1987. The Medieval Theatre. 3rd ed. Cambridge: CUP (places Everyman in ‘the last decade of the fifteenth century’). Willems, Leonard. 1934. Elckerlyc-Studien. ’s Gravenhage: M. Nyhoff (includes a text of Elckerlyc: 120-158).
Shift of meaning in the animal field: Some cases of narrowing and widening Isabel de la Cruz Cabanillas University of Alcalá
Abstract This article explores the diachronic transition of some animal denominations from the OE period onwards. The quantity and extent of the shift in meaning are remarkable at times: words which originally denoted the species were restricted to indicate the male or the female animal; while other terms which had been hyponyms turned into hyperonyms with the passsage of time. In the course of these changes several lexical units disappeared from the system, to be replaced by others. This study discusses some of the reasons traditionally adduced to explain semantic change, as well as the factors which may account for word loss and other related phenomena, such as the adoption of borrowings, the restriction of meaning of some specific items and the widening of others.
1. Introduction Linguistic change is a universal characteristic of living human languages. It is commonly agreed that lexis and semantics reflect changes quicker than any other level of the language system, so the most obvious shifts can be seen in the usage and meaning of words. In spite of this, work on semantic change has managed to shed little light on why the changes occur. Very often attempts to explain the phenomenon are not particularly successful and end up as little more than classifications of the different processes. Explanations about semantic change are sometimes even contradictory. While some scholars contend that there is no pattern (Ullmann 1957: 154; Sihler 2000: 94), others concentrate on the tendencies observed in some semantic shifts in order to work out a framework within which to give
140
Isabel de la Cruz Cabanillas
account of the general directionality of semantic change (Campbell 1998: 267-273; Traugott 1989: 34-35, 1999: 93; Traugott and Dasher 2002: passim). Although there seems to be no agreed solution in relation to the problem of what motivates semantic change, most linguists tend to group the phenomena of change according to the elements involved in the process. Thus, shifts in denotation include widening and narrowing, and shifts in connotation refer to pejoration, amelioration and taboo replacement; changes involving conceptual association comprise metaphor and metonymy (encompassing synecdoche also). Folk etymology could also be listed here, based on a connection occurring between two words, usually in terms of sound. However, this can only be classified as semantic shift when the similarity between the words has semantic consequences. Some processes seem to draw scholars’ attention more than others. In my case, I have always been struck by the idea that both widening and narrowing seem to be more frequent in the field of animal denominations than in others. In order to test the validity of this hypothesis, I have focused my research principally –though not exclusively– on domestic animals, omitting those dealt with in previous studies (cf. for example, ‘References’, below, under De la Cruz and Tejedor). This work explores the diachronic transition of certain animal denominations from the Old English period, discussing some of the reasons that have traditionally been adduced to explain semantic change, as well as the factors which may account for word loss and other related phenomena such as the adoption of borrowings, the restriction of meaning of some specific terms and the widening of others. The data have been drawn from a wide range of lexicographic sources, including A Thesaurus of Old English, the Middle English Dictionary, the Historical Thesaurus of English,1 A Dictionary of Selected Synonyms in the Principal Indo-European Languages and the Oxford English Dictionary. However, this study is not intended as an exhaustive listing of all the terms affected, nor does it have as its primary objective to discover all the possible causes of semantic change. We are all aware of the difficulties historical linguists have to face: we know the 1
I am indebted to Dr Christian Kay who generously allowed me to use materials from the ongoing Historical Thesaurus of English.
Shift of meaning in the animal field
141
meaning of a particular term in Old English and Middle English only because of its definition in the consulted lexicographic sources or because of its use in a given translation, or because we deduce it from a specific text. As McMahon (1994: 185) explains, “we cannot establish its full range of senses, or the connotational meanings it had at the relevant period”. Nevertheless, some connections can be drawn to explain semantic shift. 2. Widening or broadening Under the entry broadening, Trask (2000: 47) offers the definition: “the extension of meaning of a word to a wider range of referents than formerly, (as when English dog was broadened from denoting a particular type of canine to denoting any kind of canine)”. He also provides generalisation, widening and extension as synonyms. However, I would like to exclude the latter to avoid confusion with metaphor, as metaphor implies the extension of meaning and transfer of specific attributes or qualities from one referent to another. Henceforth, I will refer to the process as widening, broadening or generalisation. It may appear a contradiction that a word can increase its variety of contexts while experiencing the loss of some specific restrictive features of the term. An example is the case of bird which, according to the Oxford English Dictionary (hereafter OED), was originally the name for the young of any feathered creature; ‘a young bird, a chicken’. It retained that meaning through Middle English times until 1600, when it came to designate not only the young but also the adults. In this case the lost feature is [+ young]. This change probably coincides with the displacement of fowl, from OE fugol, as ‘bird, generally’. We find it in Middle English as in the fowls of the air alongside briddis of the eir (e.g. Wycliff, Matthew viii). The overlap in this semantic space was conducive to the narrowing of fowl to designate ‘domestic fowls, such as cock, hen, chicken and often also duck and turkey’ (OED). Another well-known example of broadening is that of dog, which from denoting a specific breed of dog became the general term for the whole species. Again this phenomenon is linked to the restriction undergone by OE hund which, like its German cognate, was originally the superordinate term and became specialized to refer to a breed used in fox-hunting.
142
Isabel de la Cruz Cabanillas
A different development is observed in the case of rabbit and coney. In Old English we find the item hara (‘hare’), but there is no reference to rabbit. The latter entered the language in Middle English times, as shown by its first recording in the OED (1398). It originally meant ‘the small one’, while later on it was used to designate ‘the full grown cony’ and nowadays refers to an animal of this species of any size. Though rabbit and coney might have been competitors for a while, the connotations of coney made it avoidable and redundant in this sense. According to the OED, the word was regularly rhymed with honey, money, as indicated also by the spelling coney, until the nineteenth century when the pronunciation with a long o gradually crept in. The desire to avoid certain vulgar associations with the word cunny may have contributed to the preference for a different pronunciation in reading the Bible, where this item is used to refer to the animal. Thus, the broadening of rabbit responded to the need to avoid undesirable associations. 3. Narrowing or restriction Narrowing implies that the range of meanings conveyed by a word has been restricted. Just as widening involves the loss of one or more features, narrowing occurs when some restrictive features are added. In this way, deer rendered the original Germanic idea of ‘beast’ in contrast to humans, by adding the feature of [+quadrupeds] to distinguish it from birds and fishes. This restriction went further by adding the features of [+ruminant] and [+having deciduous branching horns or antlers], or [+cervid] which marks the Cervidæ family as being different from other families of the same species. Thus, narrowing is understood as “a type of semantic change in which a word comes to be restricted to fewer cases than formerly” (Trask 2000: 223). However, there are other phenomena that also seem to be involved in the restriction process. The development of deer ‘animal’ towards its present meaning is paralleled by the assimilation of the borrowing from OF beste ‘beast’ (c. 1220). Both terms overlapped with the referential meaning of ‘an animal’, but finally beast, which replaced the Old English term, gave way to animal in the sense of ‘any living creature’, while beast was gradually reduced to denote larger domesticated animals, monsters and fabulous creatures or other metaphorical meanings as ‘a brutish or stupid person’.
Shift of meaning in the animal field
143
Apart from the interference with beast to mean ‘any animal’, deer was beginning to be used as a generic for cervids when hart collided with herte ‘heart’ and was specialized to ‘male deer’, where before it had been the generic term and even found in compounds such as hart-skin, hart-hide, hart-leather, hart-marrow and hart’s-tallow, and maybe also in hart-hunting and hart-hunter, according to Samuels (1972: 73). As well as competing against beast, deer was in use at the same time as fer was being adopted through French, and originally from the Latin feris. Sources for fer date mainly from the fourteenth and fifteenth centuries according to the Middle English Dictionary (hereafter MED), although nowadays it survives only in the adjective form fierce, ferocious: (a1398) * Trev. Barth.(Add 27944) 170b/a: Heremus is a weyeles wyldernesse, and þere wonnen but fewe, outetake wylde bestes and feers [L feris et bestijs]. a1500(?a1400) Firumb.(2) (Fil) 1537: The proud sarisin..Come with twenti thousand..And gonne to crye as wood as any fer.
We can see from the examples above that both processes, widening and narrowing, do not take place in isolation but that there are some other factors involved. The fact that there are other terms available in the language with a similar meaning causes some of the shifts. In the case of narrowing, the restrictions can be complete in the sense that the item gradually reduces the number of contexts in which it appears until it disappears entirely from the system. Let us now have a look at some other lexical units that underwent narrowing, and focus on the possible reasons for their specialisation of meaning or falling away in the English language. 3.1. Language contact When a lexical item is adopted into a system, the loanword takes over one of the central meanings of the equivalent native term, which is often retained with a marginal sense. For example, when French cuisine landed in Britain, the French terms beef, mutton, pork, veal and venison were used to designate the flesh from the living beast, which continued to be denoted by the Old English words. However, the process was a long one since, initially, native and borrowed
144
Isabel de la Cruz Cabanillas
elements were employed for both the animals and the meat. Hence, Samuel Johnson refers to a beef being killed for the house in his Journey to the Western Isles (1775); and William Cowper uses the word mutton to mean ‘a sheep’ in 1795: “A mutton, statelier than the rest, a ram, the eres and wethers, sad, address’d” (cited in Burchfield 1985: 18). The restriction did not become general before the eighteenth century and the consolidation of the new items in the lexicon until a century later. Likewise, the replacement was gradual in pig, used for the flesh even in the fifteenth century, as recorded by the MED. As in other linguistic processes, the change did not take place overnight and definitely not as early as Sir Walter Scott tried to make out. The passage in Ivanhoe (chapter I) in which this distinction between Old French and Old English terms is mentioned in a conversation between Wamba and Gurth has been claimed anachronistic because the episode occurs about a century too early. Beef, mutton and venison are first found in English around 1300 according to the MED and OED, and the action is supposed to take place during the reign of Richard the Lionheart, who died in 1199. Veal made its appearance even later. In fact, veal is attested in a recipe from 1325 in the MED and in Chaucer’s Merchant Tale from 1386 according to the OED. Pork is subject to some contradiction itself; the first OED quotations for beef and mutton date from 1300: the OED includes pork alongside mutton and beef in a quotation from c. 1290 SLegend Mary Magdalene “Huy nomen with heom into heore schip...porc, motoun and beof”, but the MED gives 1300 for the same quotation. Certainly, all five words – beef, mutton, pork, veal and venison– are out of the question for the times of King Richard the Lionheart. Pork is found alongside pig, swine and hog, the last three Anglo-Saxon words which have undergone extensive shift. Pig used to designate ‘the young of a swine’ and by extension ‘a swine of any age’, although clear examples of this usage are rare before the nineteenth century, according to the OED. From an early stage, hog seems to designate ‘a swine reared for slaughter’ and, therefore, often ‘a barrow, a castrated one’. However, it is also used to name the species and in some parts of the USA for the flesh, even today. Hughes (2000: 117) adds to this list the case of boar and brawn; the former used for the animal and the latter for ‘the flesh of a
Shift of meaning in the animal field
145
boar or swine’ (though, according to the OED, applicable to the flesh of other animals as well). Language contact is also evidenced in the case of another French word, cattle, the “collective name for live animals held as property, or reared to serve as food, or for the sake of their milk, skin, wool, etc.”, according to the OED. The tendency in recent times has been to restrict the term to the bovine genus, but the wider meaning is still found locally, and in many combinations. The Norman form, catel, from which cattle derives, was superseded at an early stage by the Parisian chattel. The original meaning of ‘sum of money’ is preserved in the French chattel, which was adopted early in legal Anglo-French, especially in the phrase goods and chattels, used for all kinds of personal property. Parallel to this, we find the development of OE feoh, fioh, feo > Mod. Eng. fee. This Old English word, whose origin traces back to Germanic *fehu, is cognate with Latin pecu ‘cattle’, which is related to Latin pecunia ‘money’. In Old English times it meant both ‘cattle and money’. Fee was still in use with this latter meaning, as well as that of ‘movable property, wealth’, until 1677 and with the sense ‘cattle’ until 1596, although the term is obsolete in present-day English. Cattle, first recorded in Cursor Mundi (1300), probably expanded at the expense of fee. Linked to the history of these words are those of other nouns used to designate big domestic animals like cows, oxen or similar. Thus, in Old English we frequently find neat, which is now only found regionally and considered archaic. In Cædmon’s Hymn we read “ond ut wæs gongende to neata scipene, þara heord him wæs þære neahte beboden”. Even more restricted is the modern use of rother < OE hroðer, hryðer, with the sense of ‘cattle’. The word is now considered obsolete except in some dialects. In Old English there are even more terms to refer to these animals: orf, hwyorf and nieten. Erf from OE orf was found until the fifteenth century with only the sense of ‘cattle’. Hwyorf is one combination of orf preceded by an element which I have been unable to retrieve either from the OED on-line or from the MED. Nieten could be related etymologically to neat and used to mean “a bovine animal; an ox or bullock; a cow or heifer” (OED) or ‘cattle’ when used collectively. Thus, the English language used to have a variety of different terms to designate ‘oxen’ or ‘the bovine species’ in Old English –feoh, hriðer/hryðer, orf, hwyorf, neat and nieten– but most
146
Isabel de la Cruz Cabanillas
died out, unable to withstand the pressure of the introduction of cattle/chattle from Old French. 3.2. Homonymic conflict When two words belong to different parts of speech, they are not likely to be confused, as the word order will not be the same for both. Homonymic elements can present different complementation patterns, in other words they show diverse collocations, determiners and modifiers. If two homonyms belong to the same part of speech and the same semantic field, interference between them can exist, but it is difficult to find homonyms which are also related semantically. Hence, even if OE buc ‘ram’ and buc ‘belly, paunch’ were homonyms, the areas they cover should be different enough to prevent any conflict. The difficulty, though, is to explain why one of them falls out of use, why precisely one is dropped and not its counterpart, and whether collision had anything to do with it. Interference can occur where a taboo word is involved. Although Lass (1997) denies this possibility on the grounds of the number of homonymic items which coexist with their reprehensible and disagreeable counterparts, traditionally it has been thought that speakers tend to avoid those terms which have a pejorative meaning; for this reason, certain words cease to be used and get replaced by others. A well-known case of replacement involving taboo words is that of cock by rooster in American English. Cock was already employed in Old English to designate the domestic fowl animal, but it was only from the seventeenth century onwards (since 1618 according to the OED) that the word acquired its new meaning. The beginning of the shift in denotation ties in with the use of rooster, which was first attested in 1772. However, as Lass (1997: 24) points out, the former is still used to refer to the male of some birds, as in cock pheasant. Another example of avoidance of unpleasant associations is the case of ass. From Old English times it was the common denomination for the animal. In fact, the Authorized Version of the Bible contains 55 occurrences of the word and not a single one for donkey. The present-day English tendency is quite the opposite. The reasons which account for the substitution of ass with donkey are both phonetic and semantic. By 1600 the /r/ in syllable final position when followed by another consonant had stopped being pronounced, with
Shift of meaning in the animal field
147
the subsequent lengthening of the previous vowel (see Dobson 1968: 724 ff). Thus, arse came to be pronounced with a long a. By the end of the seventeenth century there was also a lengthening of short a followed by the voiceless alveolar fricative, which made ass a homophone of arse. Fairman (1994: 31-34) refers to the process and dates avoidance strategies to between the years 1760-1830. He explains that the first strategy was to employ jackass, but soon failed because it ceased to mean the male of animals and “became a lexical determiner” so that speakers continued using ass. Various substitutes took its place in different parts of the country: cuddy/cuddie in Scotland, neddy in the south-west of England and dicky in Essex and Suffolk. According to Fairman (1994: 32), the first instance of donkey is in a list in Robert Nares’ grammar (1784); the OED gives his second reference, Francis Grose (1785), as the first record of the word. Gradually, donkey was gaining ground and speeding the decline of ass. However, he admits that there are still aspects of the process which are unclear and the topic remains open to discussion. Barber insists that ass underwent a modification in pronunciation in order to avoid an objectionable merger and the variant with /æ/ became standardized, although the long vowel can still be heard occasionally in expressions like silly ass (1976: 312313). 3.3. Internal displacement Even where there seems to have been no pressure from outside, some native Old English words that lived side by side disappeared from the system in Middle English times. Berndt (1989: 72) refers to the process as internal displacement and includes examples like OE hana and cock (< OE cocc), OE eosol and ass (< OE assa), and OE eofor and boar (< OE bar), where the second item takes over the function of the first. It seems difficult to determine under what circumstances a term disappeared from the system. As happens with other species, very often two or more denominations live together for some time. For instance, in the case of OE hana both cocc and carlfugol are found for the male in Old English. However, neither hana nor carlfugol is attested from Middle English onwards.
148
Isabel de la Cruz Cabanillas
Regarding assa and eosol/esol, we observe that while the former has no exact analogues in the cognate languages, the latter goes back to Latin asinus and has equivalents in other Indo-European languages; in fact, asal, assal, assald are documented in Old Northumbrian from the original Celtic forms. OE assa was coined, probably as a diminutive, taking as its base the Old Northumbrian items. It also had a feminine form, assen, which did not survive into Middle English. In the case of eofor > ME ever, both the MED and OED provide a quotation from Hali Meidenhad (c. 1225-1230) as the last occurrence of the term. Homonymic conflict here would be with the adverb ever but none of the requirements seem to be met, since the elements involved do not belong to the same word class and the probability of confusion in a specific speech act seems rather remote. The same argument applies to OE bera ‘bear’ and OE bere ‘barley’. The latter is documented in A Thesaurus of Old English, although no traces of such a word are found in the OED. Even though both are nouns, they will rarely co-occur as the semantic areas they cover are quite separate; all possible confusion can be disambiguated by the context. The basic point here is homonymic conflict as a factor capable of causing changes in the affected elements and, in some cases, their replacement by other lexical units, although the motivation for the loss of one particular item over another remains opaque. 4. Conclusions Though the field of animal denominations has not been researched thoroughly, not even the domestic animal portion,2 my original hypothesis seems to have been validated. The specific territory of animal vocabulary provides a large pool of terms that have shifted meaning since their origin; words that have disappeared from the system to be replaced by others, and occasionally some lexical units that expand to cover a wider semantic area at the expense of others. 2
In addition to the examples discussed here, other instances that have not been dealt with include: steed restricted to a specific register (poetic); other minor shifts within the same species, such as the changes observed in ox or bullock; and others in families like goat or sheep, which need further research.
Shift of meaning in the animal field
149
The data show that, while the suggested classifications cannot predict the direction a shift may take, they can to a certain extent be explanatory. Some types of semantic change may be considered predictable under certain circumstances, such as in the case of narrowing due to language contact, or of homonymic collision especially with a taboo word, or because of the widening of a term in the same semantic sphere. As has been shown in the previous pages, very often a process of narrowing seems to be closely linked to another of widening. Though we cannot be certain that both shifts started at the same time, or if one began before the other, the appearance of one undoubtedly helped to reinforce whichever phenomenon was already in progress. This study establishes some new connections in relation to semantic change but obviously new lines will need to be drawn in a systematic way to demonstrate the idea that even though every word has its own history, no word is an island; no word develops in segregation. Changes in one word will reflect closely the experience of other lexical units, so that a shift undergone by one will very likely match the changes in another, albeit in reverse. Bibliography Primary sources Buck, Charles. D. 1988. A Dictionary of Selected Synonyms in the Principal IndoEuropean Languages. Chicago and London: The University of Chicago Press. Kay, Christian, Jane Roberts, Michael Samuels and Irene Wotherspoon. Forthcoming. Historical Thesaurus of English. On-line samples at http://www.arts.gla.ac.uk/ SESLL/EngLang/thesaur/homepage.htm. Middle English Dictionary. On-line at http://ets.umdl.umich.edu/m/med. Roberts, Jane and Christian Kay. 1995. A Thesaurus of Old English in Two Volumes. London: King’s College Centre for Late Antique and Medieval Studies. Simpson, John A. and Edward S. C. Weiner (eds). 1989. The Oxford English Dictionary (on CD-ROM). Oxford: OUP. 2nd ed. and on-line at http:// dictionary.oed.com. Trask, Robert L. 2000. The Dictionary of Historical and Comparative Linguistics. Edinburgh: Edinburgh University Press. Secondary Sources Barber, Charles. 1976. Early Modern English. London: André Deutsch. Berndt, Rolf. 1989. A History of the English Language. Leipzig: VEB Verlag Enzyklopädie. Burchfield, Robert W. 1985. The English Language. Oxford: OUP.
150
Isabel de la Cruz Cabanillas
Campbell, Leslie. 1998. Historical Linguistics. An Introduction. Edinburgh: Edinburgh University Press. De la Cruz, Isabel and Cristina Tejedor. 2002a. ‘The HORSE family: On the evolution of the field and its metaphorization process’ in Díaz Vera, Javier E. (ed.) A Changing World of Words. Studies in English Historical Lexicography, Lexicology and Semantics. Amsterdam: Rodopi: 229-254. ___. 2002b. ‘Women as Evil. Animal Metaphors Referred to Women’ in Moskowich-Spiegel, Isabel (ed.) Re-Interpretations of English Essays on Language, Linguistics and Philology (I). A Coruña: Universidade da Coruña: 9-27. ___. 2006. ‘Chicken or Hen?: Domestic Fowl Metaphors Denoting Human Beings’ in Studia Anglica Posnaniensia 42: 337-354. Dobson, Eric J. 1968. English Pronunciation. 1500-1700. Oxford: OUP (2 vols). Fairman, Tom. 1994. ‘How the Ass Became a Donkey’ in English Today 10(4): 2935. Hughes, Geoffrey. 2000. A History of English Words. Oxford: Blackwell. Lass, Roger. 1997. ‘Arse Longa, Vita Brevis: Last Words on Harmful Homophony’ in Studia Anglica Posnaniensia 32: 21-31. McMahon, April. 1994. Understanding Language Change. Cambridge: CUP. Samuels, Michael L. 1972. Linguistic Evolution with special reference to English. Cambridge: CUP. Sihler, Andrew L. 2000. Language History. An Introduction. Amsterdam: John Benjamins. Traugottt, Elizabeth C. 1989. ‘On the Rise of Epistemic Meanings in English: An Example of Subjectification in Semantic Change’ in Language 65(1): 31-55. ___. 1999. ‘The Role of Pragmatics in Semantic Change’ in Verschueren, Jef (ed.) Pragmatics in 1998. Selected Papers from the 6th International Pragmatics Conference (Vol. 2). Antwerp: International Pragmatics Association: 93-102. Traugottt, Elizabeth C. and Richard B. Dasher. 2002. Regularity in Semantic Change. Cambridge: CUP. Ullmann, Stephen. 1957. Principles of Semantics. Glasgow: University of Glasgow Press.
Different aspects of the specialised nomenclature of ophthalmology in Old and Middle English1 María José Esteve Ramos Universitat Jaume I
Abstract The disease and treatment of the eyes have been recorded in texts since the earliest stages of the English language. Recent approaches to specialised discourse have favoured a closer study of the historical development of so-called languages for specific purposes. In this article, my aim is to study the nomenclature for the field found in two sources: the Old English Leechdoms and Benvenutus’ De Probatissima Arte Oculorum. The two main aspects of the investigation concern morphology and origin, following previous studies in the field (Norri 1992, 1998).
1. Introduction Medical treatises have recently been attracting the interest of the academic community. This has been reflected in different works and projects, like The Corpus of Early English Medical Writing currently under construction at the University of Helsinki. However, ophthalmology has had a distinctive place within medical practice since very early times, when the cure and treatment of the eyes were looked on as a speciality apart. It is interesting for us to note, as has been suggested, that: “although arthritis and rheumatism were common disabilities, herbals and leech books prescribed more remedies for conditions affecting the eyes than for any other single complaint”.2 In this way, ophthalmology provides researchers 1
This article has been funded by Fundació Castelló-Bancaixa, research grant 05I006.26/1. That grant is hereby acknowledged. 2 Medieval Medicine in: http://www.intermaggie.com/med/healing.php (30.11.06).
152
María José Esteve Ramos
interested in studying the evolution of medical lexicon with an opportunity to focus on a particular area of medicine from a diachronic perspective as well as from a synchronic one, something that can pose some difficulties in other subspecialities in which the whole diachronic time span may not offer so many textual sources. In this study, terms related to the diseases and anatomy of the eye have been gathered and studied according to their etymological and morphological characteristics, following previous work in the area (Norri 1992). Results from this investigation may help to shed light on the different processes through which the language of this subspeciality of medicine evolved, and how that was reflected in Old and Middle English in the two selected textual sources. 2. The texts and their context 2.1. The Leechdoms and the science of medicine in Old English In Old English, recipes were much more common that specific treatises on a disease. However, there is one remarkable compendium of medicine dating from this time –including several passages about the treatment of eye diseases– from which the terms for this study have been extracted. The work, entitled Leechdoms, wortcunning and starcraft of Early England, is a very representative example of medical scholarship of the time. According to some, it represents “the best source for the period, and comprises the first appearance in print of some of the earliest known English works on medicine and science”.3 The different texts are compiled in three volumes, totalling 1461 pages. Not all the sections relate specifically to the anatomy and diseases of the eye; therefore, only those parts of the work relating to ophthalmology have been considered for the present study. The title reproduced from the edition by Cockayne is: Leechdoms, wortcunning and starcraft of Early England, being a collection of Documents, for the most part never before printed, illustrating the history of science in this country before the Norman Conquest. The work was collected and edited by the Rev. Oswald Cockayne (M.A. Cantab), and published by Longman in London, 1864. For the present investigation 3
See Nigel Philips Rare Books at: http://www.worldbookdealers.com/books/ printablebook.asp?id=160098 (8.11.02).
Different aspects of the specialised nomenclature of ophthalmology
153
I have consulted the edition published by Thoemmes Pressin Bristol, England, 2001 (http://www.thoemmes.com) reprinted from the original. This collection includes the Herbarium of Apuleius, which opened up the herbal medicine of southern Europe to the English and also various native herbs absent from the classical herbals, the Lacnunga and St Bede’s Treatise on Astronomy and Cosmogony. Thomas Oswald Cockayne (1807–73) was a philologist, a member of the innovative Philological and Early English Text Societies, and a keen student of the Anglo-Saxon language, a subject on which he published numerous works including the Leechdoms itself. This volume, known in its time as Læceboc, together with the Old English versions of the Herbarium Apulei and the Medicina de Quadrupedibus, are the earliest English medical writings dating from the OE period.4 Shorter texts, like recipes, are also available as sources, but they present a more diffuse and difficult source for coherent study. Old English vernacular medical writings coexisted with the numerous texts written in Latin, produced either in England or on the Continent. The Latin texts were mostly the property of the monasteries and other centres of learning. The violent Viking raids (beginning at the end of the eighth century) led to a decline in learning, because many monasteries –notably Lindisfarne– and churches were burnt down and destroyed. King Alfred and religious leaders such as St Dunstan (d. 989), Athelwold (d. 984) and St Oswald of York (d. 992) helped and encouraged the revival of learning in a period of little interest in science, after becoming aware of the importance of providing books in English. It was a common practice for lay healers, for example, to visit monasteries (Norri 1992: 26) to look up texts in the vernacular wherever available. The major type of medical textbook in the Middle Ages was the leechbook and its companion volume, the surgery. Of the leechbooks, the most famous is Bald’s Leechbook, written around the tenth century. They were available on subjects such as bloodletting, diets and herbal remedies, and so were very popular and a very practical source for immediate treatments. However, other texts such as herbals were also readily 4
All three texts were edited and published by Cockayne (1864) in his Leechdoms, Wortcunning and Starcraft of Early England. Altogether it comprises more than one thousand manuscript pages. As Norri (1992: 25) points out: “This is remarkable when one considers that the first medical work known to have been composed in French dates from the thirteenth century.”
154
María José Esteve Ramos
available, and those who could read could use them for healing purposes with no other specific training. The issue of literacy in relation to medical practitioners in this period raises different questions, and different opinions have been voiced: The practice of medicine in early Anglo-Saxon history is not relegated to the backwards world of hocus-pocus magic and spells, as it is often thought, but is rather a very serious discipline. Practitioners of the science were learned physicians and, as the evidence suggests, had two main methods for the rational treatment of the sick: 1) pharmacopeia consisting of plant origin and 2) surgery. (Hamid 2002: 1)
Other authors maintain a different idea of the medieval practitioner in the British Isles. The view vividly expressed by Cockayne (1884 I: xxvi), writing about this topic in the preface to his edition of the Leechdoms, represents one attitude of his contemporary nineteenth-century period. Though he acknowledges the difference between what he calls a medicus or leech, and a lay practitioner, he also adverts to the limitations of the professionals of the time: These leeches, then, unable to use the catheter, the searching knife, the lithotritic hammer, and ignorant of the afar sought Indian drugs, were in their early practice almost wholly thrown back upon the lancet, (…). Not only the Engle and the Saxon, the warrior inhabitants of our own island, but also the races of Gothic invaders, were too rude to learn much of Galenos, or of Alexander of Tralles, though they would fain to do so. (Cockayne 1884, I: xxvi)
He also reminds us of the fact that classical scientific heritage was preserved in Arabic culture, and that this situation of limited intellectual resources was a common feature in Christian Europe, where magic spells and improvised remedies were too often substituted for learned medical practice. On the whole, Cockayne also suggests, something like a “school of Saxon medicine” could have existed: “(…) If these signs refer to native treatises, unknown to us, and now irrecoverable, they go to illustrate the existence of an English school of teaching medicines; as do the expressions “as leeches ken”, not of rare occurrence.” (1864 II: xxiii, xxxi). This view is probably more a reflection of Cockayne’s willingness to believe that such a community existed, than of any conclusive evidence in that regard. Leechdoms were written to compile descriptions of illnesses,
Different aspects of the specialised nomenclature of ophthalmology
155
symptoms and treatments, however contact among physicians at this time may have been difficult. As is widely known, the importance of Latin increased after the Norman Conquest, because of the new universities which taught in Latin, a language that was also used for textbooks. Writers of medical matters used Latin and Anglo-Norman exclusively. The first AngloNorman recipes and charms occur in the early twelfth century and this led to a wider separation between the university-trained physician and the common practitioner of medicine, who was referred to as a leech, a name of Old English origin which from then on represented a clearly differentiated professional profile. However, even the learned physician would have had to defend against the evil image spread by the Church, with ever greater intensity, over the centuries which followed: (…) John of Salisbury, echoed the sentiments expressed by the Medieval Church when he said, “they (physicians) have only two maxims which they never violate, never mind the poor, never refuse money from the rich”. In expressing this view, John of Salisbury puts the medical profession diametrically opposed to the Christian concepts of trusting in God for healing as well as the central Christian virtue of charity. (Hutchinson 2002)
Finally, after a period of splendour in Anglo-Saxon times, English became the language of the illiterate, being used only in noninstitutional contexts. It was not until the fourteenth century, in the Middle English period, that English was used again in medicine (Norri 1992: 27), first in remedy books and subsequently in original treatises. 2.2. A Middle English treatise on the diseases of the eyes: De Probatissima Arte Oculorum The Middle English ophthalmology text selected for this study is known as De Probatissima Arte Oculorum. Despite its Latin title, this text is written in the vernacular. The most recent edition (Eldredge 1996) is based on four manuscripts containing copies of the same text with minor variations. This treatise, although considered a pseudoHippocrates for a long time, is thought to have been written by Benvenutus Grassus, a practitioner of Italian origin who travelled
156
María José Esteve Ramos
around Europe.5 Benvenutus Grassus, like the bulk of practitioners, based his knowledge on experience and his work was of a very pragmatic nature. In fact, his work on ophthalmology is the only proof we have of his existence (Eldredge 1996: 1).6 His methods and also his name are first mentioned by Jean de Yperman in a treatise dated 1328, which indicates that he lived during or prior to the fourteenth century.7 Benvenutus’ methods for treating eye diseases were probably quite widely accepted, a view supported by the fact that the text was copied and distributed around Europe. The very minor importance attributed to authorship in the Middle Ages would explain the suppression of his identity as the author of the treatise. With the exception of recognised authorities like Averroës, Avicenna or Aristotle, little attention was paid to authors whose contribution to theoretical knowledge was minor, being practitioners mainly. People’s main interest was much more pragmatic and focused on effective treatments, as in the terrible period of the Black Death: And this was so indeed, when the terrible plague of the so called Black Death appeared (1347-1351) (...). It is precisely due to this necessity of medicine in the most pragmatic sense, that medical texts spread so quickly, even quicker in these dark times. It has been argued that this clear pragmatic function is what conditioned the “partial” elimination of religious and ideological values before the urgent danger of dying. (Esteve 1999: 517-18)
In this period, language marked the distance between medical practitioners –explained in the previous section– that separated the “relatively few university-trained physicians like Chaucer’s “doctor of phisik” from the unlatined others, specifically, the on-the-job-trained surgeon, barber-surgeon, apothecary, apprentice, cunning man, wise 5
Until very recently (Eldredge 1996), there was no agreement on any author for these texts. In some of the different catalogues referring to surviving manuscripts of the text, the entries were marked “anonymous” or “pseudo-Hippocrates” (see Young and Aitken 1908). 6 A well documented biography of Benvenutus is that of Noe Scalinci (1935): “Questioni biographiche su Benvenuto Grasso jerosolimitano” Atti e Memorie dell´Academia de Storia dell´Arte Sanitari, ser. 2,1. 7 Ch. Laborde was the first to observe this in his “Un oculiste du xiiie siècle: Bienvenu de Jerusalem et son oeuvre: le manuscrit de la Bibliothèque de Metz”. MD thesis (Montpellier, 1901: 8).
Different aspects of the specialised nomenclature of ophthalmology
157
woman, lay sister in a convent, and midwife” (Robbins 1970: 394). However, this distance between the university doctor and the lay practitioner or leech was not as rigid as one might think. Often, beside complicated theoretical treatises in Latin and Greek, we find lunaries, recipes and astrological recommendations, all of them in English (Norri 1992: 28). The question, then, is who used this material and what was it intended for. It would appear that the exchange of information between all levels of medical practitioners was more than sporadic. In England, up to the middle of the fifteenth century at least, many professional, specialised and technical subjects were presented in Latin. The principle applied in theology, law, alchemy, astrology and medicine. But, in some of the disciplines, English was slowly entering into recognised, terminological use. In medicine, empirical recipes and prescriptions for specified cures were normally compiled into more or less formal collections, the forerunners of the seventeenth- and eighteenth-century dispensaries (Robbins 1970: 403). Even among the medically oriented collections of recipes a number of charms appear. We find many references in late Old English literature to the graduate doctor of physick, who occupied a very high social position, as Chaucer himself explains through one of his characters in the Canterbury Tales: “In sangwyn and in pers he clad was al, Lyned with taffata and with sandal” (Canterbury Tales, I, 439-440: in Robinson 1988). In Piers Plowman, the “Phisik” is described as having “furred hodes” and “his cloke of Calabre with all the knappes of golde”, and at the end of the fifteenth century, Henryson (in his Testament of Criseyde) had his “docter in phisik cled in scarlet gown and furrit weill” (Robbins 1970: 407). Their fees were very high and not many people apart from the nobility could afford the services of a university-trained physician. In this respect, Robbins (1970: 408) suggests that “it is probably more than just a literary tradition what the Romaunt of the Rose (ca 1370), among other works, reflects”: Physiciens and advocates Gon right by the same yates; They selle her science for wynnyng, And haunte her craft for gret getyng... They wole not worchen, in no wise, But for lucre and coveitise. (vv, 5721-38)
158
María José Esteve Ramos
The excerpt clearly illustrates the pejorative attitude to the profession described earlier. On the other hand, not too much is known about the practitioners without university training who attended to the vast majority of people in England –in the fifteenth century there were only about sixty university-trained physicians in the whole country. The leech was the man who, possibly having picked up his medical knowledge during an apprenticeship in a monastery infirmary (in the best of cases), would “helpe pore folke þat falleþ in to sykenesse and beþ unconnynge to helpen hem selve and of unpower to hire hem leches” (Asmole MS 1481, f. 4v)” (Robbins 1970: 409). The surgeon, likewise, invites different portraits, as for example this description by Guy de Chauliac (1363), the author of the Chirurgie: 8 The conditions necessary for the surgeon are four: first, he should be learned; second, he should be expert; third, he must be ingenious, and fourth, he should be able to adapt himself[…].Let the surgeon be bold in all sure things, and fearful in dangerous things.
3. Data and Methodology Having described and contextualised the sources, it is now necessary to explain the process by which terms were selected and the criteria used for their analysis. As mentioned already, the examples were extracted from two texts from the Old and Middle English periods, and only passages relating to the area of ophthalmologic description and treatment were selected in the case of the Old English text. Using these resources, lexical units referring to anatomy and diseases were gathered, as can be seen in Tables 1 and 2 below. The examples were then searched and traced, both in the Oxford English Dictionary and in the Historical Thesaurus database, currently under construction at the University of Glasgow.9 In some cases, and especially with the earlier text, classifying the examples posed some problems, mainly in trying to decide whether they were diseases or symptoms. 8
Medicine in Quotations online (American College of Physicians): http://www. acponline.org/cgi-bin/medquotes.pl?subject=surgeons (8.11.02). 9 I am most grateful to Professor Christian Kay for her generosity in allowing me access to her materials.
Different aspects of the specialised nomenclature of ophthalmology
159
The criteria used for classifying and studying the examples were taken principally from Norri’s (1992) study, and a simplified version of the most comprehensive classification used by that author provided the main reference against which to define the variables used. The principles of classification adopted to study the evolution of the vocabulary in the present study are therefore two: (i) the wordformation process of the word, defining its morphology as simplex or complex, mainly on the basis of Bauer’s classifications, and (ii) the origin of the word, according to its etymological classification as Germanic, non-Germanic or mixed origin. Table 1 below presents the examples from the Old English text, a total of 44. In this period, it is important to note that the degree of specificity, although present, is not always represented by the use of a coined terminological unit; therefore, in many cases, and mostly when referring to illnesses, the author needs a whole explanatory phrase, as in “Nectalopas þat is on ure þeodum þe man þe ne mæ3e nengi 3eseo after sunna upgange ær sunna eft on setl ga”. Table 1. Examples from the Old English selected text Ablindiaþ butan ælcon sare Ad eos qui non possunt videre a solis ortu ad occasum Aswollenum eaޚum
María José Esteve Ramos Tyrende eaޚan Unscearpsynum Untrumnysse (wið ޚehwylce untrumnysse þæra eaޚena) Wænne /wenne Wemme (on eaޚum) Wyrmum
The data extracted from the second text are presented in Table 2, which includes a total of 37 examples. Table 2. Examples from the Middle English selected text Aqua putrefacta Aranea Cateract, cateractis Balle Browys Coniunctiua Cornea Cristallinus Crystalline humor Erewarde Eye, eyon,ey, eye ayen, eyen, eyren Eyelyddis Fantastical celle Festeles Fleying flyes Forhede Glassy humor Herys Hollow synews Humor uitreus Iherafrumaxyn Lacrimabili minore Lyddys
Both tables contain significant results. In Table 2, the use of phrases diminishes and there is a strong showing of new terminological units, mainly from Latin. The presence of different characteristics for the two different periods of the language calls for a more detailed discussion.
Different aspects of the specialised nomenclature of ophthalmology
161
4. Analysing the examples The examples gathered in the tables above show distinctive features of the lexicon of this medical subspeciality, especially when the two are compared. In this section, I will attempt to analyse the main peculiarities of this comparison. Most of these terms have been identified without undue difficulty with the help of the Oxford English Dictionary, as mentioned earlier. However, some curiosities remain, which may be an indication of the fluctuation of lexical stock from other languages – like Arabic– which were clearly interfering in the discourse community. However, most of these interferences do not seem to have survived beyond an initial occurrence. One example of this would be the word amesarca. Another important factor to be taken into consideration is that sometimes the borders between a symptom and a disease proper are blurred, as in “ablindiaþ butan ælcon sare” (turn blind without soreness) in the Old English source, probably in reference to the main symptom of glaucoma. This may be related to the state of medical knowledge in the different periods, denoting either a lack of terminology or a lack of knowledge (nomenclature knowledge, in this case) on the part of the writer or practitioner, although the first case is more likely. Old English terms, we find, are mostly complex, but with a distinction with respect to the complex items in Middle English mentioned above: the use of post-modifying phrases with a clear explanatory or referential purpose is highly frequent, as in the example “Cimosis (wiþ þeoradle on eaޚum þe mon ޚefiޚo hæt on læden hatte cimosis)”. Phrases are also used where a description is needed to explain the symptom, sign or disease referred to. The fact that this is so frequent evinces the lack of terminological stock in this subspeciality –above all in relation to diseases– as in tyddernyssa þe beoð on þan aemoran sara (for tenderness of eyes which are sores in the eye roots). In the case of nyctalopia (or nectalopas), the description is given in both Old English and Latin, possibly indicating that the writer obtained the information directly from a Latin source. The issue of the use of phrases in medieval writing has been tackled by a number of authors, who argue that the matter is a complicated one because “Making a distinction between compounds and syntactic phrases is difficult in mediaeval texts (…)” (Norri 1992: 91).
162
María José Esteve Ramos
However, it seems that despite the difficulty, researchers do tend to include phrases as part of their lexical studies: It therefore seems that compounds and phrases should be understood as a continuum, with combinations where the composite parts have lost their formal or semantic identity at one extreme and those in which the components preserve their individual character at the other. Looked at in this way, the degree to which the two parts have fused will become apparent, but this does not entail making a strict distinction between compounds and phrases. (Norri 1992: 92)
This question, though, can affect decisions about certain items, as researchers will always use some degree of subjectivity, regardless of the presence of pre-established criteria: Even when bearing in mind the above criteria, it is often difficult to decide whether or not to include a particular lexical item. My experience appears to parallel that of Pauline Thompson in her dissertation on the OE vocabulary of illness in hagiographic prose texts. She admits that “in the last analysis, a degree of arbitrariness in the inclusion or exclusion of certain words is almost unavoidable. (Thompson 1986: 8, in Norri 1992: 90)
My reflection, at this point, is that a deep knowledge of the specialised subject matter must at least be useful in attempting to classify these extended terms. Several other observations can be derived from the examples in Old English. Although the terminological stock is not quantitatively remarkable, I think there are grounds to say that ophthalmology had by this time already acquired a certain specificity in its lexicon. Words like glaucomata, nectalopas, ordiolum, flie or pocces on eaޚum reveal that ophthalmology, as stated in the introduction, stood more or less as a discipline of its own, and that leeches and professional practitioners used a common terminological stock to describe their cases and treatments, although sometimes inconsistently. In the Leechdoms, the presence of Latin words or expressions does not necessarily mean borrowing, but does give an indication that the different sources from which the writer was extracting the information were written in Latin, as for example in the case of cecitatem oculorum/ dymޚendum eaޚan or eaޚece/oculorum dolorem. In the latter periods, terms occurring in Latin or French do mostly indicate borrowings, although the possible
Different aspects of the specialised nomenclature of ophthalmology
163
coexistence of a native term also indicates the general tendency to adopt the non-Germanic form for specialised purposes and the Germanic term for more lay use. This fact is referred to by Getz: The fact that England had two learned textual traditions, the Latin and the vernacular, tells us something about English society as well. The growth in size of the vernacular medical corpus of texts does not suggest that written voice was being given to folk practice. Instead, it suggests that the audience for learned medical discourse was expanding, as it often had before, from one language into another. (Getz 1991: 4)
In the case of both anatomical and disease terms in Old English, all of the items refer to external elements, in other words, there is nothing that would intimate the practice of surgery –not even for cataracts– although the practice may have been possible, based on the different textual sources explaining the procedure. The anatomical terms do not refer to any internal part of the eye, with the only exception of the generic term eaޚan inneweard “the inwards of the eye”, which does not reflect in any case knowledge of the internal anatomy of the eye. Diseases are normally expressed through reference to their symptoms, generically, as in aswollenum eaޚum “swollen eyes”. Again, this was commonplace in medieval medical literature: Moreover, the present day distinction between disease, symptom and sign was vaguely, if at all, made in mediaeval texts. Peter Jones (1984: 58) speaks of the mediaeval “trend against disease theory”, in which “the description of particular diseases often amounted to no more than a list of the symptoms with which it was associated”. The name of the disorder was used for what would today be classified as a set of symptoms, not for the underlying invasive entity. (Siraisi 1990: 117, in Norri 1992: 87)
In Old English, as already indicated above, the number of terms denoting diseases is far greater than the terms which refer to anatomy, where only 6 examples can be identified. This denotes a poor anatomical knowledge that will increase greatly in the examples extracted from the text from the following period, Middle English, where anatomical nomenclature represents the highest number of the total terminological units, with 25 examples.
164
María José Esteve Ramos
Looking at the morphological aspect of the findings, the nomenclature found in Old English presents 41 complex words, out of 44. The reason, as mentioned above, is the use of phrases and compounds to explain the concepts. Norri (1992: 158) also indicates: One of the salient features of mediaeval English medical terminology is the large variety of names for particular conditions, parts of the body, and medicinal ingredients. Terms from Greek, Latin and Arabic all contributed to this diversity, and when vernacular medical writing began to be practised, the native English element added yet another layer. In response to the plethora of terms, dictionaries and glossaries were compiled, which brought together words and phrases of different origin.
At this point, the non-Germanic element is already present, its incidence highest in terminological units that denote diseases. Nevertheless, in this period Germanic resources still predominate, mainly through the use of phrases to compensate for insufficiencies in the terminological stock. Middle English terms reflect perfectly the transition from one period to the next. Lexical choices favour, in many cases, the nonGermanic root, mostly from French or Latin. Terms from the fourteenth century (Table 2) reflect a deeper knowledge of the internal anatomy of the eye and constitute the area of major terminological growth, in contrast with Old English. In this text, we see the coexistence of different variants like pupilla /pyupyl (with various spellings), and of terms of different origin used to express the same concept: guttam serenam, tunicle saluatricem or cateracte. As indicated above, the examples relating to anatomy increase strongly in number in contrast with the previous period, with 26 examples found in the Middle English text and only 6 in the Leechdoms. In some cases, we find different variants showing different degrees of adaptation of a borrowed form, as in the case of neruus opticus, nerfe obtyk, and even forms in which the post-modifying element is moved to a pre-modifying position, more in line with the prototypical Germanic phrase structure: optici nerffe. The fact that all these forms appear in the same text may be due to the different sources used by Benvenutus, or even due to a dubious personal form choice in different parts of the text. Inconsistency due to the scribe was not rare in this period:
Different aspects of the specialised nomenclature of ophthalmology
165
Many of the coreferential terms used in mediaeval scientific and medical writing represent personal choice and individual coinage rather than any fixed terminology, and equivalent is therefore a more appropriate term than synonymy. (Norri 1992: 160)
Another feature of the Middle English examples is that there is a slight increase in complex forms of mixed origin, such as watery humors or opilate holes, in which the Germanic element seems to struggle to defend its presence. Comparing the etymological choices of the two texts we witness a change in the proportions, almost an inversion, from a higher presence of the Germanic element in the first text with 33 examples, to a predominant presence of words of non-Germanic stock, with a total of 22. The fact that the words from Germanic stock decrease in the second text, to give way to examples from French and Latin origin, would seem to support Norri’s study on the names of sicknesses in which his findings show that: Simplex terms of Gmc origin typically denote common and wellknown ailments or afflictions. (...) The largest group of names of sicknesses, covering a multitude of disorders both physical and mental, consists of simplex terms and derivatives of foreign origin. Adoptions from French with no direct Latin counterpart tend to form part of the more fundamental and everyday medical vocabulary. (Norri 1992: 285)
The non-Germanic vocabulary is mostly from Latin and/or French, although other examples show influence from other languages, and therefore their presence is much more difficult to account for. One example is the term Iherafrumaxyn, probably of Arabic origin, which entered the language through a French or Latin text. There are a number of possible explanations for how Arabic words got into English medical vernacular writing, as suggested here by Getz: Perhaps the best-known medical writing in Middle English is a section of John Trevisa’s translation of the hugely popular De proprietatibus rerum of the Franciscan Bartholomew the Englishman. Bartholomew based his medical material on the writings of the monk and translator Constantine the African, who was said to have brought Islamic medical learning to the Southern Italian school of Salerno at the end of the eleventh century. For Bartholomew, as for so many medieval writers, medical knowledge was an integral part of an encyclopaedic knowledge of the created universe. (Getz 1991: 7)
166
María José Esteve Ramos
Other universally known works as the Secretum Secretorum –among others– originally written in Arabic were very famous, and translations into the different vernacular languages became very popular. This facilitated the expansion of the original vocabulary, especially to those words with difficult or no translation in the target language. 5. Further research agenda The analysis of the extracted examples indicates that there may be grounds to pursue further investigations in this field. One of the immediate possibilities is to expand the diachronic time span, and gather all texts in electronic form to create a small corpus of historical lexicon in the field of ophthalmology. The inclusion of linguistic software in the method of research would certainly improve efficiency in obtaining data and the corresponding collocations. As far as the analytical framework is concerned, the criteria would no doubt benefit from being extended to include a description of the semantic relations and mapping of the terms. Another interesting object of study would be to pursue a more precise definition of the different non-Germanic elements in this area. Further distinction, in terms of origin, within these words from non-Germanic stock seems to be a somewhat complicated task, as Norri (1992: 286) points out: “It often proves impossible to dissociate the French from the Latin influence, and for many names of sicknesses a joint source is probable. Words of French and Latin provenance include common terms, and also rarer items of a more technical kind”. However, not only Latin and French have furnished English with lexical stock. As pointed out earlier, other sources, like Arabic and other European languages, may have contributed to the vocabulary of ophthalmology as well, a circumstance which would reward further investigation. Medical treatises deserve our attention, and by analysing the historical linguistic features, together with other extra-textual factors, we can establish by what processes the lexicon for the different areas and subareas of knowledge has come about. Further studies developing on the approach adopted here will explore scientific discourse from new angles to uncover additional linguistic and non-linguistic features of specialised vocabularies. This
Different aspects of the specialised nomenclature of ophthalmology
167
will aid our understanding, not only of the linguistic devices which operate in semantic change, but of the evolution of genres and specialised terminologies in general. Bibliography Bauer, Laurie. 1983. English word-formation. Cambridge: CUP. Cockayne, Oswald. 1864. (1st ed.; 2001 ed.). Leechdoms, wortcunning and starcraft of Early England. Bristol: Thoemmes Press. Esteve Ramos, Mª José. 1998. ‘In and Out of Europe: the Arabs, the English and the science of Medicine’ in Caramés, José Luis, Carmen Escobedo de Tapia and Jorge Luis Bueno (eds) El Discurso Artístico Norte y Sur: Eurocentrismo y Transculturalismos. Oviedo: Servicio de Publicaciones Universidad de Oviedo: 513-520. Getz, Faye M. 1991. ‘Charity, translation and the language of medical learning in medieval England’ in Bull. Hist. Med., 64: 1-17. Grassus, Benvenutus. 1996. The Wonderful Art of the Eye. A Critical edition of the Middle English Translation of His De Probatissimo Arte Oculorum by Benvenutus Grassus (ed. Laurence M. Eldredge). East Lansing: Michigan State University Press. Hamid, Richard. ‘The influence of Latin in Early Saxon Medicine’. On-line at: http://www.chass.utoronto.ca/c~percy/courses/6361.Hamid.htm (consulted 07/02/2002) Hutchinson, K. s. d. ‘Physicians, clerics and Healing in the Middle Ages’. On-line at: http://stkellen.dcne.net/medievalia.htm (consulted 04.11.2002) Robbins, Rossell H. 1970. ‘Medical Manuscripts in Middle English’ in Speculum, 45: 393-415. Robinson, Fred N. 1998 (3rd ed.). The Riverside Chaucer. Oxford: OUP. Norri, Juhani. 1992. Names of sicknesses in English, 1400-1550: an exploration of the lexical field. Helsinki: Academia Scientiarium Fennica. Young, John and P. Henderson Aitken. 1908. A Catalogue of the Manuscripts in the library of the Hunterian Museum in the Univeristy of Glasgow. Glasgow: James Maclehose and Sons.
Complex predicates in early scientific writing Nuria Bello-Piñón and Dolores Elvira Méndez-Souto University of A Coruña
Abstract The aim of this paper is to describe the level of idiomaticity in early vernacular scientific writing, using for this purpose samples of late medieval scientific texts. An introductory section contextualises scientific discourse in late Medieval England, offering a brief description of the role of the vernacular in that particular register. The next section describes the corpus of data used in the study. The analysis of data considers both the complex predicates and the phrasal verbs used in the texts referred to above. The study of Complex Predicates – following the works of Hiltunen, Kyto and Tanabe (in Brinton and Akimoto 1999), among others – focuses on collocations formed with the verbs HAVE, TAKE, MAKE, DO, and GIVE. The second part of this analysis examines the etymology of such constructions. The final section gathers together the conclusions reached.
1. Introduction The importance of scientific research and the interest in scientific developments are undisputed nowadays. More and more, men and women of the twenty-first century are getting interested and involved in the discoveries and scientific advances so characteristic of what could be called a technological era. But interest in scientific research is not a recent phenomenon, since one of man’s prime concerns has always been to observe and describe how Nature works (Crespo García 2004: 157). However, science and scientific research should not only be regarded only as the study of natural phenomena and the rules and laws that govern the world around us, but also, in Taavitsainen’s words, as “a fundamental aspect of culture [which is] reflected in literature” (2000: 378). It is this aspect of scientific
170
Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
writing as a reflection of medieval culture that commands our attention in this paper. Our linguistic analysis looks at two scientific texts: Geoffrey Chaucer’s Treatise on the Astrolabe and the anonymous The Equatorie of the Planetis, which some scholars have attributed to Chaucer. The analysis will focus on the complex predicates found in both texts, together with the etymological origin of the lexical items in those collocations. Section 1 gives a general overview of the use of scientific writings as corpus material for linguistic studies. Section 2 offers a brief description of complex predicates and the theories we will follow for our analysis. Section 3 presents the complex predicates found in both texts followed by a morphological analysis of those collocations, after which we proceed to the etymological analysis of the components of the predicates. Finally we present the conclusions reached in the light of this survey. 2. Contextualising scientific writings in the late Middle Ages: a corpus for linguistic analysis Even though the most popular areas for diachronic linguistic analysis have traditionally been creative literature and certain religious writings, more and more scientific texts are finding favour as a rich source for diachronic studies on linguistics. The texts used as a corpus for our analysis represent examples of the scientific production of the last decade of the fourteenth century, a period in which the scholastic movement was starting to develop. The emergence of scholasticism led to an increase in production and the spread of scientific works over the course of the fifteenth century (Pahta and Taavitsainen 2004). The scholastic model of scientific writings was preserved right up until the beginning of Empiricism owing to two main circumstances: the lack of linguistic resources in the vernacular for certain concepts; and the socio-political interests involved in investing English with a certain prestige through the use of Latin structures and terminology (Crespo 2004). This explains why the influence of the classical languages (Latin and Greek) together with French (the language of prestige at the time) is felt in scientific literature written throughout this period, affecting both language and structure.
Complex predicates in early scientific writing
171
Another important characteristic of the scientific texts selected as the basis for our study is that the period when both were written – last decade of the fourteenth century – coincides with the beginning of the vernacularisation of scientific texts (Taavitsainen and Pahta 1998; 2004). One of the main features of scientific writing in English is the considerable number of loanwords which appear in these texts owing to the absence of adequate terms in English. Scholars, among them Taavitsainen (2000), agree that the borrowing process peaked in the fourteenth century during the vernacularisation of medical writing. Most Middle English treatises in the vernacular followed the pattern set by the classics, a model which “can be seen in lexis and in syntax, as well as in the macroforms of texts” (Taavitsainen 2000: 380). The vernacularisation of scientific texts in English had to rely on foreign prestigious texts and authors –mostly Greek and Latin– a reliance, furthermore, encouraged by the scholastic movement which was taking form around this time. The tendency was to follow the statements of existing studies rather than to attempt new research using data obtained through experimentation (Taavitsainen 2000). Nevertheless, Chaucer, the author of at least one of our texts, can be regarded as an innovative writer in the sense that he clearly states his intention of writing in “naked” English words in the Prologue to his Astrolabe (1391): wol I shewe the under full light reules and naked wordes in Englissh, for Latyn ne canst thou yit but small, my litel sone. But natheles suffise to the these trewe conclusions in Englissh as wel as sufficith to these noble clerkes Grekes these same conclusions in Grek; and to Arabiens in Arabik, and to Jewes in Ebrew, and to the Latyn folk in Latyn; whiche Latyn folk had hem first out of othere dyverse langages, and writen hem in her owne tunge, that is to seyn, in Latyn.
In this passage, there is also a hint of a nationalistic sense of the need to elevate the vernacular as a valid language for scholarly production against French and Latin, the languages of prestige used by the elite. In fact, authors such as Edmund Spencer, a poet of the sixteenth century, regarded Chaucer as “the well of English undefylled” (quoted in Leith and Graddol 1996: 143-144) owing to his role as an advocate of Anglo-Saxon words over Latin borrowings. Another feature common to both texts chosen for analysis in the present paper is that they belong to the discipline of astronomy.
172
Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
The reasons for the choice of astronomy for our study are diverse. Firstly, because it formed part of the scholastic quadrivium, the four branches of mathematics in medieval education, together with geometry, music and arithmetic, and it was studied at university. Secondly, because astronomy was, in Taavitsainen’s (2000: 379) words, “the main scientific interest of the scholastic age”. Moreover, it is the discipline within which Chaucer wrote one of his non-literary texts, A Treatise on the Astrolabe. 3. A brief note on Complex Predicates (CP) As stated in the introduction, the aim of this study is to describe the role of certain complex structures –structures comprising two different parts– found in two ME scientific texts. Since there is no general designation for such constructions, and different scholars refer to them by different names, we have decided to follow Cattell’s (1984) denomination of “Complex Predicates” (CPs henceforth). The peculiarity of this kind of construction is that it is formed from a verb (usually make, take, have, do or give) and a noun, where the verb lacks all semantic meaning (it only expresses some grammatical characteristic, such as voice and tense). It is the second part of this construction –the noun– which provides the important information, that is, both the syntactic and semantic data (Live 1973; Danlos 1992; Cattell 1984). According to Stein (1991: 31): In the development of the English language there has been a trend in the direction of splitting the verb into two parts. The first part is almost devoid of lexical meaning but embodies the associated grammatical information, being the bearer of the inflectional endings (thus indicating tense, number and person). The second part carries the lexical load, conveying verb-like meaning, although its form is not that of a verb; it is usually preceded by the indefinite article and may be marked by other stigmata of the substantive: the stress pattern […] and the possible co-occurrence of an adjective.
Other scholars (including Brinton and Akimoto, 1999; Brinton, 1996; Live 1973; and Nickel 1968) argue that these constructions comprise not just two parts –verb and deverbal noun– but that a case can be made for tripartite constructions if we take into account the different modifiers that can appear with these predicates.
Complex predicates in early scientific writing
173
According to Matsumoto (1999: 82), modifiers are used in these cases to vary the style of the texts. Matsumoto (1999: 83) establishes three different types of modifiers. In the first place, he identifies adjectives, which are used as emphasizers, good and great being the two used most frequently with CPs. In addition to adjectives, he mentions adverbs, used as emphasizers or amplifiers, and, finally, articles. The indefinite article (the preferred type in these constructions), used with the deverbal noun, gives aspectual meaning to the sentence, even though it (or the definite article) does not necessarily change the meaning of the CP. Matsumoto’s classification (1999: 61) of CPs comprises two different patterns: V + N, and V + N + P. For Brinton and Akimoto (1999: 2-3), however, in addition to the three categories mentioned above, modifiers may also include “possessives, quantifiers, compound adjectives, nouns, and even relative clauses”. The use of CPs is not peculiar to a particular period of the history of the English language. In fact, the use of such constructions evolved considerably from ME up to the Modern English period (Brinton and Akimoto, 1999: 21). They were already used in OE, even though their forms were less productive than in later periods. However, as Brinton and Akimoto (1999: 22) explain, “little research has been done on the topic of the composite predicate in Old English”, so we have very little information on this topic. What is clear is that these collocations were –not always, but most of the time– “calqued on, or influenced by Latin” (Brinton and Akimoto 1999: 28). This fact changed, however, in ME, as a result of the influence of French vocabulary on English and of borrowings into the language. 4. Corpus material and analysis of data For our corpus of data, we have chosen two treatises on Astronomy: A Treatise on the Astrolabe, a text of 14,908 words written by Chaucer in 1391, and a sample of 7013 words from the anonymous The Equatorie of the Planetis, written in 1393, taken from the well-known Helsinki Corpus. In this study we will approach the analysis of the CPs on two fronts. The first will be to analyse the number of occurrences found in the texts under survey; the second will be to find out the etymological
174
Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
origin of the nouns that appear in these constructions, in order to determine if the vernacular was the preferred language or if, instead, it was Latin and French that predominated. To identify the CPs in the two texts, we decided to highlight constructions which use the verbs make, take, have, give and do, in line with previous studies on this subject by scholars such as Nickel (1968), Live (1973), Stein (1991), Algeo (1996), Brinton (1996), Brinton and Akimoto (1999), Matsumoto (1999) and Tanabe (1999), who in their different works focus their attention on the constructions comprising these five verbs and a deverbal noun. Tables 1 and 2 below present the total number of CPs found in both texts: Table 1. The Astrolabe
Table 2. The Equatorie of the Planetis TOTAL Nº OF CPs: 5 TYPES TOKENS #/% #/% 1 / 25% 1 / 20% 1 / 25% 1 / 20% 2 / 50% 3 / 60% 0 / 0.0% 0 / 0.0% 0 / 0.0% 0 / 0.0%
The percentages obtained indicate that the majority of CP types in the Astrolabe are constructed with the verbs make and have. However, the frequency of tokens with the verb take is higher. In the case of the Equatorie, the verb have shows the highest proportion of types and tokens. Significantly, out of the five types of verbs commonly used in CP constructions, no examples of give and do are recorded. Following Matsumoto’s (1999) classification of CP patterns, Tables 3 and 4 show that the preferred structure of CPs in The Astrolabe is V+N+P. Nevertheless, in the Equatorie the V+N pattern is just as common.
Complex predicates in early scientific writing
Table 3. The Astrolabe TYPES Nº / % TOKENS Nº / %
175
Table 4. The Equatorie of the Planetis
V+N pattern
V+N+P pattern
V+N pattern
V+N+P pattern
2/ 20% 5/ 38.5%
8/ 80% 8/ 61.53%
2/ 50% 2/ 40%
2/ 50% 3/ 60%
The second part of the analysis of the CPs in both texts investigates the etymological origin of the nouns in each predicate. To assist in this task, we have consulted Skeat’s An Etymological Dictionary of English and the Middle English Dictionary. Our objective is to establish whether the nouns used in the CPs are of English origin or adopted from other languages. Although the origin of the nouns in the collocations is not representative of the origin of the rest of the words in both texts, this aspect of the study will add a new dimension to our understanding of CPs in early scientific writing in English in general. For the etymological analysis of the collocations, only types were taken into account. The number of types, as stated above, is thirteen and most of the nouns have a non-English origin, as recorded in Table 5: Table 5. Nouns appearing in CPs, classified by etymological origin. A Treatise on the The Equatorie of the Astrolabe Planetis Latin conclusyoun equacioun, equacoun Latin > French Old French Old French > Latin English
measure, excused Nombre Excepcioun beginning, see, kep, knowing, mynde.
Nombre
Figures 1 and 2 also illustrate the percentages of borrowings and their languages of origin in both texts:
176
Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
Figure 1. A Treatise on the Astrolabe: Origin of nouns in CPs. 10% 20% Lt. Lt. > Fr.
50%
O. Fr. 10% 10%
O. Fr. > Lt. E.
Figure 2. Equatorie of the Planetis: Origin of nouns in CPs. 0; 0% 0: 1: 33%
Lt. Lt. > Fr. O. Fr. O.Fr. > Lt. E.
0: 0%
2: 67%
The graphs show that while nouns with an Anglo-Saxon origin predominate among the CPs found in the Astrolabe, in the Equatorie there are none. It is also significant that in the Astrolabe half of the nouns found in the CPs have been borrowed from a Romance language –Latin, Old French or French– directly or indirectly. However, given that one of the texts alone doubles the total number of words selected for this study, it is important to bear in mind that the number of CPs found is not representative of the etymological origin of the rest of the words used in both texts. It is interesting to note, though, that despite Chaucer’s championing of the vernacular in his Introduction to the Astrolabe, the presence of words with a classical origin is still obvious.
Complex predicates in early scientific writing
177
5. Conclusions The analysis of both texts highlights some interesting points. Even though Matsumoto (1999) states that in Middle English CPs are very frequent, we must assume that this is the case only for certain types of texts. In the analysis he performs on Chaucer’s works, for instance, there are many examples of CPs. However, in the present study the number of CPs found is not significant in relation to the number of words used in our corpus. This discrepancy may be due to the fact that Matsumoto focuses his study on Chaucer’s non-scientific texts, a possibility reinforced by that same author’s claim that: “Chaucer uses the CP form as a poetic, stylistic device” (1999: 61). Since we are dealing with a non-stylistic text, CPs are not used as frequently. Another explanation for the relatively small number of CPs found in our scientific texts is that during these early periods of the language (OE, ME), CP constructions were not highly idiomatised and interchangeability of verbs was still possible (Brinton and Akimoto, 1999; Matsumoto, 1999). Consequently, one possible reason for the low occurrence of constructions with do and give is that the influence of French and Latin, for example, conditioned the use of certain verbs to the detriment of others. Regarding the etymological analysis of the CPs, we observe that most of the nouns used have an Anglo-Saxon origin. This, again, contrasts with Matsumoto’s claim that loan words from Latin and French outweigh native words (1999). We agree with this claim in relation to the whole text and accept that the low number of CPs which we have identified cannot be taken as representative of the rest of the words in the texts. Further analysis on the different lexical categories remains to be carried out but this would go beyond the aims of the present study. However we can say that on the basis of our analysis of CPs in non-scientific texts at least, Matsumoto’s statement does not apply. Bibliography Primary Sources Chaucer, Geoffrey. 1968. ‘A Treatise on the Astrolabe’ (ed. W.W. Skeat, The Early English Text Society Extra Series). Oxford: OUP. Matti et al. 1991. The Helsinki Corpus of English Texts. Diachronic Part. Helsinki: Helsinki University Press. (Chaucer’s The Equatorie of the planetis.)
178
Nuria Bello-Piñón and Dolores Elvira Méndez-Souto
Secondary sources Algeo, John. 1996. ‘Having a look at the expanded predicate’ in Aarts, Bas and Charles F. Meyer (eds) The Verb in Contemporary English. Cambridge: CUP: 203-217 Brinton, Laurel J. 1996. ‘Attitudes Towards Increasing Segmentalization: Complex and Phrasal Verbs in English’ in Journal of English Linguistics 24 (3): 186-205. Brinton, Laurel J. and Minoji Akimoto. 1999. ‘The Origin of Composite Predicate in Old English’ in Brinton, L.J. and M. Akimoto (eds) Collocational and Idiomatic Aspects of Composite Predicates in the History of English. Studies in Language Companion Series (SLCS). Amsterdam and Philadelphia: John Benjamins Publishing Company: 21-58 Cattell, Ray (ed.). 1984. ‘Composite Predicates in English’ in Syntax and Semantics, Vol. 17. Crespo-García, Begoña. 2004. ‘General Survey of the Growth of Scientific Culture. A Historical Approach’ in Woodward Smith, Elizabeth (ed.) About Culture. A Coruña: Servicio de Publicacións Universidade da Coruña: 157-165. Danlos, L. 1992. ‘Support verb constructions: linguistic properties, representation, translation’ in Journal of French Language Studies 2: 1-32. Graddol, David, Dick Leith, and Joan Swann. 1996. English History and Change. London: The Open University. Jespersen, Otto. 1954. A Modern English Grammar on Historical Principles (1909-1945). London: George Allen and Unwin. (V. VI – Morphology.) Live, Anna H. 1973. ‘The take-have Phrasal in English’ in Linguistics 95: 31-50. Matsumoto, Meiko. 1999. ‘Composite Predicates in Middle English’ in Brinton, L. J. and M. Akimoto (eds): 59-95. Nickel, Gerhard. 1968. ‘Complex Verbal Structures in English’ in IRAL (International Review of Applied Linguistics) 6(1): 1-21. Stein, Gabriele. 1991. ‘The Phrasal Verb Type “To Have a Look” in Modern English’ in International Review of Applied Linguistics in Language Teaching 29: 1-29. Taavitsainen, Irma. 2000. ‘Science’ in Brown, Peter (ed.) The Chaucer Companion. Oxford: Blackwells: 378-396. Taavitsainen, Irma and Päivi Pahta 1998. ‘Vernacularisation of medical writing in English: a corpus-based study of scholasticism’ in Early Science and Medicine 3(2): 157185. ___. 2000. ‘Vernacularisation of scientific and medical writing in its sociohistorical context’ in Taavitsainen I. and P. Pahta (eds) Medical and Scientific Writing in Late Medieval English. Cambridge: CUP: 1-19. Tanabe, Harumi. 1999. ‘Composite Predicates and Phrasal Verbs in The Paston Letters’ in Brinton, L.J. and M. Akimoto (eds): 97-132.
Sixteenth-century glosses to a fifteenth-century gynaecological treatise (BL, MS Sloane 249, ff. 180v205v): A scientifically biased revision Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez University of Las Palmas
Abstract In the sixteenth century, Doctor John Wotton revised a fifteenth-century gynaecological manuscript of his property (BL, MS Sloane 249) by modifying, deleting and adding suffixes, words, phrases and even complete paragraphs. Although Wotton’s task was apparently to update what was an old-fashioned language for his time, most of the changes were devised to achieve greater precision in the expression of medical contents, hinting at some emerging awareness of scientific discourse. This article analyses Wotton’s lexical substitutions and stylistic changes as a particular method of adapting the language of the manuscript to a text type that was then in the process of configuration, i.e. the medical treatise.
1. Introduction British Library MS Sloane 249, ff. 180v-205v,1 (henceforward S249) is a fifteenth-century English manuscript2 in the tradition of Gilbertus Anglicus’ thirteenth-century Compendium medicinae, which has been listed by Green (1992: 81-82) under the heading of “The Sekenesse of 1
Reference as in eVK 1956.00 (Voigts and Kurtz 2000). S249 is written predominantly in English, except for folios 202v-204r where three Latin chapters are included, namely: (inc.) Ad menstrua prouocanda (ff. 202v-203r), (inc.) Ad restringendum coytum (ff. 203r-203v) and (inc.) De tumore mamill (ff. 203v-204r). 2
180
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez
Wymmen” Version 2. Originally an independent fascicle on gynaecology and obstetrics, in the sixteenth century an almost unknown Dr John Wotton incorporated S249 into his own collection of remedies entitled Collectiones medicinales ordine alphabetico instructae (Green 1992: 62; Green and Mooney 2006: 474, note 54). The text was subjected to heavy correction and annotation, probably due to John Wotton himself, who attempted to update its fifteenthcentury English to make it more familiar and comprehensible to a Renaissance reader.3 The replacement, insertion and deletion of multiple letters, affixes, words and even complete sentences indicate his very great divergence from the manuscript language. Thorns, for example, were replaced with the spelling
almost throughout; <-h> forms for the third-person plural pronouns were substituted for <-th> forms; the negative particle not superseded the double-negative construction ne...nought; obsolete words disappeared in favour of new ones, etc. A study and classification of these glosses have already been carried out by Rodríguez-Álvarez and Domínguez-Rodríguez (2005). In this article we will address both the lexical and stylistic changes introduced in S249 by Wotton, exposing the glossator’s awareness of the peculiarities required by English medical language in the sixteenth century. By this stage, a large amount of lexicon had entered medical treatises to meet the demands of a rapidly evolving discipline (Norri 2004: 109-110; Peters 2004: 213-215). McConchie confirms that “from about 1560 there is a spectacular increase in the scope and sophistication of vernacular medical texts, and therefore in the amount of terminology put into use” (1997: 93). This was partly the result of a complex process of social and scientific transformation. The Renaissance witnessed radical changes in the field of medicine. Most importantly of all, the Church ceased to play a major role in the restriction of scientific advances. The origin and healing of physical disorders began slowly to be disassociated from questions of divine will and punishment (Grell and Cunningham 1993: 101-102, 104; Wear 2000: 30-32). This gradual distancing from the Church became definite in England when The Royal College of Physicians 3
The number and accuracy of Wotton’s annotations and modifications suggest his intention to publish the text (Rodríguez-Álvarez and Domínguez-Rodríguez 2005: 47; Green and Mooney 2006: 479, note 68).
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
181
(London) decided to exclude all members of religious orders (Simpson 1962: 50-60). Emerging theories in anatomy and physiology led to innovative approaches and experimentation involving the human body, and developed new trends in the study of medicine (Wear 2000: 37-40). The dissections of Andreas Vesalius corrected some Hippocratic and Galenic beliefs on internal anatomy (Inglis 1965: 81-83; Siraisi 1997: 142-144), while a number of other discoveries –such as the circulation of blood and the different functions of the heart by William Harvey, or the transmission of contagious diseases by Girolamo Fracastero (López-Piñero 1985: 21, 27-28; Bellamy 1992: 128, 116)– helped to advance understanding of the human structure and its physical workings. These early contributions to the field could be made accessible to a wider audience thanks to the emergence of the printing press, which, at that time, had a double influence on science: First […] it sowed knowledge, providing a wider audience than could ever have been without printing, while serving as well to emphasise the authority of the written word. Secondly, it peculiarly influenced the development of biological sciences, by making possible the dissemination of identical illustrations. […] It became increasingly normal to publish one’s discoveries, thus ensuring that one’s ideas were not lost, but were available to provide a basis for the work of others. (Boas Hall 1994: 29)
It was against this background that sixteenth-century physicians could begin to dismiss from their minds the long-established burden of medical traditions, religion and superstition, which had been repressing and shaping scientific thought for so long. Regarding prose style, sixteenth-century English was distinct from late Middle English. The new scholars educated in the humanist doctrine abandoned “the trailing fifteenth-century ‘French’ sentence” (Gordon 1966: 79) as well as features of the so-called curial prose, not only present in legal texts (Rodríguez-Alvarez 1996) but also in other types of writing during the fourteenth and fifteenth centuries (Burnley 1986b, Mueller 1984: 111-158). Instead, they adopted the qualities advocated by the ‘rhetorical’ prose proclaimed, principally, by Thomas Wilson in his The Arte of Rhetorique (1553) (Gordon 1966: 79). It seems inevitable, then, that a Renaissance physician such as Wotton himself should not feel comfortable with the language of a
182
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez
scientific text compiled and copied a century before. As a consequence, he modified all the linguistic elements he considered necessary to achieve rigour and preciseness in the rendering of the contents, or according to Wilson’s rhetorical terms, plainness and aptness. Without question, Wotton’s interference in the original text marks a step forward in the configuration of medical discourse in English. 2. Lexical substitutions Most of the changes in S249 relate to the lexicon. On the whole, they fall into three types: (i) deletion of words from the original manuscript, (ii) introduction of new terms, and (iii) lexical substitutions. Sometimes, the replacement of one lexical item for another also requires that the relevant phrase or sentence be rewritten, which can have an added effect, for instance, on collocation or number agreement (see Rodríguez-Álvarez and DomínguezRodríguez 2005: 56, 61). Although lexical substitutions are relatively abundant in this manuscript, we have confined our study to those that contribute to the reformulation of the text within a scientific frame. For this reason we have not included substitutions for words that were already considered obsolete in the sixteenth century, since they simply represent an updating of the language in the text. Such is the case in example (1) below, where then, potage and taketh awaye are preferred to the oldfashioned sethen, wortes and fordo: (1) & sethen ªthenº quenche them in the mylke of Almondes in the mylde of a gote that is fedde with gode herbis or with cowes mylke (186r, 15-17)4
4
In editing the original text, we have decided to retain Wotton’s techniques as far as possible; we also underline and cross out words whenever the glossator does it. Words, phrases or clauses that he added appear between special square brackets in our transcription: ª…º. The caret (^) before some of the brackets are a typographical representation of the symbol used by Wotton to indicate the exact place where his addition was to be inserted. Finally, the numbers at the end of each example correspond to the folio and line(s) in S249.
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
183
And amonge all thinges that men vse Rys & whete thikketh moche a mannes blode & wortes ^ ªpotageº y made of drage (186r, 19-21) Also doth xv · graynes of pyonys soden in wyne & ydronke fordo þe ªtaketh awayeº suffocacion of the modir (192v, 27-28)
Our classification of lexical substitution in S249 is just a first step in our study of the glossator’s revision of the vocabulary of the text. In order to group the changes into coherent sets, we have taken into account the nature of the words removed and have thus distinguished three categories: 1. High-frequency generic empty-words in everyday language 2. Words describing emotional affections or relating to the field of feelings 3. Too positive or categorical adjectives 2.1. High-frequency generic empty-words in everyday language Generic words almost empty of meaning –such as any form of the verbs do, have, make and be, or the noun thing– are either replaced with medical terms or with words more appropriate to the gynaecological discipline. For example, the verb have in (2), which has the general meaning of ‘possess’ and can be found in a variety of contexts, gives place to the more specific suffer, which complements the essential idea of ‘having’ with the feelings of pain, discomfort and great sorrow that are so often associated with illness. The glossator, then, conveys this sense of experiencing something unpleasant to the body through the new verb: (2) it have ªsufferº the grettest sekenesse of the body the whiles they leven than to ben y heled (188v, 1)5
In example (3), there is a double substitution. First, the too general verb to be is changed for the formal proceed, so that cold and the evil humours occasioning grief in the female body are explicitly 5 Some of the excerpts taken from S249 to illustrate certain points of our discussion contain several other glosses that are not relevant to the topic in focus; we have removed them for clarity’s sake.
184
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez
linked by a cause-effect relationship. Second, the causative expression maketh her to haue þat greuance is turned into cause this greefe, in order to emphasise the physiological aetiology of the disease: (3) And yif the humours be ^ ªproceedeº of colde that maketh her to haue þat greuance ªcause this greefeº (189r, 11-12)
In a similar way, the overused verb maketh gives way to discribethe in (4), a more appropriate term to describe the action of depicting something (namely, trocissis of mirre) in words: (4) Or els yeve hir trocissis of mirre that Rasis maketh ªdiscribetheº in his boke of Almasorum the which ben these (200v, 19-20)
Finally, in excerpt (5), all the vagueness of meaning and reference implied by thinges is solved by using medicines instead. The two lines also contain an adverbial change where the formal and archaic herefore is substituted for the idiomatic expression for the purpose, to indicate that the ensuing list of natural elements can be used to heal the organism: (5) And such thinges ªmedicinesº that ben good herefore ^ ªfor the purposeº is Gallia muscata, muske, Ζilocassia… (188v, 30-31)
2.2. Words describing emotional affections or related to the field of feelings Our second group of discarded words belongs to the domain of feelings and emotions. On this occasion, the glossator attempts to render the physical disturbances of the female body as a medical reality separate from any sort of affective or personal involvement. In (6), for example, the word greuance is replaced with payne. Here, Wotton is right in his perception of meaning differences: whereas to be in greuance is more properly related to a state of mental distress, payne reflects ‘physical suffering associated with bodily disorders’ (Merriam-Webster Online). (6) & oþer whiles they beten the erth with thire handes & with there fete ^ ªby reson of the grete payneº for the greuance that they have / (188r, 15-16)
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
185
In the same vein, in (7), greuaunce is replaced with difficultie, a term that Wotton could have believed more accurately descriptive of the symptom in question. Changing greuance for difficultie eliminates any trace of gloom and mental distress, and denotes only that the woman’s breathing is laborious: (7) & hir sides swellen & the haven greuance to brethen & to dr ªfeele a difficultie of breath whenº awe wynde Also But ªand drawene there winde &º the grettest greuance that they haven is abouten there share & ther raynes (194v, 5-7)
Again, in (8), the glossator prefers the word disease to sorowe for similar reasons. Disease denotes ‘harm or affection in the body’; and even though sorowe could refer to ‘physical pain’ in texts ranging from the fourteenth to the sixteenth century (OED), it was mainly found in contexts where extreme spiritual affliction was experienced: (8) Also doth xv · graynes of pyonys soden in wyne & ydronke fordo þe suffocacion of the modir & helpeth that sorowe ªdiseaseº · (192v, 2728)
A further instance is (9), where, apart from the replacement of the verb haue for the more specific feel (‘to perceive physical sensation of something’), pain is introduced to substitute woo because the latter is closely connected to a state of mind or intense feelings of regret and would be more fitting in a non-scientific document: (9) And yif she haue ªfeelenº moch woo ªpayneº in hir wombe it ys a signe þat she is with child (195v, 35-36)
Lastly, the lines in (10) illustrate Wotton’s dismissal of poetic overtones in the medical discourse of S249. During the Middle Ages, the term wynter preceded by a numeral was used to indicate age; the OED asserts that this commonplace construction was synonymous with a numeral plus the expression years old, and coexisted in the language until the sixteenth century. However, from that time onwards, the use of wynter as an age-marker was progressively restricted to poetry, a circumstance that may have motivated its removal from S249 and the introduction of the more neutral year:
186
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez (10) And thay have such purgaciones from ^ ªtheº tyme of twelve wynter ^ ªyeares of theireº age vnto the age of fyfty wynter ªyearesº (181r, 2-3)
In spite of all these corrections, we should bear in mind that the three words discussed in this section –‘grief’, ‘sorrow’ and ‘woe’– could be used in reference to unhealthy bodily conditions or to physical sufferings in the sixteenth century as perfectly as nowadays, although Peters (2004: 208-213) notes that the number of occurrences with the physical sense is noticeably lower. It is no wonder, then, that the glossator of S249 decided to assume the dispassionate, direct and unambiguous writing expected of a scientific treatise, moving away from the realm of feelings and figurative language. 2.3. Too positive or categorical adjectives In keeping with the medical spirit Wotton wished to instil into S249, he was equally consistent in resorting to plain, impartial adjectives, so that too positive or categorical qualifications were scrupulously avoided. This tendency evinces, once more, the cautious and noncompromising attitude typically ruling the composition of medical writings. Consequently, in (11 (a, b, c)), the adjective profitable – which denotes certain advantage or usefulness– is deleted and the moderate conveniente (in (a)) and good (in (b) and (c)) are substituted in its place. It seems that Wotton considered profitable more appropriate to material than to therapeutic benefits. (11) (a) CURA ffor to help women of these sekenesse there ben many diuers medycenes as blode lettynge in oþer places to deliuere hem of blode that they maw nought ben ypurged of & that is ^ ªconvenienteº profitable list ªleasteº they fall in to a Cardiacle (182v, 26-30) (b) Also there with gotes mylke þat is full profitable ªgoodº for this sekenesse (186r, 10-11) (c) maketh a woman ben sone deliuered of child thogh it be dede in her wombe & it is profitable ª& in this case it is goodº to make them to fnese with poudre of peper & of castory & cast it in her nose (199r, 31-33)
A final example is found in (12), in which the fifteenthcentury adjective worshipfull appears to be excessively grand for a
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
187
syrup preparation, and so the neutral and less reverential good is interposed in its stead: (12) Also a worshipfull ªgoodº serup that myghtlich bringeth forth the corupt blode fro the moder (184r, 15-16)
3. Direct and concise prose As explained above, Wotton’s exhaustive revision of the prose of the original manuscript was guided throughout by his desire for a clear, concise and direct medical language. This was also true of the features that were to characterise the syntax and style of the text. In this section, we will examine the modifications affecting the basic features of the so-called curial prose, i.e. a distinctive prose style that flourished during the fourteenth and fifteenth centuries; those syntactic changes that were common to the language of the period will be not included (for an account of these changes in S249 see Rodríguez-Álvarez and Domínguez-Rodríguez 2005: 53-58). Curial prose, which emerged as a product of medieval chanceries (Burnley 1986a: 3), features an elaborate style characterised, in syntactic terms, by a “tendency to extend the length of groups by lists and by doublets, and sometimes by extended modifiers and qualifiers” (Burnley 1986b: 597), giving way, eventually, to a trailing sentence-structure (Gordon 1966: 50-52; 56-57). This prose style spread beyond chancery documents to pervade the language of works composed according to Latin and French models (Davis 1961: 175-178; Rodríguez-Álvarez 2006: 190-193). Lexical repetitions and lexical doublets –two devices that gave cohesion to curial texts (Burnley 1986b: 601-602)– were originally present in S249, but Wotton elided them by removing and rewriting certain passages. His efforts resulted in a more concise prose supported by shorter and simpler clauses that sought to transmit the contents in a more straightforward manner. To illustrate Wotton’s “editorial changes” to minimize the presence of rhetorical contrivances in S249, we will first reproduce part of the original phrasing from the chapter on “withholdyng of this blode” (or amenorrhea) and its modified version (numbered (13) and (14), respectively, below):
188
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez (13) And yif it be of moche fastyng oþer of moche walkyng let diete her with moche with gode mettes & good drinkes that mowe make herre to have good blode / & lete here make her mery & glade & leven her hevinesse of hir hevy thoughtes (184r, 9-12) (14) And yif it be of moche fastyng oþer of moche walkyng diete her gode mettes & drinkes that breede before good blode / & let her put awaie sadnes (184r, 9-12)
The short, direct text resulting from Wotton’s changes contains neither lexical repetition nor semantically related words. On the one hand, he removes one of the occurrences of good since the coordinate construction makes it clear that the adjective is qualifying both mettes and drinkes. On the other hand, he shortens the lines & lete here make … thoughtes to a briefer & let her put awaie sadness, managing to eliminate both the lexical doublets mery & glade and the lexical repetition of hevinesse/hevy. The following examples also show how our glossator deliberately sought to avoid lexical doublets and repetitions. In (15), foule is taken away since the intended meaning of manner is clearly stated by the adverb horribly, while in (16) the locative expression of that towne of london becomes of that cite because the endophoric reference in the immediate context leaves no doubt as to which English city is being referred to. (15) thinges that stynkyn horribly & foule (188v, 12) (16) And ther was a woman in london & in hadde this y dropsie & she was holden mouvable thurgh after leches ªof all physiciansº of that ªciteº towne of london (193v, 13-15)
Likewise, in (17), the object of the author’s prayers in both occurrences is quite clearly God; aware of this double reference, Wotton elides the second occurrence of the word god and introduces the alternative noun phrase His grace: (17) Praying <…> god & to his blessefull moder marie full of grace to sende me grace trewly to wryte to the ^ ªHis graceº plesaunce of god & to all womannes helpyng (180v, 11-14)
Sometimes, repetition reached the level of clauses (Bornstein 1977: 376-377), as in (18). The use of the discourse deictic such
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
189
before blode makes redundant the presence of that they shuld haue in purgacion and, accordingly, Wotton removes it: (18) The first ^ ªgreefeº is stoppyng of the blode that they shuld haue in ªbyº purgacion and be purged ^ ªofº as I haue saide The seconde is to moche flowyng of suche blode that they shuld haue in purgacion (181r, 24-27)
Other times, he merely avoids long sentences resulting from the inclusion of modifiers or qualifiers. In (19), concision is achieved through syntactic reorganization and the substitution of the adverbial clause as sone As men may for the adverbial phrase with spede: (19) but yif ^ ªexcepteº a man oþer a woman be more y febled ªverie muche weaknedº therby for than men shall sesen it as sone As men may ^ ªit should be stopped with speedeº (185r, 31-32)
In (20), Wotton prefers a two-element noun phrase to replace the long sentence lykyng in Bathes also So that they… Thus, he is able not only to express the message more directly, but also to confer it with syntactic harmony by providing another noun phrase as the third element of coordination after rest and slepe: (20) moche rest & slepe & lykyng in Bathes ªmoderate bathingº also So that they dwell naught long yn the Bathe (189r, 36-38)
Wotton’s solution in (21) is more practical and obvious. In the original text the author wanted the reader to visualise a particular shape and so he recalled two familiar images: a narrow long ball (an euelonge balle) and an egg (an neye). (21) & make it of the shappe of an euelonge balle after the shappe of an neye & put that balle in to the priue membre (191r, 23-25)
Considering both options, Wotton probably reckoned that the image of an egg could easily represent this shape and therefore crosses out the remaining possibility to leave the text like this: (22) & make it of the shappe of an egge & put that balle in to the priue membre (191r, 23-25)
190
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez
Finally, in examples (23), (24) and (25), complex causative constructions are withdrawn and replaced with transitive verbs plus shorter direct objects: (23) & let her vsen porselane letyse // mynte plantayne sorell & roses in dedycyns to make þe flux to sese ªstayed the fluxº (185v, 25-26) (24) yiff these may nought ben hadde make a / plastre of mogwede y soden in water & emplastre the women therwith from the navell to the priue membre / for it maketh a woman some to be de ªcauseth speedie deliveranceº liuered of hir child quyk yif it be oþer dede / in her wombe (198v-199r, 38; 1-3) (25) Also it maketh a woman that is stopped sone to be deliuered by the purgacion of herre blode ªand provokethe her monethlie coursesº (199r, 10-11)
4. Conclusions As we have observed over the course of these pages, an analysis of the multiple glosses in S249 can give us some insight into the mental workings and linguistic strategies of a self-conscious physician in search of a specific jargon and prose style for his profession. In relation to the lexical substitutions, we can conclude that the editorial decisions of John Wotton, if he is our glossator, tend towards the elimination of generic terms almost empty of meaning, words carrying emotional connotations and categorical adjectivization. His changes confer the “new” text with precision, neutrality, simplicity and clarity. Likewise, the stylistic refinement of the original is intended mainly to avoid unnecessary repetitions and syntactic elaboration with a view to achieving a concise and unambiguous scientific discourse. The present study –which focuses on certain lexical substitutions and stylistic modifications found in a selection of excerpts from S249– complements findings on this manuscript included in a more comprehensive analysis which revealed further linguistic changes involved in the creation of a proper medical writing: the introduction of a terminology appropriate to the gynaecological context, the use of deictics and repetition to make references clear, the absence of first-person pronouns, the adoption of passive structures as a formula of detachment, and the need for an accurate time sequence to elaborate and apply each medical treatment
Sixteenth-century glosses to a fifteenth-century gynaecological treatise
Mª Victoria Domínguez-Rodríguez and Alicia Rodríguez-Álvarez
Peters, Hans. 2004. ‘The vocabulary of pain’ in Kay, Christian J. and Jeremy J. Smith (eds) Categorization in the History of English. Amsterdam and Philadelphia: John Benjamins: 193-220. Rodríguez-Álvarez, Alicia. 1996. ‘Latin for specific purposes and Latinized English in 15th-century vernacular deeds’ in Revista de Lenguas para Fines Específicos 3: 255-272. ___. 2006. ‘Oral features in Late Middle English legal texts’. Neuphilologische Mitteilungen 107(2): 187-197. Rodríguez-Álvarez, Alicia and Mª Victoria Domínguez-Rodríguez. 2005. ‘A Middle English text revised by a Renaissance reader: John Wotton’s annotations to British Library MS Sloane 249 (ff. 180v-205v)’ in International Journal of English Studies 5(2): 45-70. Simpson, Robert R. 1962. Shakespeare and Medicine. London: E. and S. Livingstone Ltd. Siraisi, Nancy G. 1997. The Clock and the Mirror: Girolamo Cardano and Renaissance Medicine. Princeton (New Jersey): Princeton University Press. Wear, Andrew. 2000. Knowledge and Practice in English Medicine, 1550-1680. Cambridge: CUP. Voigts, Linda E. and Patricia D. Kurtz. 2000. Scientific and Medical Writings in Old and Middle English: An Electronic Reference. Ann Arbor: The University of Michigan Press.
Contents Introduction Isabel Moskowich Spiegel-Fandiño and Begoña Crespo García Part 1. Linguistic aspects of early English The (im)possibility of stacking adjectives in early English Agnieszka Pysz Lists in letters: NP-lists and general extenders in early English correspondence Ruth Carroll
Part 3 Philology and the study of medieval texts
Rewriting eleventh-century English grammar and the editing of texts Donald Scragg University of Manchester
Abstract This paper looks at the current state of Old English grammar textbooks, which are largely based on Early West Saxon, and proposes their rewriting on the basis on eleventh-century Old English, which is so much more fully recorded. The paper then goes on to show how such grammars are now in a position to avail of a web-based database, recently created at Manchester, which comprises a high percentage of eleventh-century manuscript documents presented in a searchable form. Finally, a short sample text found in a number of late manuscripts is edited to demonstrate how that database can also be used to aid textual criticism.
Alistair Campbell (1959) in his Old English Grammar lays out perfectly the situation in the study of grammar of the Old English period. He says that early West Saxon ‘has come to be regarded as a grammatical norm’ (1959: 8). In fact it has been so since Henry Sweet used it in his influential readers and dictionary at the end of the nineteenth century.1 The result is that the grammar of Old English is almost invariably based on early West Saxon, with all other ‘dialects’ –if that is what we should call them– being regarded as variations on this norm. This is the case in the three major grammars of Old English: Campbell’s (which is still the most complete in English), the even more complete German one of Sievers-Brunner (1965), and the 1 These influential readers (because they were intended for students) are An AngloSaxon Reader in Prose and Verse (1876) and An Anglo-Saxon Primer (1882). Sweet also noted the consistent use of Early West Saxon spellings in the head-words of his The Student’s Dictionary of Anglo-Saxon (1896: x).
196
Donald Scragg
most recent, that of Richard Hogg (1992); and most of the shorter and introductory grammars do the same.2 As Campbell states, however, there are only four manuscripts in early West Saxon: the two early copies of King Alfred’s translation of Pope Gregory’s Cura pastoralis (BL, Cotton Tiberius B. xi, and Oxford, Bodleian Library, Hatton 20), the Parker manuscript of the Anglo-Saxon Chronicle (Cambridge, Corpus Christi College 173), which is written in one hand as far as the entry for 891 and in another up to 924, and the Lauderdale manuscript of the translation of Orosius’ History (BL, Add. 47967), written by the same scribe as Hand 2 of the Parker Chronicle. This means that early West Saxon consists of four manuscripts written by four scribes, which is not a great deal of evidence on which to base the study of a whole language. Against that we should set what used to be called late West Saxon, but which is now better called late Old English or even Standard Old English.3 If we take the period 980 to 1100, there survive to us a minimum of 275 manuscripts and documents containing English, written by around 500 scribes. This excludes a number of large manuscripts which are dated around 1100 itself, and an even larger number found in the twelfth century. By and large, all of the eleventh-century texts are written in a uniform language. To quote Kenneth Sisam (1953: 153), who established the basis of the modern study of late Old English: the early eleventh century was the period in which West Saxon was recognized all over England as the official and literary language. The York surveys of about 1030 supply a good instance in local documents from the North. The prayers to St. Dunstan and St. Ælfheah in MS. Arundel 155 give an equally striking example from Kent, for though they were certainly copied at Christ Church, Canterbury, yet they are in normal West Saxon. Dialect does break through, the more frequently as the eleventh century advances; but good West Saxon may be written anywhere in its first half.
It would seem, then, that this late West Saxon would be a much better basis for a grammar of Old English than early West Saxon which is recorded in so limited a way. So why do the grammars continue to insist on using early West Saxon as their base? 2 3
An exception is Quirk and Wrenn (1955). Cf. Gretsch (2001).
Rewriting eleventh-century English grammar
197
Quite simply, because there has been no detailed study of later West Saxon. There have been studies of particular authors and texts, a list of which up to the 1950s can be found in Campbell (1959: 365366). Late Old English vocabulary has been studied in Germany, especially in Munich, and Bruce Mitchell has produced two large volumes on Old English syntax to cover all periods. Many editions of late Old English texts, including my own of the Vercelli homilies (Scragg 1992a) include linguistic analyses, especially of sounds or spellings and inflections. But we need a synthesis, a bringing together of material of this sort, and the addition of studies of the language of texts that have been edited without benefit of linguistic notes, such as the many short texts published in periodicals. Furthermore, linguistic introductions to editions are very often concerned with the base texts. This is true of my own Vercelli homilies edition, the introduction to which concentrates entirely on the language of the Vercelli Book, and the more recent Introduction, Notes and Glossary to Ælfric’s Catholic Homilies published by Malcolm Godden (2000), which contains a very detailed glossary to the two manuscripts used for the First and Second Series respectively, BL, Royal 7 C. XII and CUL Gg 3.28, but no reference to the language of the many other manuscripts and scribes who copied all or part of the Catholic Homilies. There remains a significant gap in our knowledge: we still have no grammar of eleventh-century English, even though there are more texts and manuscripts surviving from this period than from any comparable period of English before the fourteenth century. I first drew attention to this problem in a lecture at a Harlaxton conference on the eleventh century in 1990 (Scragg 1992b), and I have been slowly working towards producing such a grammar for the last fifteen years, amongst the other various and manifold affairs of my kingdom, as King Alfred famously said.4 But the first concern is to assemble the evidence, and that is clearly beyond the scope of one individual. And so, for the last seven years, initially with the help of a large grant from the Arts and Humanities Research Board of Great Britain, a team of six at Manchester, led by my colleague the AngloSaxon palaeographer Alex Rumble and myself, have constructed an electronic database which is now public, the Inventory of Script 4
‘ongemang oðrum mislicum ond manigfealdum bisgum ðisses kynerices’, Preface to the translation of Gregory’s Cura pastoralis.
198
Donald Scragg
Categories and Spellings in Eleventh-century English, or more simply the Manchester eleventh-century database. The url address is and I would encourage everyone interested in Old English to investigate it for themselves. Over the last two years, since the AHRB-funded project ended, data has continued to be assembled thanks to a grant from the British Academy. This database has been widely advertised because the project team members have variously been talking about it at conferences and publishing papers which draw upon its data. But I need to explain its workings again here, partly because it is being expanded constantly, but also because I want to demonstrate its usefulness to editors of eleventh-century texts. First let me quote a few statistics. When I asked the AHRB for money to fund this project, I naively believed that it would be possible to assemble an inventory of scribes and spelling from the eleventh century in three years. The weight of evidence when we came to look into it overwhelmed me. Allowing for multiple copies of texts surviving in different manuscripts, there are almost 2000 in the period 980 to 1100. We decided to take 980 as our starting point because we could then include all the early manuscript copies of Ælfric, the first of which is dated around 990. Many of these 2000 texts are reasonably short, if you count an Ælfric homily as a short text. But in the Catholic Homilies alone there are 600,000 words, and when you add to that copies of longer texts, such as the Anglo-Saxon Chronicle, the Old English Bede or Ælfric’s Grammar, translations of the Old Testament and the Gospels, the Benedictine and other Rules, Gregory’s Dialogues and so on, you begin to see the scale of the task. There are twelve eleventh-century copies of Ælfric’s Grammar, each of them containing well over 25,000 words. All of this excludes marginalia and interlinear alterations made during the eleventh century to both manuscripts of this period and earlier ones such as the early West Saxon texts listed at the beginning of this essay. My own estimate now is that we have surviving from the eleventh century at least five million words in English, not different words, nor even different spellings, but words actually written in manuscripts and documents that survive. That is a huge corpus of language to integrate into a grammar.
Rewriting eleventh-century English grammar
199
We decided to limit our investigation. First, we excluded all alterations by eleventh-century scribes, and all marginal and interlinear additions. This left us with, as I noted earlier, around 275 manuscripts and documents written by around 500 scribes. If you were to add the alterations and additions, I estimate that at least another 100 scribes would be involved. If you think about it, that is an extraordinary legacy to have survived for 1000 years. Apart from Latin, how many other languages are recorded so fully by so many individuals from such an early period? As far as writing a grammar of the language is concerned, the first task is to assemble a complete catalogue of spellings and inflections. Earlier grammars associate these with dialects and with texts or authors. The concept of dialect in Old English studies is a problematic one, because we have no native informants, and therefore we have only very limited information on sounds. But we do have, as we have just seen, a huge resource of written material. So we should be thinking in terms of written dialects, the products of specific scriptoria, if these can be localised. And rather than studying the language of texts and authors, we should be looking at the language of individual scribes, who are themselves the products of specific scriptoria. This is the thinking behind the database. First, it identifies scribes by palaeography: the letter forms of each scribe in each manuscript are described or in many cases illustrated. Then the spellings of each copy of a text can be associated with the scribe who wrote it. At the same time, all the available spellings of an individual word are listed. For example if you look at spellings of the word for ‘king’, historically cyning, you will find more than 40 alternative spellings. I have chosen this word because there are so many variants: one syllable or two, or , or , and many ways of writing the consonant group . The number of occurrences of each spelling is listed in the database, together with the number of different texts in which it occurs, so that we can easily see which spelling occurs most frequently. In every case, the ‘Show Details’ button allows the user to access a full sentence context in which the word occurs, with reference to the Toronto Dictionary of Old English sense unit to allow for the sentence to be found in the text. There is a further possibility here. So far, what I have shown offers more information than you can get from the Dictionary of Old English, but not a great deal more. The statistics are fuller, for example, but the list of
200
Donald Scragg
spellings should, if the Dictionary is doing its job properly, be exhaustive. However, the Dictionary works with edited texts, which means only the text from the base manuscript of the edition chosen by the Dictionary compilers for their references is included. By and large, the Dictionary does not include variant readings. The Manchester database is designed to include all copies of every text, which you can see in collation where more than one copy of a text has been uploaded. The great advantage of the database is that it allows the user to track the occurrence of infrequent spellings, to see if they are the work of scribes known to have copied a large body of work or ones about whom we know very little, or perhaps scribes working at a particular time or in a specific location. There are other features which an editor might also find useful. As well as tracking the variants of particular words, the user can carry out a complete search of the data for alternative graphs, e.g. <eo>/, to get a list of all words containing these combinations, together with statistics, and again with the possibility of seeing the words in context. Without devoting too much time here to the details of the resource, I wish simply to highlight that it allows the user to interrogate the database in a variety of ways and is in my opinion a most important and useful electronic tool. One thing I should mention, however, is that the computer, because it has no mind, makes no distinction between an uninflected word and a word that has, for example, a final <e>. This means that the user can track inflections as well as the spellings of stressed syllables. I should also add that the database is still being populated. Currently we have around a million and a half words uploaded, in almost 100,000 different spellings and the complete surviving corpus of the work of many scribes is included, along with some part of the work of all the major scribes. The database can be said, therefore, already to contain a representative selection of the forms used in eleventh-century texts. This means that it is now possible to begin work on a grammar of eleventh-century English –of late Old English– and also that an editor of an eleventh-century text can now operate with a fuller understanding of the grammar of other eleventh-century texts when contemplating the grammar of his or her own text. The inclusion of variant readings in the database has allowed us to recover occasional spellings lost or obscured in the standard editions. Let me give just one example. In my Harlaxton lecture fifteen years ago I cited the
Rewriting eleventh-century English grammar
201
example of the word godcund ‘divine’ with for the initial (codcund) occurring seven times in three different manuscripts. The database shows that we can now add to that what must surely be a related example, godgund, which occurs twice in the Cambridge, Corpus Christi College 198 copy of Ælfric’s Catholic Homilies. Peter Clemoes’ (1997) edition of the two relevant First Series homilies does not cite these godgund instances in his list of variant readings, presumably because he assumed that they were just scribal slips. Perhaps they were. But if all such aberrant forms are similarly suppressed in all editions, how will we ever know if they are simply occasional slips (that perhaps the second reflects confusion with the in the previous syllable) or if they have greater significance (in this case perhaps some phonological significance)? Because if you put the seven instances of codcund I have collected so far with these two of godgund, it would seem that the word has a more regular instability than any grammar or dictionary at present allows. I can add one more example to underscore the point, a single instance (so far) of godspel ‘gospel’ written with for , codspel in an early eleventh-century copy of Ælfric’s Grammar. We don’t need many more examples to get the sense that something unusual is happening here. So my first plea to all editors is that they should not suppress anything. All manuscript readings should be recorded in the apparatus, and ideally those readings, even if the editor decides not to include them in the edited text, should be included in the glossary. If multiple copies survive of the text being edited, again nothing should be suppressed. We have so few copies of most Old English texts that there can be no reason not to give full variants. And I reiterate, all variant readings which contain unusual words or spellings should be included in the glossary, so that the glossary provides a complete record of the text in all the surviving copies. This will give lexicographers the opportunity to collect all examples of all words in every available spelling before they decide which to use in the Dictionary of Old English, and will give grammarians the knowledge they need to compile a complete grammar of the language. I maintain the same position with regard to inflections. Again, I take an example from my Harlaxton lecture, the occurrence of for thorn or eth in the third-person singular of present tense verbs, and the reverse spelling of thorn or eth for after <s> in the second-
202
Donald Scragg
person singular. Walter Skeat (1901) suggested 100 years ago that this was one of the characteristics of the Anglo-Norman scribe, but this cannot be the case because there are examples in the Vercelli Book which is dated around 975, and I defy anyone to show the influence of Anglo-Norman scribes in the days of King Edgar.5 There are many eleventh-century examples, and I will quote from just two manuscripts which date from before the middle of the century, hence before Anglo-Norman scribes are likely to have been commonplace. There are two examples of for thorn or eth in the Cotton Tiberius B.v version of the Wonders of the East, one in the singular weaxet ‘grows’ and one in the plural weaxat. This scribe made two marginal notes: God me helpe and God helpe minum handum. It seems to me extremely unlikely that an Anglo-Norman scribe would be writing such prayers in English. The second example, two instances of – <esð> written for <est>, is in a copy of Vercelli homily IV which was one of a series of texts entered in the margins of a copy of the Old English Bede in Cambridge, Corpus Christi College 41. We cannot know why a scribe used the margins of this manuscript to record homilies and charms but the idea that it might be the work of an Anglo-Norman seems to me far-fetched. I do not intend to argue the point further here. All I will say is that by using the Manchester database, it is possible to check how widely what may be thought to be unusual inflections occur. So far, I have drawn examples from manuscripts, but the majority of surviving manuscripts are written by more than one scribe, and it is necessary for editors to note carefully any differences that occur between the language of different scribes if more than one is involved in the writing of the text being edited. This may seem an obvious comment, but the truth is that even the most significant editions of recent times have sometimes neglected this aspect of the work. For example, no-one can doubt that Peter Clemoes’ (1997) edition of the First Series of Ælfric’s Catholic Homilies and Malcolm Godden’s (2000) of the Second Series constitute two of the most important editions of Old English texts of recent decades. Clemoes identifies three scribes in his base text, and Godden two in his. Yet apart from determining the units of work of the respective scribes, 5
More plausible is the suggestion in Brunner (1965: §357, Anm. 3) that it may be a Kentish form, but until all of the eleventh-century evidence is collected, no conclusion can be drawn.