'Gyf hyr his medycyn': analysing the middle-english recipe medical discourse

"Gyf hyr ^is medycyn ": Analysing the Middle-English Recipe Medical Discourse Francisco Alonso Almeida Universidad de Las Palmas de Gran Canaria Within the historical studies of the English language, only the literary genre has been traditionally under the spotlight. However, there is a numher of texts which still remain virtually unstudied, at least, as far as their internal structure is concerned. Nevertheless, studies on this particular matter by Manfred Górlach, and Paivi Pahta have enjoyed some attention. In this paper, I present the analysis of the recipes contained in Yale MS 47 (ff 60r71v) pointing out the structure of the texts, and the different linguistic devices used in the different parts of the medical recipe. 1. The aim of the study' The aim of my study is to describe the fifteenth-century medical recipes contained in MS Yale 47. The interest of such a study is to understand the organisation of recipe texts within a larger body text, that is, how they are inserted within the overall medical treatise and how a reader can identify those recipes by means of certain linguistic elements which appear recursively in the medical recipe genre. At the same time, I shall describe a set of features which are typical of such a genre. 2. Material In this study of the recipe genre, the material used consists in 53 recipes (see Appendix) excerpted from a fifteenth-century gynaecological treatise housed at the Harvey Cushing/ John Hay Whitney Medical Library (Yale University), and catalogued as MS Yale 47 under Bond and Faye (1935-40: 62), folios 60r-71v. The manuscript constitutes a translation into English of some parts of Gilbertus Anglicus's Compendium medicinae^, written in Latin, and it deals with ' I would like to express my gratitude lo Dr Mercedes Cabrera for the painstaking work of reading earlier drafts of this paper. ^ A comprehensiva discussion of gynaecological and obstetrical manuscripts in Middle English is given by Monica Green (1992). This American scholar classifies the manuscripts Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998-1999) 49 Francisco Alonso Almeida common obstetrical diseases. By considering the significant number of extant copies, one may assume that the text circulated well beyond the mediaeval academic medical communities, and possibly into the worid of midwifes and rural practitioners. However, these copies remain untranscribed, save for MS Yale 47 —transcribed by Hallaert in 1982— and MS Sloane 2463 which was edited by Beryl Rowland in 19813. In my study of the recipes contained in MS Yale 47, I will depart from Hallaert's edition, since her transcription fails to be accurate in some particular cases. For instance, Hallaert ignores the rnarginalia contained in the manuscript and offers no transcription for them. The marginal notes are partially written in Latin, and they constitute an unavoidable component of a manuscript like this, since «they help to identify the organisation of Information in an effective way» (Alonso Almeida 1997: 65)*. In these circumstances, I shall follow the edition proposed in Alonso Almeida (1997). 3. The mediaeval medical recipe text: the state of affairs The use of recipes is quite common during the mediaeval period both as receptaría (a collection of recipes), and as part of larger academic medical treatises. Library collections in the United Kingdom and the United States of America house a large number of medieval receptaría manuscripts written in the vernacular and which according to the sources of the manuscripts, mainly the female Saleinitan physician called Trotula, and Gilbertus Anglicus's Compendium medicinae (ca. 1230). ' Rowland's edition was reviewed by Linda Voigts and Jerry Stannard in Speculum 57 (1982). They found several mistakes and omissions in her work at diffeient levéis, namely, codicological, palaeographical, lexical, and historical. In the case of Hallaert's edition, her study of the codicological side of the manuscript lacks an in-depth analysis. Moreover, she fails to regard the manuscript as a Trotula, and her edition is marred with transcription errors every other page. ' Keiser (1995; 207) points out the importance of marginal notes in mediaeval texts, since they help the readers to find Information quickly and, they highlight the ordinatio of the text. Regarding the use of marginal notes and other visual devices, see Parkes (1976), Rouse and Rouse (1979), Keiser (1991), and Alonso Almeida (forthcoming). 50 Revista de Lenguas para Fines Específicos N"'' 5 y 6 (1998 y 1999) "Gyf hyr {jis medycyn": Analysing the Middle-English Recipe Medical Discotirse mostly remain untranscribed, or even worse, some of them remain uncatalogued. As Linda Voigts (1995: 185) claims below: The neglect of fifteenth-century Middle English texts can be at least partially understood because the large majority of these texts remain unidentified in infrequently consulted fifteenth-century manuscripts. Voigts and D. Kurtz are in the process of compiling information on scientific and medical Middle English manuscripts in order to remedy this situation, and are carrying out a catalogue of incipits of Oíd and Middle English medical and scientific texts. However, and according to Schmidt (1994: 428), this catalogue of incipits would be also insufficient for the scholar working on Middle English recipes, because unfortunately, the policy foUowed in their catalogue of incipits exeludes those recipes which are not included in a «group» of recipes. Here, «group» of recipes refers to groups of, at least, three recipes written together by the same scribe; so, any «loóse» recipe will be lost. In addition, the catalogue gives only an approximate (rather than the exact) number of recipes in each group. Schmidt also disagrees with the actual policy of the Index of Middle English Prose (IMEP). For the IMEP editorial board, recipe collections should be indexed by transcribing «at least fifty words, including the first ten or twelve of each of the first three recipes. One should then transcribe the lines which conclude the coUection» (Schmidt 1994: 423-4). Again, this indexing procedure proves unsatisfactory, since the scholar consulting the Index is not provided with any information about the contents'of the recipe collections. The initial principies of the IMEP are later modified by the guidelines in Hargreaves's «Some Problems in Indexing Middle English recipes» (1981). Then, the new indexing of recipes consisted in «transcribing the first at least fifty words, including the first ten or twelve of each of the first three recipes in a given collection, as Revista de Lenguas para Fines Específicos N°* 5 y 6 (1998 y 1999) 51 Francisco Alonso Almeida well as the last one, and by noting the number of recipes in that collection» (Schmidt 1994: 424). Schmidt suggests yet another indexing procedure which consists in transcribing the first fifteen words of each recipe in a coUection, leaving the rest unstranscribed. To my knowledge, this proposal is the best ever presented for the indexing of recipes, since this method will also offer an account of the receptarium contents, and the number of recipes in each receptarium. However, the time required to accomplish such a project would be undoubtedly long! The matter which is left to decide in the issue of recipes indexing is whether recipes which form part of a medical treatise, as it is the case here, are also relevant to be indexed and recorded for further studies of the recipe genre as such. 4. What is a medical recipe text? The structure of the recipe in MS Yale 47 In order to define the medical recipe text, I shall follow this process: first, I put forward a recent definition of text in terms of a set of characteristics offered by Gorlach (1992); after that, I study the recipes so as to identify such features. Gorlach (1992) defines text-type as «a specific linguistic pattern in which formal/structural characteristics have been conventionalized in a specific culture for certain well-defined and standardized uses of language» (738). This definition of text-type draws attention to the structural side of the text and the sociological reasons which are involved in the production of such texts. If taken literally, one assumes that texts are cultural dependent, and they may present differences cross-culturally and cross-linguistically. In other words, a same text-type produced in a particular geographical área can show variations according to the time it is produced in and the audience it is addressed to. Given this definition, at the structural level, a recipe text could be defined simply as a question-response interaction, as a sort of 52 Revista de Lenguas para Fines Específicos N°* 5 y 6 (1998 y 1999) "Gyf hyr fiis medycyn": Analysing tíie Middle-English Recipe Medical Discourse dialogic structure, where the author answers a hypothetical question. Or, to put it differently, a recipe text presents a two-part división structure: the problem and the solution. According to Michael Hoey (1994: 26), the problem-solution type of text presents a problem which requires a theurapeutic solution. To exemplify this idea, let us turn to the foUowing example from the Liber de Diversis Medicinis (Odgen 1938: 31-2) S: For |5e menyson... Tak a pott of erthe & do it full of whete flour & do it in a hate owen & late it bake ais lange as brede dose. ]3an tak of t)at flour a mesure in a coppe & tak pe ^olkes of iij hen eggis & pe jewse of mynt, of bathe ilike mekill of J)e jewse & of pe ^olkes, & make a cake & bake it & ete f)er-of, bot nott ouer mekill, & it sal helpe the. For the disentery... Take an erthen pot and put it in wheat flour and put it inside a hot oven, and bake it as long as bread does. Then, take from that flour the quantity of a cup and take the yolks of three hen eggs and mint juice, 1 like much of both the juice and the yolks, and make a cake and bake it and eat thereof but not very much, and it will help you''. In the previous recipe, the problem is given in the title, that is, dysentery, visually indicated by the use of rubrication. The solution to this problem is offered after the title, that is, the medicine made up of flour, eggs, and mint juice. Admittedly, this is a very simplistic view of the structure of the recipe, and also this structure is not generic only to recipe texts; in fact, it can also be found in other instructive text types. For that reason, Gorlach (1992: 746) proposes some linguistic features, which he considers to be generic of the recipe text. Such features are the following: (1) the form of headings, (2) the length of sentences, (3) the type of verbal forms, (4) the use of possessive ' Abbreviations have been silently expanded in all examples of this paper. Word for word translation offered. Translation mine. Revista de Lenguas para Fines Específicos N°* 5 y 6 (1998 y 1999) 53 Francisco Alonso Almeida pronouns with nouns referring to ingredients and implements, (5) deletion of objects, (6) temporal sequence and adverbs used, (7) sentence complexity, and (8) loanwords or genteel diction. Of all these, I shall describe only those which are relevant for my study. I leave this until section 4 below. The sociological components in the medical recipe are related to «the language used and the audience addressed» (Carroll 1997: 49). During much of the mediaeval period, the language of science and instruction was mainly Latin, and, to some extent, Anglo-Norman''. Thus, scientific texts were mainly accesible to that minority instructed in those tongues. However, from the thirteenth century on, scientific texts (in which I include medical recipes) were undergoing a process of «Englishing» ^: After the Conquest, vernacular scientific and medical texts did not reappear in England until the thirteenth century, and then in Anglo-Norman. In the mid-fourteenth century, English-language texts began to reappear, and in the last third or quarter of the 1300s eclipsed Anglo-Norman texts. In these last decades of the fourteenth century, English-language remedy books were joined by serious reworkings in English of Latin texts and Middle English writings of some originality (Voigts 1995: 184). The vernacularisation of scientific texts constitutes an obvious response to the growing demand of practical books by the vast laity and an incipient middle-class who needed instruction at all levéis. In the case of the medical recipes, these texts were translated from La- ' For some examples of Anglo-Norman recipes, see Tony Hunt's Popular Medicine in Thirteenth-Century England: Introduction and Texts (1990). " This process of Englishing of mediaeval scientific manuscripts is favoured by the increasing demand of practical and recreational books to suit the needs of the growing middleclass: «The steady increase in literacy brought with it a steady increase in the number of people who wanted to read for recreation and profit. As a result, in the 14th and 15th centuries there was an increasing number of translations from French and Latin into English, as well as new works composed in English» (Parkes 1973: 565). 54 Revista de Lenguas para Fines Específicos N™ 5 y 6 (1998 y 1999) "Gyf hyr Jjis medycyn": Analysing tbe Middle-English Recipe Medical Discourse tin and Anglo-Norman into the vernacular, and, to judge by the impressive number of receptaría circulaüng in English during the mediaeval period in the Isles, we may claim that they were greatly appreciated by the population. The consequence of those translations, and the new compilation of recipes in the vernacular, was the disappearance of any linguistic barrier among different language communities; and, ultimately, the availability of quick theurapeutic information in the form of (small) books which could be understood by any English-speaking person. As expected, this process of Englishing of scientific and medical recipe texts was carried out with interferences from Latin and Anglo-Norman. To this, Voigts (1996: 815) adds: Popular as well as learned medicine appeared in English, often in the company of Latin, and sometimes Anglo-Norman, in texts from this early period of vernacularization. An example... is a ieaf surviving from what appears to be a late-fourteenth-century receptarium... The text on the surviving Ieaf begins with an English recipe. That recipe is foUowed by Anglo-Norman instructions, introduced by a Latin rubric, on how to make and use a healing amulet. In short, we find three languages in three lines. In the following section, I present a structural description of the recipe texts contained in MS Yale 47 in order to illustrate the pattern which underlies those texts. This done, I shall study some the linguistic elements in greater detall. 4.1. The structure of the recipes in MS Yale 47 As noted above, the recipe text presents a generic pattern which may slightly differ from recipe to recipe. The structure of the recipe is characterised by the fact that it contains a number of stages of which some deem necessary for the text to be considered a recipe. Thus, the schematic structure of the recipes in MS Yale 47 may be outlined, as follows: Revista de Lenguas para Fines Específicos N""* 5 y 6 (1998 y 1999) 55 Francisco Alonso Almeida PROBLEM SOLUTION (remedy and preparation) TITLE = sickness 4.1.1. Ingredients 4.1.1.1. Quantities 4.1.2. Preparation 4.1.3. Application: dosage, duration 4.1.4. Evaluation use and treatment By applying this pattern to the contents of a recipe from MS Yale 47 (f. 63v, 11. 237-243), we get: Title: Ingredients: Preparations: Application and treatment duration: A gode suppositores to purge wymmen Take J3e meyle of cockle and meduU yt with hony medull ... & make yt sad as paste is... wynde yt yn a soft lynnen clo{)e put {)at in hir priuy membre • And let yt lygge J)er all nyght or lenger if it nede Let US now look into each section separately. 4.1.1. Ingredients The ingredients section aims at presenting the material needed to prepare the remedy. This section is easily recognisable because the opening element is a verb in the imperative form, such as tak (1. 238), take (1. 319), let (1. 351), followed by a number of noun phrases connected by «and» (sometimes in the form of a tironian nota here represented by &), or a punctum: Tak poudre of galles • notmygge • spykanard • & of clowes... (f. 67v, 11. 544-5) Take powder of gall, mitmegs, spikenard, and claves... 56 Revista de Lenguas para Fines Específicos N"' 5 y 6 (1998 y 1999) "Gyf hyr f)is medycyn": Analysing the Middle-English Recipe Medical Discourse for it woll make hyr a noon to hafe purgacyon (f. 62v, 1. 188) sínce it will make her have a purgation at once it wyll deliuer hir of corrupt sede (f. 67r, 11. 473-474) she will get rid of her corrupt seed For the sake of clarity, I include below an example of a mediaeval medical recipe parsed as indicated above: MS Yale 47, 11. 183-188 Title Ingredients + Quantities Preparation Application: use, dosage and duration Efficacy An o|)er Take of pelettre • and pyrettre ana • 3 iij of cockull • 9 j of diagredion as mych poundre al J3es • and put ]3em yn a lytull lynyn bagge let hyr put f)e bagge as depe yn as sche may pulí it cute lyhgtly it woll make hyr a noon to hafe purgacyon To sum up, the schematic structure of the recipe may be represented in a linear diagram as foUows (under Eggins 1994: 40): Title ^ Ingredients ^ (Quantities) ^ Preparation ^ (Application *(Use) * (Dosage) * (Duration) ^ (Efficacy) Key: ^ indicates fixed order, * unordered stage, ( ) optional stage 5. Linguistic features in the mediaeval medical recipe Recipes are also characterised by the use of certain recursive linguistic ¿lements. Those linguistic features are the forerunners of modern academia writings, and they are the use of the imperative mood, the passive voice, causative verbs, relative constructions, and temporal clauses. I concéntrate on each one of them in the foUowing sections. Revista de Lenguas para Fines Específicos N™ 5 y 6 (1998 y 1999) 63 Francisco Alonso Almeida 5.1. The imperative mood The use of the imperative mood is extensive in the recipe. The reason for this lies in the fact that it allows depersonalisation which is also a common feature of scientific treatises. By resorting to the imperative, the use of personal pronouns is partially avoided. The imperative mood also favours the instructive method, since it is a simple way of listing instructions. Besides, it avoids the use of wordy expressions, which may render difficult to understand for the reader of the medical receptaría. Such a reader expects straightforward practical guidelines to help him to solve a problematic situation. 5.2. The passive voice The use of the passive voice favours the exclusión of personal constructions in the recipe, as the agent is totally removed from the sentence. Quirk (1985: 165) notes that «since the agent phrase is usually left unexpressed, the identity of the agent may be irrecoverable.» Moreover, the passive voice allovv's the focalisation of relevant information in the recipe, in the sense that, the most important elementes) is emphasised at a particular point in the recipe. This is exemplified in the following instances: tyl J)ei be well soden (f. 62r, 1. 146) till they are well soaked when JDflt sche is fallen doun (f. 66v, 1. 458) when she falls down sowe it to gedre pat ys brokon with a silken |3rede (f. 68r, 1. 558) sew together the piece that is broken with silk thread In the first example, the use of the passive voice <be well soden> allows the focalisation of <\)ti>, and so it filis the thematic position in the sentence. This position normally implies that new information to the reader is given; or it just constitutes a way to high64 Revista de Lenguas para Fines Específicos N°' 5 y 6 (1998 y 1999) "Gyf hyr pis medycyn": Analysing the Middle-English Recipe Medical Discourse light the most important information in the sentence. Similarly, in the second and third examples, the thematic positions in the sentences are filiad by <sche> and the relative pronoun <J)at>, respectively, by using the passive voice. 5.3. Causative verbs In nearly all stages of the recipe, we can find many instances of verbs with a causative meaning. The common verbs expressing this causative meaning in our recipes are: make, do and let, they are followed mostly by a bare infinitive which expresses the resultant action. Causative verbs also raise an object which obtains the benefits from that resultant action, as shown in the examples below: gyf hyr ]3is medycyn to make pe mater more abull to passe a way (f. 62r, 11. 141-142) give her this medicine to make the matter more able to pass away let hir smell stynkynge |)inges as felthe y brent (f. 66r, 1. 433) king things as burnt filth do sef)e pem in watre • and do pe woman syt in pe. watre vn to pe pappys • (f. 70r, 11. 698-699) soak them in water, and make the woman sit in the water up to her nipples In the previous examples, the causative verbs are make, let and do. In the first sentence, the causative verb make is followed by <J)e mater> which eventually receives the action of the second verb: <passe a way>. In the same vein, the object <hir> in the second example obtains the benefits of the verb <smell>. Finally, in the last statement, do is the causative verb followed by the object <|)e woman>, which performs the action of sitting in the water. Revista de Lenguas para Fines Específicos N°' 5 y 6 (1998 y 1999) 65 Francisco Alonso Almeida 5.4. Relative clauses The relative clause is another feature typical of mediaeval scientific texts. In our recipes, relative constructions are solely introduced by ^at, and they are used as a means to mark the identity of a specific element from a wide range of possibilities. There is not one single instance of a relative clause to define any difficult concept; this is due to the absence of complicated technical words in medical recipes. Instances of relative clauses are the following: Take a gourde f)at is ripe... (f, 69v, 1. 668) Take a gourd that is ripen... if hir warme wyne to drinke |)at comyn haj)e be soden yn • (f. 70r, 11. 707-708) give her warm wine in which cumin has been boiled In these examples, the relative clauses are introduced by the relative particle <J)at> which are placed immediately after their antecedents. Here, the function of the relative clauses is to mark the identity of the antecendents. In other words, the writer refers to a specific amount of <gourd> which must be ripen. In the second example, the <warme wyne> referred to is one which has been boiled together with <comyn>. There are also underlying relative constructions expressed here by the use of non-finite participle constructions: «take J)e poudre of castorye and galbanum resolued yn vynegre» (f. 66v, 1. 436). Some of these relative constructions may be also understood as an apposition where «no apposition marker is used» (Pahta & Nevanlinna 1997: 124), and whose function is to identify an element within a larger range of possibilities. «Take hertes horne wel y brent and egges schelles well y brent» (f. 65r, 11. 343-4) and «Also J)e ius of mugwort drunkert • or herbé y plastred» (f. 65r, 11. 349-350) are examples to illustrate this point. 66 Revista de Lenguas para Fines Específicos N°' 5 y 6 (1998 y 1999) "Gyf hyr ¡jis medycyn": Analysing the Middle-English Recipe Medical Discourse 5.5. Temporal clauses Temporal clauses are needed in recipe texts, as they mark the chronological sequence in the instructive process they describe. Arguably, the most common way to introduce time clauses is by means of and; though when also occurs. Anrf-temporal constructions make no structural difference with those introduced by when, or then, that is, the predícate position is filiad by a finite verbal form. During the Middle English period, it was also possible to find these anJ-constructions as non-finite clauses to indícate temporal relation of simultaneity, as already claimed by Klemola and Filppula (1992: 310-2). The examples below illustrate the use of and with a temporal meaning; often likely to be replaced by then: Take radyche • fenell • parcell • dauke • marche • |)e rotes of ]3ese and not J)e leuys • (temporal) And \>tv to take |)e leuys of rué • mugwort sauyn • calament & origanum (f. 62r, 11. 142-145) Take the roots of fennel, parsley, wild carrot, celery, the roots of these, but not the leaves. And take the leaves of rué, mugwort, savin, calamint, and orégano take f)e rote of marche • of }3e quantite of Jjy fynger all grene and larde it with pe rote of pelettre of spayne • (temporal) And syth put f)at rote in erthe A Zeyne (f. 62v, 11. 170-172) take the root of celery, completely green, of the same size of your finger and grease it with the root of pellitory of Spain. And put the root in the earth again Aftur ward take it oute and (temporal) a noynte yt with oyle of lorer or with mete oyle & (temporal) put it yn efte sonys • and (temporal) let it lye tyl sche hafe hyr purgación (f. 62v, 11. 174-176) Afterward, take it out and anoint it with the oil of laurel or with meat oil and put it in again. And let it lie until she has her purgation Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998 y 1999) 67 Francisco Alonso Almeida 6. The expression of involvement in the Yale recipes'" It is certainly difficult to imagine linguistic features showing any emotional meaning in scientific pieces. Taavitsainen (1994: 331) argües that «modern scientific discourse is often described as depersonalised, with frequent use of the "supressed person", the passive voice, and avoidance of emotive features.» However, there is already a set of «affective features» (Taavitsainen 1994: 331) in early scientific writings, namely pronouns, verbs of cognition, and some specific expressions. Personal involvement is expressed in the Yale recipes by means of the use of personal pronouns: / and you (2"'' person singular). In the case of you, the reader of the book is addressed, although it could be argued that, very often, you is potentially used to form impersonal constructions. Examples from the coUection of recipes show that this is not the case here: Take J)e leuys of rué • mugwort sauyn • calament & origanum • if |)ou hafe not ynowe of al • take J)em JDat may be goton (f. 62r, 11. 144-145) Take the leaves of rué, mugwort, savin, calamint, and orégano. If you do not have enough of all, take those that you may have. Put it yn • in like wyse ye may do with the rote of lupynes ]3at ys mych bettur (f. 62v, 11. 169-170) Put it in the same way that you do with the root of lupin. That is much better. But be for • and aftur |)ou must gif hir medycynes to make hyr hafe a purgación be for sayde (f. 66v, 11. 459-460) But, befare and after, you must give her the medicines to make her have her purgation aforementioned. The use of the first person singular / is restricted to the following statements: y lyke mykull (11. 430-431), and the textual comment as y sayde befor (11. 563-564). '"The partial formulation of this heading is taken from Taavitsainen (1994: 331). 68 Revista de Lenguas para Fines Específicos N°-' 5 y 6 (1998 y 1999) "Gyf hyr f)is medycyn": Analysing the Middle-English Recipe Medical Discourse Apart from the use of the above mentioned pronouns, there is hardly any other occurrence of any linguistic feature which may imply interpersonal, or emotive meaning. 7. Inserting the recipes in the treatise Once I have described the internal structure of the recipe, the following question is how recipes are incorporated into the framework of a larger body text. These recipes are normally inserted by means of linking thematic and pragmatic elements both in the expository text of the ginaecological treatise and in the recipe text. There is a clear thematic relationship between the gyneacological text and the recipes. The gyneacological text is rather an expository account of the diseases in women, and the ultimate reason for such exposition of diseases is to present remedies which may offer Solutions to overeóme those states of illness. For instance, in the chapter of the dropsy of the uterus (ff. 63r-63v), the disease is described as an excess of bodily fluids, and the solution to restore good health is by means of purgation. That purgation is achieved by using the suppositories prescribed in a recipe of the said chapter <A goda suppositores to purge women> (1. 237). For the sake of clarity, I present a partial analysis of one recipe showing how it links into the overall treatise by means of certain lexical Ítems: A goode fumygacion for to wakynn pem yn tyme of hir accesse / Take of pensadanum 5 • vij of galbanum • 5 xij of coost • as mych of peper xj cornes and a lytull bawme • ^o put may be stomped «stompe ptm to poudre But resolue galbanum in vynegre & cast ])cv to doo p&r to on pe coles and do hange hir hed ouer J)e smoke • pese, medycyns schal be don when sche ys in hir accesse • when pat sche is fallen doun but be for • and aftur pu must gif hir medycynes to make hyr hafe a purgación be for sayde // (f. 66v, 11. 452-460) A good fumigation to make her be awake duriiig her attack: take seven drams of hog 's fennel, twelve drains of galbanum, and also that much of Revista de Lenguas para Fines Específicos N"*" 5 y 6 (1998 y 1999) 69 Francisco Alonso Almeida costmary, twelve grains of pepper, and a little of balín that may be pounded. Pound thein to powder, but dissolve galbanum in vinegar and throw it on the coals and hang her head over the smoke. These remedies will be done when she has her attack; when she falls down. But, befare and after, you musí give her the medicines to make her have her purgation aforementioned. The title itself is an indication of cohesiveness, in the sense that it links thematically the recipe with the rest of the chapter. The word <accesse> in the title represents a lexical link between the expository text and the recipe, as it is repeated many times throughout the chapter in Unes 391 {accyse), 427 (accesse), and 429 (accesse). Another word connected to this one is the hyperonym <sekeness> which represents a lexical link between the expository text and the recipe where we can find the word <accesse>. From a pragmatic standpoint, recipes are also inserted by using endophoric referential devices. These devices «instructs the hearer/ reader to look inside the text to find what is being referred to» (Brown and Yule 1983: 199). In the recipe above, the anaphoric pronouns pem (453), hir (453), hir (457), sche (458), hir (458), sche (458), hir (459), hyr (459) are the examples found. ítems hir and sche refer back to woman (379) in the first line of the beginning of the chapter where this recipe is included. And so is ^em. However, in the case of \>em which is clearly plural, its identity could either be found back in woman (453) or in wymmen in line 1: Her be gynes pe sekenesse of wymmen. It should be noted that, very often, the scribe shifts from the singular to the plural, but he is always referring to the concept of "woman" as such. Another common strategic way of inserting the recipes is by means of conditional clauses, as could be observed in the following examples: But if this sekenesse be of colde • furst gyf hyr t)is medycyn (f. 62r, 11. 140-141) 70 Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998 y 1999) "Gyf hyr jjis medycyn": Analysing the Middle-English Recipe Medical Discourse But if this sickness is because of cold, first give her this medicine. And if sekenes be of febulnes of f)e wommanys complexión • let hir vse confortabul metes (f. 65r, 350-351) And if this sickness is because of the woman's feeble complexión, allow her to eat nice food. But and J)e greuaunce be of ony of)er encheson • make hir a baj)e (f. 70v, 11. 735-736) But if the pain is due to any other reason, make her a bath. 8. Conclusions As shown above, the medieval recipe text presents a well-defined internal structure whicli is divided into different stages which are easily identified by the reader. Underliying this structure, there is a group of certain linguistic features that appear to be recursive in a set of mediaeval texts, and can be taken as a characteristic features of the recipe text. Such linguistic elements stand as the forerunners of modern scientific discourse, namely the use of the passive voice, the relative clause, and causative verbs. Besides, it has been shown that the mediaeval medical recipe text is characterised by a high degree of depersonalisation. This paper constitutes a partial contribution to the study of mediaeval English medical recipe texts. Notwithstanding, there are elements which need furthering up, especially those pragmatic elements such as referents, or the use of cross-references in the mediaeval receptaría. The study of these features will shed light both on the process of writing medical recipes throughout history, and on the process of compiling medieval recipes. WORKS CITED Alonso Almeida, F. (forthcoming) «Finding Information Aids in Mediaeval English Medical Manuscripts: The Cases of MS Hunter 117 and MS Hunter 185.» 11''' International Selim Conference. Revista de Lenguas para Fines Específicos N'"* 5 y 6 (1998 y 1999) 71 Francisco Alonso Almeida — 1997. Edition and Study of a Middle English Gynaecological Treatise: Yale Medical Library, MS 47, jf. 60r-71v. Tesina inédita. Facultad de Filología. Universidad de Las Palmas de Gran Canaria. Bergner, H. 1995. «The Openness of Medieval Texts.» In A. H. Jucker (ed.) Historical Pragmatics: Pragmatic Development in the History of English. Amsterdam: John Benjamins Publishing Co., 37-54. Bond, W. and C. Faye. 1935-40. Censas of Medieval and Renaissance Manuscripts in the United States and Canadá. New York: H. W. Wilson. Brown, G. and G. Yule. 1983. Discourse Analysis. (Cambridge Textbooks in Linguistics) Reprinted 1996. Cambridge: Cambridge University Press. Carroll, R. 1997. Cooking Verbs in Fourteenth-Century English: A Semantic Class and Its Syntactic Behaviour. Unpublished Doctoral Dissertation. University of Oxford. Eggins, S. 1994. An Introduction to Systemic Functional Linguistics. London: Pinter. Gorlach, M. 1992. «Text-types and Language History: The Cookery Recipe.» In J. Svartvik and H. Wekker (eds.) History of Englishes. New Methods and Interpretations in Historical Linguistics. Berlin and New York: Mouton de Gruyter, 736-761. Green, M. 1992. «Obstetrical and Gynaecological Texts in Middle English.» Studies in the Age of Chaucer 14: 53-88. Hallaert, M. R. 1982. «The "Sekenesse of wymmen." A Middle English Treatise on Diseases in Women.» Scripta 8, Brussels. Hargreaves, G. 1981. «Some Problems in Indexing Middle English Recipes.» In Edwards and Pearsall (eds.) Middle English Prese: Essays on Bibliographical Problems. New York: Garland Publishing, 91-113. Hoey, M. 1994. «Signalling in Discourse: A Functional Analysis of a Common Discourse Patterns in Written and Spoken English.» In M. Coulthard (ed.) Advances in Written Text Analysis. London: Routledge, 27-46. Hunt, T. 1990. Popular Medicine in Thirteenth-Century England: Introduction and Texts. Cambridge: D. S. Brewer. Jones, M. C. (forthcoming) «Formula and Formulation: "Efficacy Phrases" in Medieval English Medical Manuscripts.» Neuphilologische Mitteilungen. — 1997. An Analysis of the Language and Style of a Late-Medieval Medical Recipe Book: Glasgow University Library MS Hunter 117. Dissertation. Glasgow University Library. Keiser, G. (1991). «Ordinatio in the MSS of John Lydgate's Lyf of Our Lady. Its valué for the Reader, Its Challenge for the Modern Editor.» In T. W. 72 Revista de Lenguas para Fines Específicos N" 5 y 6 (1998 y 1999) "Gyf hyr jjis medycyn": Analysing the Middle-English Recipe Medical Discourse morter with gotys talowhe • aftur cast J)er to a lytul pyche and frye hem wel to gedur • so make a plastre JDer of and ley yt vp on a pece of Icper • and ley it warme to pe wombe • Recipe nr. 33: 594-598 Or take pe erthe be for a bestys mawnger pat ys troden & be pyssed with bestes • heete J)at gerthe a geynes pe fyre and ley it wher pe greuaunce ys • And let pem absteyne J)em from metes J)at ben wyndy • As froo pesys • benys • fycches • fryte • and rawhe erbys§§ Recipe nr. 34: 629-632 When pe postem ys full waxen • make plastres to make yt rype & nesch to breke As J)us make a plastre of whete melé • and lynsed • or of fenygrekes meduld with lye • but let not pe lye be ful stronge • Recipe nr. 35: 632-633 OJ)er medull with syche lye • barlye melé and make per of a plastre • Recipe nr. 36: 636-638 And pe by gynnynge of a colde postem • make plastres of origane • ysope • centorye • rué • & celydoyne • y soden yn wyne and ley it to all warme • Recipe nr. 37: 640-641 Or make a plastre with snayles with hony and whete melé soden to gedur • Recipe nr. 38: 641-643 Or make a plastre of pe rotes of affodyll stamped with reysinges and dry fyges • and soden yn oyle with lye • wyne and oyle Recipe nr. 39: 643-645 • Or make a plastre of sowrdowhe soden yn oyle • or of rué or of fenell soden in wyne with oyle or fenell seede so ysod • Recipe nr. 40: 653-654 Take J)en methe warmode and do |3er to iij partes of dispumed hony and let hir receyue pe weyjte of a pounde J)er of J)orogh Recipe nr. 41: 668-674 Take a gourde J)at is ripe and pare hym with oute & cíense hym of pat put ys with yn hym • pen stampe hym small • and put yt in a pot ouer pe fyre • with oyle and wax • cast Jjer to pe poudre of mastix and schepis talogh • and when J)ey wel soden to gedur • cast f)er to oli banum and let J)em boyle to gedur Revista de Lenguas para Fines Específicos N™ 5 y 6 (1998 y 1999) 79 ' Francisco Alonso Almeida f \.\ well ptn cíense it porogh a cloJ)e • And J3er wíth a noynte it with yn forth • And f)is oynement ys gode for brennynge and sóaldyng Recipe nr. 42: 686-691 ., And herfor take pyllyall • origanum • and bay ieuys • calamynt • & hokkes • sethe ]3em & |3er with wasche liir wombe fr'oin \)e. nauyll doun wart to Ipe pryuey membre • Jjen take cloys • and spycanárd • notmyges • and galyngale • and make liir of hem a fumygacion be nepe forpfe Recipe nr. 43: 692-695 ' ! Take rué & mugwort • & camphory • stampd tiem wel to gedre yn a morter with oyle of pyliall • and hete yt at pe fyre • and wrappe it wel with a clojDe from pe nauyl dounwart • '• Recipe nr. 44: 695-699 Also for ache pat comeJ)e of hardnesse of pe moder • Take saxifrage & ground swyly • olde caule • mugwort • hokkes betayn • and do se|)e J)em in watre • and do pe woman syt in pe watre vn to JDe pappys • Recipe nr. 45: 704-708 ffor J)at take pe croppys of ellern stampe JDem & wring out pe ius • with J)at ius & melé and with eyren make Jjynne cakys * íand fry ]3em with fresche grece • and Jjat schall cese pe ache • & if hir warme iwyne to drinke JDat comyn hajíe be soden yn • §§ • Recipe nr. 46: 736-741 l make hir a ba]3e of hokkes • fenygreke • lynseed' • warmot • so]3ern wode • & • mugwort • soden in watre • and doo hir bape ¡Hir {)er yn a gode while • And when sche comep oute of pe baf)e let a noynt hir from pe nauyll doun wart with buttur and with deuty • & with arogon •, & be hynde • also • |3en make hir a fumygacion of spynacard & of pe rotes of c'oste • Recipe nr. 47: 743-745 Also when sche comef) oute of pe baj)e and ¿'che be ryche gif hir 3 S '^ of bawme with wyne warmed • Recipe nr. 48: 743-745 if yt be a por woman sej)e pe rotes of cooste • & mugwort in wyne & do J)er to 3 ij of boles gall • & let hir drinke pat when sche comeJ)e oute of pe baf)e '^ This symbol stands for the Latin «semi» "half" 80 Revista de Lenguas para Fines Específicos N"' 5 y 6 (1998 y 1999) "Gyf hyr J3is medycyn": Analysing the Middle-English Recipe Medical Discourse Recipe nr. 49: 746 Or elles temper borage • with wyne • & gif hir JDat to drinke • Recipe nr. 50: 746-749 or make a plastre of mugwort soden in watre • from ]3e nauyll to hir priuy membre • ffor J)at make]3e a woman soné to be delyuerd of child • Recipe nr. 51: 797-800 Also take pe ius of percely and of leke • medel jDem with oyle of pulyall & gif hir to drinke • or J)e ius of Borage & J3ese wyll drawe oute |)e Seconndyne§§ Recipe nr. 52: 815-817 And if J)ei be yonge wymmen pen do sepe karlokkes • or skyrwyttes • in wyne & let hir syt ouer |)e smoke per of put it come to hir priuey membre • Recipe nr. 53: 817-820 Or take pulyall & make poudre per of and put yt in a bagge so brod and so longe pat it may hyll boJ)e pe priuey membre of pe woman & ley it to & bynde yt well {)at yt fall not A wey §§§ Revista de Lenguas para Fines Específicos N"'* 5 y 6 (1998 y 1999) 81