SELIM 15.indb Teresa Marqués-Aguado, Selim 13 (2005-2006): 47—68 OLD ENGLISH PUNCTUATION REVISITED: THE CASE OF THE GOSPEL ACCORDING TO SAINT MATTHEW1 Abstract Punctuation has been traditionally neglected by scholars and editors of Old and Middle English texts due to the apparent ambiguity and lack of consistency of the system, to the extent that it is often silently modernized in contemporary editions. However, recent studies have shown that there exists certain regularity in the use of these punctuation marks. In the light of this, the aim of this paper is to offer an account of the use and function of such marks in the Old English version of The Gospel according to Saint Matthew (Cambridge University Library, MS Corpus Christi College 140). For this purpose, the analysis is organised into four levels: macrotextual, sentential, clausal and phrasal. Keywords: function, modernization, Old English, punctuation, The Gospel according to Saint Matthew. Resumen Tradicionalmente, la puntuación ha recibido escasa atención por parte de académicos y editores de textos en inglés antiguo y medio debido a la aparente ambigüedad y falta de consistencia que muestra tal sistema, hasta el punto de que se moderniza en las ediciones contemporáneas. Sin embargo, estudios recientes han mostrado que existe una cierta regularidad en el uso de los signos de puntuación. De acuerdo con esto, el objetivo del presente artículo es ofrecer un análisis de los usos y funciones de dichos signos en la versión en inglés antiguo del Evangelio según San Mateo (Cambridge University Library, MS Corpus Christi College 140). Para ello, el análisis se organiza en cuatro planos: macrotextual, oracional, frasal y sintagmático. Palabras clave: función, inglés antiguo, modernización, puntuación, Evangelio según San Mateo. INTRODUCTION Punctuation in Old and Middle English manuscripts has eluded detailed study, which could be put down to a number of attested facts, among them: a) the lack of systematization in punctuation, as practices may vary from scriptorium to scriptorium, from scribe to scribe and from text-type to text-type, so that, as Mitchell has pointed out, “each manuscript and / or text may demand individual treatment” (1980: 412), a view also shared by Heyworth when signalling the non-systematic introduction of these marks in many manuscripts (1981: 139); b) the overlapping functions of punctuation marks in 1 The present research has been funded by the Spanish Ministry of Science and Technology (grant number HUM2004-01075/FILO). This grant is hereby gratefully acknowledged. Teresa Marqués Aguado 50 Old and Middle English (Lucas 1971: 19); and c) the outstanding differences between the mediaeval and PDE punctuation systems (Zeeman 1956: 11). However, this situation changed in the 1970s after the publication of two ground-breaking articles, by Lucas (1971) and Arakelian (1975), as both concluded that punctuation in these early periods was far from haphazard, even though the latter also hinted that 100% consistency should not be expected (1975: 616). More recently, other scholars have contributed to the study of Old and Middle English punctuation, evincing the existence of certain tendencies in the use of punctuation within the same hand, i.e. particular symbols are likely to feature sentential relations (e.g. nominal, adjectival and adverbial clauses). This is the case of Heyworth (1981) or Mitchell (1980), as well as Gradon (1983), Parkes (1978), Alonso-Almeida (2002), Rodríguez- Álvarez (1999), Calle-Martín (2004), Esteban-Segura (2005) and Calle-Martín and Miranda-García (2005). Two recurrent issues in the literature on historical punctuation are invoked: its function and its modernization. As for its function, there has been a traditional opposition between the grammatical and the rhetorical ones. The first one refers to punctuation as a means to make grammatical structures explicit and, therefore, to convey the correct meaning. On the contrary, the rhetorical function implies that punctuation indicates the pauses that should be introduced when reading aloud. Lucas added a third function to this catalogue, the macro-textual one, according to which punctuation helps to clarify “the arrangement and lay-out of the text” (1971: 5). This issue of function has been dealt with by several scholars, such as Arakelian, who attributes punctuation a grammatical function (1975: 615)2, as opposed to Parkes (1992: 36) or Morgan (1952: 164), who opt for the rhetorical one. In general, the received view seems to veer towards the 2 In the 13th century, Bene of Florence argued against the possibility of punctuation being used to mark intonation or accent, although he was not against the rhetorical function of punctuation (Parkes 1992: 45). Old English punctuation revisted 51 rhetorical function, as Blake (1979: 67) or Strang (1994: 343-345) defend. Likewise, Grünberg, when analysing the West-Saxon version of the Gospels, concludes that the grammatical function should be excluded, asserting that “in considering these symbols it should be clearly borne in mind that they served to denote intonation: the Gospels were used for liturgical reading” (1967: 27)3. Nevertheless, this is not a clear-cut distinction, as some other scholars have highlighted. This is the case of Zeeman (1956: 18), Harlow (1959: 2) or Mitchell (1980: 393), who defend that it is a combination of both functions that we find in most texts. Regarding modernization, the dilemma lies not only on whether punctuation should be modernized or not4, but on how this process should be eventually carried out, given the unlikeness between the mediaeval and the PDE systems: whereas in mediaeval punctuation the rhetorical function plays an important role, in PDE punctuation is essentially syntactic (Quirk 1999: 1611; Blake 1979: 67). In this line, Mitchell offers three possibilities: “the manuscript punctuation, modern punctuation, or a compromise between the two” (1980: 388), clearly preferring the first one for scholarly audiences5. Contrary to his opinion, most modern editions have modernized punctuation without making explicit the criteria followed. A revealing example is Goolden’s edition of the Old English Apollonius of Tyre (1954). As opposed to these methods, the uses of either critical apparatuses for punctuation variants (Heyworth 1981: 155) or of functional equivalents (Alonso-Almeida 2002: 227-228; Calle-Martín 2004: 421) have been recently proposed as transparent methods to modernize manuscript punctuation. In this vein, Calle-Martín suggests that “the modern equivalent, therefore, depends on the ultimate 3 Although the use of the rhetorical function is obvious, the grammatical one should not be excluded when analysing this text, as we argue in the conclusions. 4 For instance, Blake opposes modernization (1979: 70). 5 Ronberg also concluded that when editing Renaissance literary works, texts should be presented “in accordance with the original views of rhetorical syntax, suggested so powerfully by the original punctuation” (1995: 61). Teresa Marqués Aguado 52 function of each mark of punctuation” (2004: 421), so that the classification of uses will be useful to propose modern counterparts for mediaeval punctuation (Marqués-Aguado 2005: 333-339). These counterparts will be ultimately drawn from Quirk’s description of the uses of PDE punctuation marks (1999: 1609-1639) and from Truss’ account of PDE punctuation usage (2003). In the light of this, the present study analyses the punctuation found in the West-Saxon version of the Gospel according to St. Matthew, which is found in the Corpus manuscript (folios 2r-45v), housed in the Corpus Christi College Library (Cambridge) under the reference 140. The results obtained from the study of punctuation in context will help us to deal with the function it fulfils, as well as to offer a proposal for modernization, as suggested above. METHODOLOGY In order to describe the uses of punctuation marks in context, the complete set of examples must be first obtained. For this reason, we resorted to the annotated version of the Gospel, where each item of the text was provided with the vowel-length marker so as to prevent ambiguity, as in the case of minimal pairs, such as þÕ vs. þe. The annotations, in turn, comprised not only the lemma, but also the tag (which included information as regards class and accidence) and the translation into PDE (Marqués-Aguado 2005). This version was taken as input for the OEC (Old English Concordancer Miranda-García & al. 2006), which is a software tool purposely designed to retrieve morphosyntactic information from properly annotated Old English corpora. In our case, all the instances of punctuation symbols were obtained from the application, together with a context of 5 words before and after each of them so as to determine their uses. These examples were pasted onto an Excel spreadsheet, separating the different marks to prevent confusion. Moreover, each example was split into three parts (the context before the mark, the mark itself and the five words following it) which were subsequently allocated to three cells of the same row. Finally, these examples were sorted Old English punctuation revisted 53 according to the first word after the mark so as to ease the subsequent task of classification. INVENTORY OF PUNCTUATION MARKS The symbols found in this version of the Gospel according to St. Matthew, which was probably written in the 11th century6, illustrate the system of positurae which was developed in the 8th century and was used until the 11th. With the term positurae we refer to the set of punctuation marks that progressively replaced the Latin system of distinctiones. The ultimate reason for this gradual substitution is to be associated with positurae’s univocal marking to distinguish a statement from a question and with their eye-catching value (Parkes 1992: 37). This system comprises four marks, which are found in the text under study with the exception of the punctus interrogativus: The punctus versus (;) (1,053 occurrences), which is the most common one. The punctus elevatus (ঠ) (344 occurrences)7. The punctus (.) (995 occurrences), which is sometimes placed slightly above the line of writing. In addition to these three positurae, the use of the section marker (represented here by the paraph ڧ¶ڦ) has to be noted. It may appear either in isolation, thus separating chapters, or in conjunction with the tilde (represented here by ڧ~ڦ), which is placed in the margin of the folio to highlight the presence of the former. USES AND FUNCTIONS OF PUNCTUATION MARKS Once the examples of punctuation marks have been obtained, a classification of their uses becomes essential in order to fulfil our initial objectives. For 6 This date has been suggested by Skeat (1871: vi) and Liuzza (1994: xxvi). 7 The use of the punctus elevatus might be taken as a stylistic feature in favour of the existence of two hands (or two different people inserting punctuation marks) in the text, since less than 40 examples are found from chapter 21 onwards (11.62%). Teresa Marqués Aguado 54 practical purposes, these will be grouped according to four levels (namely, macro-textual, sentential, clausal and phrasal). In each case, a commentary on the possible function of punctuation, as well as a proposal for modernization, will be offered. The numbers between brackets in front of the examples serve to identify them, the references of which include both the folio and the lines where they are found. 1) MACRO-TEXTUAL LEVEL The macro-textual level is concerned with macro-textual units, i.e. chapters, paragraphs and sense-units. The prevailing marks at this level are the punctus versus and the section marker. The latter normally co-occurs with positurae, since it is chiefly intended as a visual marker. The main uses of punctuation at this level comprise: A) TO SEPARATE CHAPTERS The punctus versus is the commonest marker to separate chapters, as shown in (1), since 25 out of the 28 chapters of this Gospel are marked off in this way8. Chapter divisions are sometimes (11 instances) additionally indicated by means of the section marker without the tilde, as illustrated in example (2)9: (1) Òà gemiltsode hÕ him. and hyra Õagan æthrÃn. and hig sďna gesÃwon. and fyligdon him; And þà hÕ genÕalƥhte hierusalem. and cďm tď bethfage tď oliuetes dĭne þà sende hÕ hys twÕgen leorningcnihtas (f.29v, 20-25) (2) tď his rípe; ¶ And tďsomne gecígydum (f.13r, 27) B) TO SEPARATE PARAGRAPHS Paragraphs in the Gospel must be understood as merely physical units, that is, they do not necessarily contain a complete thought, as in current usage 8 The end of the remaining 3 chapters is signalled by means of a punctus elevatus (f.21v, 8 – chapter 14), Ø (f..23r, 20-21 – chapter 15) and a punctus (f.45v, 9-10 – chapter 28). 9 The distribution of the section marker to separate chapters is an uneven one, as it appears in chapters 10, 11, 13, 14, 15, 16, 17, 22, 24, 26 and 27. Old English punctuation revisted 55 (Quirk 1999: 1624). Paragraphs in the text are usually indicated by the presence of the punctus versus, as seen in example (3). The exception is f.2v, 12, where the punctus is found10: (3) and hí gegaderigað hys gecorenan. of fÕower middaneardes endum of heofona hÕahnyssum oð hyra gemƥru; Leornigeað bígspell be þÃm fíctrÕowwe þonne hys twig byþ hnesce and lÕaf Ãcennede. gÕ witun ঌ sumor ys gehende (f.36r, 14-20) C) TO MARK SEMANTICALLY-INDEPENDENT SENSE-UNITS Sense or topic changes are introduced in 6 occasions by means of the section marker accompanied by the tilde, along with either the punctus (1 example) or the punctus versus (5 instances), as observed in example (4) below: (4) ~ þƥr byþ wďp and tďþa grist|bitung; Witodlíce manega synt gelaþode and fÕawa gecorene; ¶ Òà ongunnon þà pharisei rƥdan ঌ hig woldon þone hƥlend on hys sprƥce befďn (ff.32r, 27 - 32v, 4) However, the section marker is not compulsory, and in that case the punctus versus appears isolatedly (5): (5) Eornostlíce ealle cnÕoressa fram abrahame oð dauid. synd fÕowertyne cnÕoressa. and fram dauide oð babilonis gelÕorednysse fÕowertyne cnÕoressaঠ and fram babilonis gelÕorednesse oð críst. fÕowertyne cnÕoressa. Sďþlíce þus wæs crístes cnÕores; Òà þæs hƥlendes mďdor maria wæs iosepe beweddod. ƥr hí tďsomne becďmun hÕo wæs gemÕt on innoðe hæbbende. of þÃm hÃlegan gÃste; (f.2v, 8-18) Punctuation may also be used to call attention to what follows, i.e. a conclusion or explanation of the preceding fragment, which may highlight an important idea from the religious standpoint. More than half of the examples retrieved include the punctus versus, as in example (6). The section marker, along with the punctus versus, is also encountered on two occasions, as in (7): 10 In the first four chapters, all paragraphs end with a Latin inscription in a different hand. Given their probably late date of insertion, they have not been taken into consideration. Teresa Marqués Aguado 56 (6) TwÕgen bÕoþ on bedde. Ãn byð genumen and ďþer byð lƥfed; Wacigeað witodlíce forþÃm þe gÕ nyton on hwylcyre tíde Õower hlÃford cuman wyle; (f.36v, 7-10) (7) ~ ¶ ; Se þe nys mid mÕ hÕ is ongÕn mÕ. and se þe ne gaderaþ mid mÕ hÕ tďwyrpð; (f.17r, 23-25) The analysis of punctuation at the macro-textual level indicates that the different markers help to clarify the general layout of the text in large units, thus fulfilling the macro-textual function referred to by Lucas (1971: 5). As far as the equivalence with PDE punctuation marks is concerned, the OE positurae can be rendered by a full stop or by a colon when dealing with conclusions or explanations (Quirk 1999: 1621-1624). 2) SENTENTIAL LEVEL At sentential level, independent sentences as well as the relationships established between clauses are included. In this case, the function of punctuation will be dealt with at the end of the section, unlike the proposal for modernization, which is individually suggested for each use. The main uses at this level comprise: A) TO MARK INDEPENDENT SENTENCES Punctuation may be used at sentential level to mark independent sentences, both simple and complex ones, as in examples (8) and (9), respectively, where the punctus versus accomplishes this function. Notice that in example (8) the subordinate clause precedes the main one, and between them a punctus elevatus has been inserted. The other two positurae are used sometimes: for instance, the punctus may appear to signal the end of interrogative sentences (10): (8) Eornustlíce þonne ðĭ þíne ælmessan sylleঠ ne blÃwe man bǴman beforan þÕ swà líceteras dďð on gesomnunegum and on wícum ঌ hí sín geÃrwurþode fram mannum; Sďð ic secge Õow hí onfÕngon hyra mÕde; Sďþlíce þonne þĭ þíne ælmessan dďঠ nyte þín wynstre hwæt dď þín swǴþre ঌ þín ælmesse sǴ on díglum and þín fæder hit Ãgylt þÕ se þe gesyhþ on díglum; (f.8r, 1-7) Old English punctuation revisted 57 (9) Sďþlíce ic secge Õow bĭton Õower rihtwísnyss mÃre sǴ þonne þƥra wrítera and sundorhÃlgena. ne gà gÕ on heofonan ríce; GÕ gehǴrdon ঌ gecweden wæs on ealdum tídum; Ne ofslÕh þĭঠ se þe ofslíhð se byþ dďme scyldig; (f.6v, 15- 19) (10) þà embe þà endlyftan tíde hÕ ĭtÕodeঠ and funde ďþre standende. and þà sƥde hÕ; HwǴ stande gÕ hÕr eallne dæg ídele. þà cwƥdon hig forþÃm þe ĭs nÃn mann ne hǴrode; (f.28v, 4-7) In cases such as the ones described above, the stop is to be taken as the most appropriate modern counterpart (Quirk 1999: 1623). Question marks should be used for questions (Truss 2003: 141). B) TO MARK JUXTAPOSED SENTENCES Though syntactically independent, juxtaposed sentences retain semantic links between them. For this purpose, the three positurae overlap and we may encounter not only statements —see example (11)—, but also commands or questions: (11) Ealle heora worc hig dďð ঌ menn hí gesÕon; Hig tďbrƥdaþ hyra healsbæc and mƥrsiað heora rÕafa fnadu; (f.33v, 12-14) Special attention should be devoted to the connection established between the sentences constituting the genealogy of Jesus Christ, which are signalled by means of any of the three positurae—there are 40 instances—, as in (12). Similarly, there is an enumeration of miracles marked with puncti elevati (13). The beginnings of the 9 Beatitudes are highlighted with puncti versi, as observed in (14). In this example the punctus elevatus is used to mark the beginning of a subordinate clause, as we will discuss in sub-heading d): (12) Ãbia gestrǴnde Asaঠ Asa gestrǴnde iosaphath; Iosaphath gestrǴnde ioram. Ioras gestrǴnde oziam; (f.2r, 14-16) (13) blinde gesÕoþঠ healte gÃðঠ hrÕofe synt Ãclƥnsudeঠ dÕafe gehǴraþঠ (f.15r, 17- 19) (14) Õadige synt þà gÃstlícan þearfanঠ forþÃm hyra ys heofena ríce; Ôadige synt þà líðanঠ forþÃm þe hí eorðan Ãgun; (f.6r, 10-12) Teresa Marqués Aguado 58 In any case the likeliest PDE equivalents are either the stop or the semicolon, which relates semantically-linked sentences lacking connectors (Quirk 1999: 1623). Question marks should be inserted at the end of questions. C) TO MARK COORDINATE CLAUSES Coordinate clauses can be signalled by means of whichever positurae. In (15) the copulative coordinator and is preceded by a punctus versus, whereas the clauses introduced by the correlative coordinators ne… ne (16) and oþþe… oþþe (17) are marked by either a punctus or a punctus elevatus: (15) manega wítegan and rihtwíse gewilnudon þà þing tď gesÕonne þe gÕ gesÕoþ and hig ne gesÃwon; and gehǴran þà þing þe gÕ | gehǴrað. and hig ne gehǴrdon; (f.18v, 25 – 19r, 1) (16) Witodlíce ne wífiað hig. ne hig ne ceorliaþ on þÃm ƥryste (fol.33r, 4-5) (17) Ne mæg nÃn man twÃm hlÃfordum þÕowian oððe hÕ sďðlíce ƥnne hatað and ďðerne lufaþঠ oððe hÕ bið Ãnum gehǴrsum. and ďðrum ungehyrsum; (fol.8v, 22-24) and may introduce main (18) and subordinate (19) clauses when followed by an adverb or a subordinator, and it may also appear in anastrophes11 (20): (18) Òà herodes ঌ gehǴrde ðà wearð hÕ gedrÕfed and eal hierosolimwaru mid himঠ and þà gegaderode herodes ealle ealdras þƥra sÃcerda and folces wríterasঠ (f.3r, 12-15) (19) Gyf þín hand oððe þín fďt þÕ swícað. Ãceorf hyne of and Ãwurp fram þÕ; Betere þÕ ys ঌ þĭ gà wanhÃl oþþe healt tď lífe. þonne þĭ hæbbe twà handa and twÕgen fÕt. and sǴ on Õce fǴr Ãsend; And gyf þín Õage þÕ swícað Ãhola hyt ĭt 11 Anastrophe is defined in the Oxford English Dictionary as “inversion, or unusual arrangement, of the words or clauses of a sentence” and in the Diccionario de la Real Academia de la Lengua as “inversión en el orden de las palabras de una oración”. According to its Greek etymology, anastrophe refers to the practice of changing the standard element order for the sake of emphasis. Here we refer exclusively to prepositional anastrophes wherein the preposition follows the object. In this case, the preposition is termed postposition (Fakundiny 1970: 31; Mitchell 1985: 448). Old English punctuation revisted 59 and Ãwurp hyt fram þÕ; Betere þÕ ys mid Ãnum Õage on lífe tď gÃnne | þonne þĭ sí mid twÃm Ãsend on helle fǴr; (ff.25v, 22 – 26r, 1) (20) HÕ genÕalƥhte þà and hig æthrÃn. and him tď cwƥþ. Ârísað and ne ondrƥdaþ Õow (f.24v, 10-12) From these instances we may conclude that the most suitable PDE counterparts are either the comma (Quirk 1999: 1615) or no punctuation symbol at all. D) TO MARK SUBORDINATE CLAUSES The different types of subordinate clauses are associated with different punctuation marks. Thus, for instance, the punctus versus is almost exclusively used in direct speech12, although other marks are also possible in this context, as pointed out in (20). In (21), for instance, direct speech begins after a punctus versus, and it finishes with the same mark: (21) and hÕ sƥde him; Cumað æfter mÕ and ic dď ঌ gyt bÕoð manna fisceras; And hí þƥrrihte forlÕton hyra net and him fyligdon; (f.5v, 17-19) The same pointing is observed with vocative expressions which are included within direct speech and introduced by the interjections ÕalÃ, wà or lÃ. In this case, the punctus versus appears in 33 examples —as in (22)—, whereas the remaining 7 are preceded by either a punctus elevatus or a punctus. Nevertheless, there are also examples where punctuation is missing (18% of the instances), as shown in (23): (22) Òà andswarode hÕ him; Ôalà gÕ ungelÕafulle and þwǴre cnÕores hĭ lange bÕo ic mid Õow (f.25r, 5-6) (23) þà se hƥlend hyra fÃcn gehǴrde þà cwæð hÕ là lícceteras hwǴ fandige gÕ mín ætgǴwað mÕ þæs gafoles mynyt. (f.32v, 10-12) 12 As for direct speech, Warner (1982: 158) places it outside the boundaries of subordination. However, Quirk includes it in the chapter on complex sentences and examines several arguments to consider it subordinate (1999: 1020-1024). Teresa Marqués Aguado 60 Parables are also included in speeches. Of the 10 instances found, 4 begin after the verb cweþan followed by a punctus (24), whereas the other 6 are independent sentences inserted within a longer speech, wherein a punctus versus appears, as in (25). Notice that in this example, a punctus is found after gelíc, although this is the only instance: (24) HÕ rehte him þà gǴt ďþer bígspel þus cweþende. heofena ríce is geworden gelíc senepes corne (f.19v, 3-5) (25) þà hÕ funde ঌ Ãn dÕorwyrðe meregrot þà Õode hÕ and sealde eall ঌ hÕ Ãhte and bohte ঌ meregrot; Eft is heofena ríce gelíc. Ãsendum nette on þà sƥ and of ƥlcum fisccynne gadrigendum (f.20r, 11-13) In turn, most relative clauses introduced by invariable þe or by the demonstrative se, sÕo, þæt lack punctuation (26), and only some examples are preceded by the punctus (27). In (28) the end of the relative clause is marked with a punctus, maybe in an attempt to prevent confusion, owing to the repetition of the verb gŬþ. Likewise, punctuation is also absent between headless clauses (introduced by se þe and usually placed in front of the main clause) and main ones, although, again, the punctus may appear. Both possibilities are illustrated in (29): (26) Sďðlíce þà hyt ƥfen wæs cďm sum welí mann of arimathia þæs nama wæs iosep. se sylfa wæs þæs hƥlyndes leorningcnihtঠ (f.44r, 14-16) (27) Sďþlíce se þe beswícð ƥnne of ðyssum lǴtlingum. þe on mÕ gelǴfað. betere him ys ঌ Ãn cwyrnstÃn sí tď hys swǴran gecnytt. and sí besenced on sƥs grund; (f.25v, 16-19) (28) ne ongyte gÕ ঌ eall ঌ on þonne mĭþ gƥþ. gƥð on þà wÃmbe and byþ on forþgang Ãsend (f.22r, 19-20) (29) Sďþlíce se þe sÕgð hys brÕðer þĭ Ãwordena. hÕ byð geþeahte scyldig; Se þe sÕgð þĭ stunta se byþ scyldig helle fǴres; (f.6v, 20-22) Regarding adverbial clauses, the punctus is by far the most widely used mark (30), as it more than triples the instances of the punctus elevatus (31), as well as (8) above. No punctuation is also possible, as in (6): (30) Geblissa þĭ gďda þÕowa and getrǴwa. forþÃm ðe þĭ wƥre getrǴwe. ofer fÕawa (f.37v, 14-15) Old English punctuation revisted 61 (31) Gyf þĭ wylt bÕon fullfremedঠ gà and becǴp eall ঌ þĭ Ãhst and syle hyt þearfum and þonne hæfst þĭ goldhord on heofone. (f.27v, 22-24) In most cases, the PDE equivalent is the comma —especially where the subordinate precedes the main clause—, or Ø. Inverted commas (Quirk 1999: 1630-1631) are also required for direct speech. As for the specific function of punctuation at this level, we can find cases in which the grammatical function prevails, as in uses a), b) and c), and others where the rhetorical function plays an important role, as in anastrophes or the Beatitudes. Nevertheless, in subordination both functions meet. 3) CLAUSAL LEVEL At this level, punctuation is employed to signal the relations established within the clause domain, where the punctus clearly prevails over the punctus elevatus. In this case, the PDE counterpart will be offered at the end of this section, together with the discussion about the function of punctuation. The main uses listed here are the following: A) TO DISTINGUISH THE VOCATIVE EXPRESSION FROM THE REST OF THE CLAUSE Vocative expressions are distinguished from the rest of the clause by means of puncti elevati and, mostly, by puncti (32), although 6 of the 22 examples found lack punctuation, as in (22) and (23) above, and (33) here: (32) and þonne hÕ gewordyn byð gÕ gedďð hyne helle bearn. twǴfealdlícor þonne Õow; Wà Õow blindan lÃttÕowas. gÕ secgeað swÃnmn hwylc swà swereþ on temple ঌ hÕ ys nÃht (f.34r, 5-8) (33) Wà Õow bďcyras and pharisei líccetteras forþÃm. gÕ befarað sƥ and eorþan (f.34r, 2-4) In (33), forþÃm should be taken as an adverb of result and not as a conjunction of cause, because, if the latter (either forþÃm or forþÃm þe), punctuation comes first most of the times, as commented in relation to subordinate clauses. Teresa Marqués Aguado 62 B) TO MARK ELEMENTS BELONGING TO THE SAME CLAUSE The punctus (and, occasionally, the punctus elevatus) fulfils this function in the following contexts: 1. To separate different elements. For instance, in (34) and (35) it separates the VP from the NP. (34) Sďþlíce þà se hƥlend inÕode on capharnaum. þà genÕalƥhte hym. Ãn hundredes ealdor. hyne biddende and þus cweðende (f.10v, 14-17) (35) ঌ wƥre gefylled. þæs wítegan cwyde ic ÃtǴne mínne mĭþ mid bígspellum (f.19v, 14-15) (36) nĭ þín cyning þÕ cymð tďঠ gedæfte. and rít uppan tamre assene and hyre folan. (f.30r, 6-7) The last example included under this sub-heading, (36), is an instance of anastrophe (type C according to Mitchell 1985: 447) which has passed unnoticed to Bosworth, who even rearranges the word-order, rewriting it as Òín cyning cymþ to ðé (1991: 383). We assume that punctuation here averts the reading þe cymð to gedæfte where þe is a relative particle rather than a 2nd person pronoun. Actually, it clearly indicates that preposition tď does not govern adjective gedæfte, thus highlighting the preposition stranding, as well as the rhetorical and grammatical values. 2. To distinguish a long element from the subsequent one. A clear example is (37), where the punctus marks the NP: (37) Eornostlíce ealle cnÕoressa fram abrahame oð dauid. synd fÕowertyne cnÕoressa (f.2v, 8-9) 3.- To relate the two particles in correlative constructions such as fram… oþ / tď, þe… þe, and… and, ne… ne, Ãn… ďþer, and Ãn… Ãn Old English punctuation revisted 63 In spite of not being fully systematic throughout the Gospel, these constructions are rendered by means of puncti (38), excepting Ãn… Ãn, which entails the use of the punctus elevatus followed by and (39): (38) ঌ ofer Õow cume ƥlc rihtwís blďd þe wæs Ãgoten ofer eorþan. fram abelys blďde þæs rihtwísan. oð zacharias blďd barachias suna (f.34v, 21-24) (39) sÕge ঌ þÃs míne twÕgen suna sittan Ãn on þíne swíþran healfeঠ and Ãn on þíne wynstran on þínum ríce; (f.29r, 10-12) All in all, the function of punctuation at clausal level is the grammatical one, excepting vocative structures, where punctuation also signals the end of their rising tone, and could therefore be also interpreted as a rhetorical marker. As for the PDE counterparts, Ø is the most common one, though commas may be encountered with vocatives and when marking long elements (Quirk 1999: 1627-1628). 4) PHRASAL LEVEL At this level, the scope of punctuation marks comprises the different elements of a given phrase. The inventory used at this level comprises both the punctus elevatus and the punctus. Once again, both the proposal for modernization and the function will be discussed at the end of this section. The most important uses are the following ones: A) TO RELATE THE ELEMENTS BELONGING TO A NOUN PHRASE Punctuation is used to mark the relations established within one NP. For instance, the punctus elevatus is used twice in order to highlight the connection either between a determiner and the head (40), or between a noun and its noun complement. As for the punctus, it enhances the connection between the noun and either a genitive or, mostly, an apposition (41): (40) and þonne sďna finde gyt Ãneঠ assene getíggede and hyre folan mid | hyreঠ (ff.29v, 26 – 30r, 1) (41) and hí nemnað his naman. emanuhel (f.3r, 1) Teresa Marqués Aguado 64 B) TO ENUMERATE PHRASES There is a clear example of enumeration, which is a list of evil actions (hence, all of them NP’s), joined by means of puncti elevati and, especially puncti, as shown in (42): (42) Of þƥre heortan cumaþ yfle geþancasঠ mannslyhtas. unrihthƥmedu. forligru. stale. lÕase gewitnyssa. tÃllíce word þis synt þà ðing þe þone mann besmítað; Ne besmít þone mann þÕah hÕ unþwogenum handum ete. (f.22r, 22-26) C) TO MARK COORDINATE PHRASES Under this heading we include phrases coordinated by inserting both punctuation and a coordinator. These coordinators are copulative for the most part, although adversative and disjunctive ones are also present. We may refer to, for instance, the full list of the twelve apostles, whose names are connected by inserting puncti and the coordinator and (43), where enumeration is also present. Although the punctus clearly prevails, the punctus elevatus may also be found, as in (44). This tendency clearly contrasts the results rendered for copulative clauses, where any of the three positurae may appear: (43) Òis synt sďðlíce þƥra twelf apostola naman; Se forma is simon þe ys genemned pÕtrus. And Andreas hys brďðor. Iacobus zebedei. and Iohannes hys brďður. Philippus. and Bartholomeus. Thomas. and Matheus. Puplicanus and Iacobus alphei and Taddeus. Simon chananeus. and Iudas scarioth þe hyne belƥwde; (f.13v, 3-9) (44) Òĭ sďðlíce þonne ðĭ fæste smyra þín hÕafod. and þwÕah þíne ansǴne ঌ þĭ ne sǴ gesewen fram mannum fæstendeঠ Ac þínum fæder þe ys on díglumঠ and þín fæder þe gesyhð on dǴglum hit Ãgylt þÕ; (f.8v, 17-18) D) TO MARK OFF THE WORD ū Given its shortness, the word Ŭ is graphically separated from the immediate text to prevent confusion in 3 out of 10 occurrences. It is enclosed by puncti, which are a visual device employed to separate them from the surrounding Old English punctuation revisted 65 context (45), although Grünberg also suggested that these instances “possibly mark a more solemn intonation” (1967: 27-28): (45) ne gewít fram þƥre. ƥ. ƥrþÃm ealle þing gewurðan (f.6v, 9-10) With the exception of d), whose function has already been described, it is the grammatical function that prevails at this level, excepting enumerations, which may also belong to the field of rhetoric. According to this description, only enumerations imply the insertion of commas (Quirk 1999: 1619). CONCLUSIONS In the previous section, the uses of punctuation marks in the Gospel according to St. Matthew have been classified into four levels and discussed accordingly, specifying the role of each mark. Taking into consideration the information offered here, some conclusions may be drawn from the study of the punctuation system: FIRST. Particular and consistent uses have been identified, thus revealing that the use of mediaeval punctuation is far from haphazard. In this vein, punctuation symbols can be allocated to particular levels: the section marker is an exclusively macro-textual indicator and the punctus versus is virtually restricted to the macro-textual and sentential levels, whereas the punctus elevatus and the punctus frequently overlap, although the latter is more common, especially at clausal level. Therefore, symbols might be ranked according to the level where they are found. SECOND. Notwithstanding this classification of symbols and their clear consistency at macro-textual level, for instance, overlapping is still to be noted at some points, as in the above-mentioned genealogy of Jesus Christ or the examples of juxtaposition and coordination. This phenomenon reveals that consistency relates to the function and uses of punctuation symbols, and not necessarily to a particular mark, as Rodríguez-Álvarez has noted in relation to Teresa Marqués Aguado 66 15th-century legal documents (1999: 29). Nonetheless, the uses and functions outlined here for the different marks are not fully consistent in all the contexts identified, as Arakelian suggested (1975: 616). THIRD. In view of this analysis, the prevailing function of punctuation in the Gospel should be determined. On the one hand, the high number of vocative structures, instances of direct speech, etc., feature a text to be orally transmitted, and would thus indicate a prevalence of the rhetorical function. So does the use of discourse markers or the punctuation inserted in central religious tenets such the Beatitudes. On the other hand, punctuation in juxtaposition is eminently grammatical, as well as that found at phrasal level. However, the general tendency for punctuation is to mark both syntactic relations and pauses. This is the case of subordination, as discussed above. This mixture of functions suits to the main aim of the Gospel: this is a religious text aimed at Christening people, whose main access to culture was via oral transmission (rhetorical function), so that the message had to be properly organised in order to convey the orthodox doctrine of the Church (grammatical function). FOURTH. As regards modernization, functional equivalents have been sought. There is not a one-to-one relationship between an 11th-century punctuation symbol and a PDE equivalent, owing to the limited repertory of punctuation marks, which gives way to overlapping. Table 1, divided into the four levels dealt with here, summarizes the proposal for modernization. The punctuation marks for each use have been ranked according to their frequency. Old English punctuation revisted 67 Table 1. Proposal for modernization PUNCTUATION MARKS USES AND FUNCTIONS PDE COUNTERPA RT ; / ¶ To separate chapters . ; / . To separate paragraphs . ¶ / ; /. To mark semantically-independent sense- units . / : ; / ঠ / . To mark independent sentences . / ? ; / ঠ / . To mark juxtaposed sentences . / ; ; / ঠ / . To mark coordinate clauses , / Ø ; / ঠ / . To mark subordinate clauses , / “”/ Ø . / ঠ To distinguish the vocative expression from the rest of the clause , . / ঠ To mark elements belonging to the same clause , / Ø . / ঠ To relate the two particles in correlative constructions Ø ঠ / . To relate the elements belonging to an NP Ø ঠ / . To enumerate phrases , . / ঠ To mark coordinate phrases Ø / , . To mark off the word Ŭ Ø In the light of this analysis, as well as of those carried out by other scholars, more texts belonging to the Old and Middle English periods should be revised, bearing in mind that modernization, though complex, is possible, and that punctuation is not haphazardly used, but follows a relatively systematic set of principles. Teresa Marqués Aguado University of Málaga Teresa Marqués Aguado 68 REFERENCES Alonso Almeida, F. 2002: Punctuation Practice in a Late Medieval English Medical Remedybook. Folia Linguistica Historica 12. 1-2: 207-232. Arakelian, P. G. 1975: Punctuation in a Late Middle English Manuscript. Neuphilologische Mitteilungen 76: 614-624. Blake, N. F. 1979: The English Language in Medieval Literature. Methuen, London – New York. Bosworth, J. & T. N. Toller 1991[1898]: An Anglo-Saxon Dictionary. Oxford University Press, Oxford. Calle Martín, J. 2004: Punctuation Practice in a 15th-Century Arithmetical Treatise (MS Bodley 790). Neuphilologische Mitteilungen 4: 407-422. Calle Martín, J. & A. Miranda García 2005: Aspects of Punctuation in the Old English Apollonius of Tyre. Folia Linguistica Historica 26: 45-64. Esteban Segura, M. L. 2005: The Punctuation System of the West-Saxon Version of the Gospel according to Saint John. Linguistica e Filologia 21: 29-44. Fakundiny, L. 1970: The Art of Old English Verse Composition. Review of English Studies 21: 129-142. Gradon, P. 1983: Punctuation in a Middle English Sermon. In Stanley, E. G. & D. Gray eds. Five Hundred Years of Words and Sounds. A. S. Brewer, Cambridge: 39-48. Grünberg, M. 1967: The West-Saxon Gospels: A Study of the Gospel of St. Matthew with Text of the Four Gospels. Scheltema & Holkema NV, Amsterdam. Harlow, C. G. 1959: Punctuation in Some Manuscripts of Ælfric. Review of English Studies 10, 37: 1-19. Old English punctuation revisted 69 Heyworth, P. L. 1981: The Punctuation of Middle English Texts. In Heyworth, P. L. ed. Medieval Studies for J. A. W. Bennett. Clarendon Press, Oxford: 139-157. Liuzza, R. M. ed. 1994-2000: The Old English Version of the Gospels. 2 vols. Oxford University Press, Oxford. Lucas, P. J. 1971: Sense-units and the Use of Punctuation-Markers in John Capgrave’s Chronicle. Archivum Linguisticum 2: 3-24. Marqués Aguado, T. 2005. El Evangelio según San Mateo en sajón occidental: edición electrónica, estudio léxico y sistema de puntuación. Unpublished MA Dissertation. University of Málaga, Málaga. Miranda García, A., J. Calle Martín & D. Moreno Olalla 2006: The Old English Apollonius of Tyre in the light of the Old English Concordancer. In Renouf, A & A. Kehoe eds. The changing face of corpus linguistics. Rodopi, Amsterdam – New York: 81-98. Mitchell, B. 1980: The Dangers of Disguise: Old English Texts in Modern Punctuation. Review of English Studies 31, 124: 385-413. Mitchell, B. 1985: Old English Syntax. 2 vols. Clarendon Press, Oxford. Morgan, M. M. 1952: A Treatise in Cadence. The Modern Language Review 48: 156-164. Parkes, M. B. 1978: Punctuation, or Pause and Effect. In Murphy, J. J. ed. Medieval Eloquence. Studies in the Theory and Practice of Medieval Rhetoric. University of California Press, Berkeley - Los Angeles – London: 127-42. Parkes, M. B. 1992: Pause or Effect. An Introduction to the History of Punctuation in the West. Scolar Press, Hants. Quirk, R., S. Greenbaum, G. Leech & J. Svartvik 1999[1972]: A Comprehensive Grammar of Contemporary English. Longman, London. Teresa Marqués Aguado 70 Rodríguez Álvarez, A. 1999: The Role of Punctuation in 15th-Century Vernacular Deeds. Folia Linguistica Historica 19, 1-2: 25-51. Ronberg, G. 1995: They Had Their Points: Punctuation and Interpretation in English Renaissance Literature. In Jucker, A. H. ed. Historical Pragmatics: Pragmatic Development in the history of English. John Benjamins, Amsterdam – Philadelphia: 55-63. Skeat, W. W. ed. 1871-1887: The Old English Gospels in Anglo-Saxon and Northumbrian Versions. Cambridge University Press, Cambridge. Strang, B. M. H. 1994[1970]. A History of English. Routledge: London – New York. Truss, L. 2003: Eats, Shoots and Leaves. The Zero Tolerance Approach to Punctuation. Profile Books, London. Warner, A. 1982: Complementation in Middle English and the Methodology of Historical Syntax. Croom Helm, London. Zeeman, E. 1956: Punctuation in an Early Manuscript of Love’s Mirror. Review of English Studies 7. 25: 11-18. * † *