(Re)writing orality: editing the preaching of the Compileison de Dis Commandemenz

: Sermons are a type of text ideally suited for analyzing historical methods of writing orality as a rhetorical technique, illustrating the rhetorical techniques employed to invoke the spoken word on the written page and the responses that the impression of orality was intended to elicit from a sermon’s envisioned audiences. Artes praedicandi, guides to composing and performing sermons, flourished during the High Middle Ages; reading them alongside contemporary sermons allows us to explore how the advice that they gave was carried out in practice. By combining this theoretical approach with the study of the material text—looking at cues particular to the manuscripts in which these works were found, such as punctuation—we can consider how, why, and for whom these texts set out to construct themselves as oral, and perhaps learn something about the methods and intentions of writing (or, perhaps, ‘feign-ing’) orality in the process.

Here begins the fifteenth chapter of the Compilation on the Ten Commandments.It talks about the three orders of salvation.It has three paragraphs.I have previously shown you that God gives his blessing to eight kinds of people in the Gospels.And afterwards, here I will talk to you about three orders of salvation.
One of the most significant developments in the textual criticism of written orality has been the shift from considering oral elements in a text as residual evidence of its spoken composition to regarding it as a deliberately employed rhetorical strategy in which the spoken word is evoked for aesthetic or functional purposes. 1Therefore, "written" orality can refer to both orality as a means of textual composition and orality as a means of textual performance; the latter category, which evaluates how the framing of a text's delivery as oral and its reception as aural affects its interpretation, is the focus of this article.As Simon Gaunt notes, 'a [medieval] written text will often assume an audience of listeners while nonetheless invoking a written text or even specifically signalling the mediation or presence of a book' (Gaunt 2005: 124).This blending of the oral and the written complements the integral role of speech in medieval reading practices, both public and private. 2 Analyzing the ways in which a written text deliberately constructs itself as spoken-what Gaunt aptly terms a 'fictional façade of orality'-allows us to consider the significance of 'written orality' as a literary mode in which the impression of speech is as significant to a work as its literal, spoken performance (Gaunt 2005: 138).The above quotation, excerpted from a thirteenth-century Anglo-Norman treatise on the Decalogue, exemplifies a medieval compiler's deliberate engagement with fictive orality. 3While the chapter rubric firmly grounds the text on the manuscript page by dividing it into chapters and paragraphs, its opening sentence makes use of numerous rhetorical elements to frame itself as spoken: first-and second-person pronouns constructing a speaker and listener(s), markers of time linked to narrative progression ('pardevant', 'isci', 'aprés'), and verbs of speaking ('parout', 'parlerai').The first two words of the chapter syntactically juxtapose the 'je' of the compiler with the 'vus' of the audience, directly linking speaker and listener(s).Both the rubric (the written text) and the narrator (the textual voice) are further connected by the successive use of the verb parler: both exist simultaneously as written and spoken, read and heard.This compiler clearly made a deliberate decision to frame his text as oral-to write orality-by using specific rhetorical techniques, with a specific intention in mind.Outlining the methods that he used to elicit this impression of orality can, therefore, indicate functions that written orality was intended to fulfil, and suggest how contemporary audiences would have responded to orality as a literary mode.The impression of fictive orality is a feature characteristic of medieval literature of religious instruction.A significant proportion of this literature was composed in Anglo-Norman, a historically neglected field that has benefitted from a surge in academic interest over the previous three decades.Traditionally encountered almost exclusively through its romances and chansons de geste, the 'French of England' has been brought to light as a vernacular that was extensively used in texts of religious instruction that encompassed a wide range of subject matter, forms, and intended audiences.The sustained emphasis in clerical circles on the need for vernacular pastoral literature that could be disseminated to aid the spiritual reform of the laity-an emphasis whose culmination is often seen in the Fourth Lateran Council of 1215-resulted in the proliferation of such texts in the languages of Western Europe, and Anglo-Norman was no exception (Melville and Helmrath 2017;Boulton and McCann 2019).For Jocelyn Wogan-Browne and Nicholas Watson, Anglo-Norman was defined by a 'linguistic and ideological lability' that made it an ideal medium for composing and disseminating vernacular pastoralia (Watson and Wogan-Browne 2004: 40).They continue, Nearly half the items in Dean's list of 986 Anglo-Norman texts are nonhagiographic texts of religion.So far as England is concerned, during the thirteenth and early fourteenth centuries, Anglo-Norman pastoral writing deserves to be recognised as the principal medium in which the formation of the self and the politics of access to vernacular texts were thought and experienced (Watson and Wogan-Browne 2004: 41-2).
Inherent in this body of literature, the majority of which remains unpublished, is a marked emphasis on orality and the interrogation of the contemporary 'politics of access' to education and latinity in non-male and non-clerical spaces.4Claire Waters sees thirteenthcentury Anglo-Norman pastoralia as marked by the use of 'innovative … means and forms … repeatedly characterised by a turn to dialogue', with this dialogic mode facilitating audience engagement with a text in contexts far removed from its original reception (Waters 2015: 19).Writing on La vie de sainte Modwenne, a verse hagiography composed c. 1230, Tony Hunt observes that 'one of its striking features [is] a high density of "oral" elements, which are suggestive of a special relationship between its style, mode of performance, and intended audience' (T.Hunt 2005: 104).Even in genres traditionally associated with oral composition, the figures of the author/narrator and the audience(s) have been reinterpreted as rhetorical constructs that feign spoken performance rather than as evidence that a text was orally composed.Gaunt's analysis of 'fictional façade[s] of orality' centres on a reassessment of the ways in which chansons de geste-composed orally if contemporary sources are taken at face value-were written as well as transmitted by medieval scribes (Gaunt 2005: 138).Sophie Marnette similarly considers how in a related genre, the lai, 'lais construct the speakers of the narrative' in such a way that the voices of the narrator and/or characters evoke the impression of spoken performance within the seemingly written constraints of verse form (Marnette 2013: 21).Integral to these analyses is the close reading of features characteristically associated with written orality, such as the use of phatic elements, a first-person narrative voice, verbs of speech and listening, and the construction of a listening audience.Such valuable assessments of fictive orality, however, are limited by two factors.The first is that, with very few exceptions, they focus solely on verse: of the many Anglo-Norman works examined in Waters's substantial study, a relative minority are written in prose, while the studies of Gaunt, Hunt, and Marnette centre solely on verse by definition.The primary source material analyzed in Suzanne Fleischman's (1990) seminal study of the development of narrative structure and tense in medieval French works as mediations of spoken performance is almost exclusively verse romance.Therefore, the ways in which rhetorical techniques used to write orality in prose texts might differ from those used in verse texts remain comparatively unexplored.Second, the theoretical bases of these readings are grounded principally in modern critical theory.While these presentday approaches are vital to understanding the hermeneutic aspects of written orality, they carry the inevitable caveat that they cannot reflect the interpretative approaches that the contemporary composers and audiences of these texts used in the act of making meaning.
Here, therefore, a different literary form will be considered as evidence of historical techniques of writing orality: the sermon.

Written orality as genre: the medieval sermon
Although they form one of the largest corpora of surviving medieval texts, sermons generally receive significantly less attention than other genres, perhaps because they are not as easily thought of as 'literary' in the conventional sense.And yet, sermons, by definition, are a type of text whose composition and function are dependent on orality: regardless of whether they were actually performed, they must be written as if performed, and thus evoke the spoken word on the manuscript page.Sermons exemplify that it is perhaps more profitable to approach historical texts as writing orality rather than written orality.In their construction of a preacher and his audience(s), they convey the immediacy and orality that Gaunt attributes to idealized readings of chansonniers, 'an oral parole … characterised by immediacy, presence, authenticity' (Gaunt 2005: 121).While the academic study of sermons has traditionally sidestepped questions of literal performance due to the impossibility of reconstructing historical realizations of the written texts that survive, it is precisely the idealized orality inscribed in these texts that makes them suited to reading fictive orality as a mode of literary composition.This consideration complements Arnold Hunt's presentation of Early Modern sermons as 'essentially a study of hearing rather than preaching, focused not on the lone figure of the preacher but on the two-way relationship between the preacher and his audience' (A.Hunt 2010: 5).Developing the earlier work of Michael Baxandall on the role of visual art in shaping cultural receptions of literary images, Lina Bolzoni identifies sermons as a type of text that 'permits a precise reconstruction of … the cognitive frameworks [audiences] were conditioned by … They clarify the whole ensemble of interpretative frameworks intended to guide the actions of the observer' (Bolzoni 2004: 2;cf. Baxandall 1985).'Preachers words', she continues, '… are therefore an essential guide in the reconstruction of various modes of reception' (Bolzoni 2004: 2).Critical readings focused on the ways in which the narrative voice is inscribed thus circumvent what Hunt considered to be the most insurmountable obstacle to studying sermons as illustrations of written orality: 'as early modern preachers liked to point out, vox audita perit, litera scripta manet-the spoken voice perishes, and only the written word remains' (A. Hunt 2010: 8).The multimodal function of sermons as texts intended to be both read and heard is what makes them perfectly suited to analyzing how and why oral features are woven into written texts.In addition to sermons themselves, there survives an extensive body of contemporary literature dedicated specifically to the composition and preaching of sermons: the artes praedicandi. 5.Artes praedicandi offer first-hand information regarding how the compositional and the performative aspects of 'the art of preaching' were perceived by medieval preachers and their audiences.These guides functioned both as models of rhetorical composition and as performance manuals for the aspiring preacher, offering him not only techniques considered essential to writing an effective sermon but also the corresponding dramatic and elocutionary methods that would enable him to perform it successfully.Their treatment of these topics, while inevitably idealized to some degree, can still be taken as reflective of literal performance to an extent: Siegfried Wenzel points to the sheer quantity of artes praedicandi and the wide circulation of especially influential examples as evidence that, 'as their authors clearly indicated, [they] followed actual practices used by preachers' (Wenzel 2015: 1).The rhetorical and compositional techniques described in artes praedicandi were applied not only to sermons, but also to sermon-like texts: texts that articulated the same performative, didactic, and functional aspects as sermons, but were not necessarily delivered in the pulpit.Of these genres, the most significant is the compilatio.Like a sermon, a compilatio amalgamated diverse scriptural, patristic, and homiletic sources into a single work that could be either read or performed for the spiritual benefit of an audience; unlike a sermon, however, a compilatio's length was not constrained by the need of keeping churchgoers engaged, nor was it tied to the specific thema, or biblical lesson, of the day.Hic docet, a fourteenth-century ars praedicandi composed in England, observes, Quia in hiis duobus principaliter predicandi et conferendi versatur negocium, non impertinens est videndum qualiter in istis debet peruigilare studium et intencio predicantis et eciam conferentis.Vbi primo sciendum est … a modo et arte faciendi collaciones, quia quod exigitur ad collacionem requiritur ad sermonem.
Since the task of preaching and of making collations lies chiefly in [the same] things, it is not irrelevant to investigate how the preacher's as well as the collation maker's study and intention must be watchful in these matters.Hence, we must first understand … the way and technique of making collations, because what is necessary for a collation is also required for a sermon (Wenzel 2013: 148-9).
What Hic docet implies is that the compilatio and the sermon were composed to the same end and with the same rhetorical techniques, the only difference being that one was preached formally, and the other was not.The ways in which compilationes could function as works of religious instruction that were also defined by oral performance complements Waters's argument that the 'imagined dialogues' evoked in such texts 'enable a transfer not only of the content of learning, but of its process' (Waters 2015: 23-4).In reading an Anglo-Norman compilatio through the lens of both the verse literature surveyed above and the prose models of the artes praedicandi, we find a continuity of techniques and intentions in writing orality between the two seemingly disparate literatures of 'education' and 'entertainment', thus illustrating universal aspects of medieval approaches to writing orality. 6e also, furthermore, can appreciate the ways in which the form and rhetoric of the socalled 'university' or 'scholastic' sermons, defined by their latinity, erudition, and indebtedness to Scripture, were adapted to suit the needs of the diverse audiences of a vernacular equivalent (Wenzel 2015: xv).

Written text, oral mode: the Compileison de Dis Commandemenz
The Compileison de Dis Commandemenz (Compilation about the Ten Commandments, hence Dis-Commz) is the fourth of a five-part series of Anglo-Norman pastoral works composed in the second half of the thirteenth century now known as the Compileison for the religious life. 7 As the title indicates, the text is a compilatio (Anglo-Norman compileison) that is concerned with providing an exposition of the Decalogue in an accessible, vernacular form; this commentary is supplemented by three additional chapters on the Beatitudes, the three states of marriage, widowhood, and virginity, and advice for maintaining the state of virginity.The primary source of DisCommz is the Summa de virtutibus et vitiis, composed by William Peraldus (d.1271), a Dominican friar who was also known to his contemporaries as a composer of sermons. 8.Recalling the-near equivalence of the sermon and the compilatio articulated by Hic docet, DisCommz tellingly defines itself simultaneously as a 'compileison', 'escrit' ('treatise'), and 'sermon' within its first paragraph.Its audiences, likewise, are presented as both listeners and readers: in the opening sentence of each compileison, the compiler directs the text 'a tuz iceus e icelis ki lirrunt cest escrit u deuoutement de bon quer de autre lire l'orrunt' ('to all the men and women who read this text or hear it read by someone else devoutly and with a good heart'). 9This technique of inscribing a fictive audience into a sermon was one of the primary means of inviting an actual audience's textual participation.Waters sees vernacular preaching as an activity that 'implied a necessarily joint instructional undertaking between the clergy and the laity': a cleric transferred Latin doctrine into an accessible vernacular form, which a lay person then drew on 'to work toward his or her own salvation' (Waters 2015: 1).This shared experience of textual reception is the sole feature that the compiler uses to define his audience-a notable contrast to the analogous Anglo-Norman works surveyed by Waters, in which the identification of their audience(s) with particular social groups 'makes [these] discussions in the prologues of didactic works … more widely relevant than they might at first seem' (Waters 2015: 12).The duality in DisCommz's presentation as both a narrated text and a written text, as illustrated in the opening quotation of this paper, complicates this distinction between heard sermons and read sermons, in which the act of reading facilitates a move from 'the public experience of hearing sermons into a more intimate, self-directed, and textual mode' (Waters 2015: 14).The verbs that the compiler uses to indicate the reception of the work are almost always This is the most recent editorial title given to the work in Dean and Boulton (1999: 644).In this article, Compileisoncapital-C refers to the compilation of treatises in its entirety, and compileison-small-c to the textual form, corresponding to the Latin compilatio.On Peraldus' surviving sermons, see Wenzel (2017) Cambridge, Trinity College, MS R.14.7, f. 1r; virtually identical repetitions of this address recur on ff.36r, 111v, and 115r.All further references to the text of the Compileison, unless stated otherwise, will be to this manuscript.related to hearing rather than reading, thus positioning orality as the intended means of conveying the text: the most frequent verbs of communication are parler, tocher, and moustrer and those of reception entender, ecouter, and oir; every reference to a reader is invariably accompanied by one to a listener.The orality that characterizes DisCommz as a narrative text is not related to its reception as either public/plural or private/singular-an ambiguity perhaps benefitted by the polysemous use of Anglo-Norman vus as the formal secondperson singular and the second-person plural-but to its intention to impart effectively and memorably the teaching that it contains.That all these features can be analysed from the opening sentence of the text recalls Ruth Evans's highlighting of the significance of the prologue as a performative textual space.'Prologues are not', she writes, 'inert texts but cultural performances with recognisable cultural effects … opening up a space for the "theatre" of social performance' (Evans 1999: 377).Of interest here, however, is the primarily visual language that Evans (1999: 372) uses to express this reading: Designed to stand outside the works they introduce, prologues offer frames for reading those works, frames that promise the reader a certain transparency of the ensuing text.A frame demarcates the boundary between outside and inside, showcasing a visual image as well as confining it.
Here, orality is more than a rhetorical flourish or a script for performance: the voice of the text has the additional effect of universalizing its audience by means of its very accessibility, more open to readers and listeners alike than a treatise intended solely to be read ever could be.It is, therefore, unsurprising that one of the three surviving manuscripts of DisCommz, Oxford, Bodleian Library, MS Bodley 654, does not contain the entire Compileison, but rather incorporates it into a collection of Latin sermons addressed to fratres karissimi, suggestive of use in actual preaching (f.24r).What will be shown in this article is how DisCommz exemplifies the ways in which a thirteenth-century text might use specific literary devices to construct itself as orally delivered and received, and how performancewhether real or imagined-is an essential part of successfully conveying literary function and meaning of such texts.

From theory to practice: inscribing speech in artes praedicandi and DisCommz
The authors of artes praedicandi were primarily concerned with producing effective sermons that affected their audiences.Writing in the last quarter of the thirteenth century, the Dominican friar Jacobus de Fusignano clearly outlines the ways in which the intention, form, content, and effect of a successful sermon are dependent on each other: Omni operi omnique actioni non solum debetur efficiens quod opus agit, verum etiam exigitur finis quem agens intendit, materia quam agens assumit, forma quoque quam suo operi agens imponit.
Every deed and every action must not only have something efficient that carries it out, but also an end which its agent intends, a subject matter that its agent works with, and a form which he imposes on his work (Wenzel 2015: 10-11).
If a sermon is to be 'something efficient that carries out' a particular action-i.e., preaching-it requires three things: a specific goal, an appropriate subject matter, and a suitable form.If the most fundamental function of preaching a sermon was to benefit spiritually the souls of its hearers, then the most fundamental function of writing a sermon was to capture their attention by evoking this performance on the manuscript page, either to be literally reproduced in a later act of preaching or imaginatively recreated in the mind of a reader.The onus on the composer was, therefore, to create a work that was as effective when read silently as it was when read aloud.
Consequently, it is unsurprising that artes praedicandi consistently emphasize the performative and persuasive facets of the sermon: unlike other types of devotional texts that were as well suited to private contemplation as to public reception, sermons by necessity had to be produced for and evocative of performance.The necessity of seizing and maintaining an audience's engagement to the success of preaching is reflected in the repeated references that artes praedicandi make to the effects that certain rhetorical or structural techniques should (or should not) elicit in their envisioned hearers.In chapter 7 of his Libellus, Jacobus de Fusignano supplies twelve methods of elaborating on the thema of a sermon (Wenzel 2013: 38-41).While some of these techniques are concerned with content, such as including quotations from external sources or developing on the etymology of a particular word in the thema, others relate to rhetoric: devices like synonyms, similes, or apposition, he suggests, are particularly useful in fleshing out the content of a particular section of the sermon, or division (Wenzel 2013: 40-1).The author of Quamvis, an ars praedicandi contemporary with Hic docet, warns his readers against rhetorically overcomplicating their sermon divisions: 'I consider this [practice] far too intricate and superfluously confusing, and therefore I won't provide an example of it'. 10 Hic docet instructs a preacher, using a plethora of verbs related to speech, to consider whether a chosen quotation utrum dicatur ... vel per modum exhortationis seu excitantis, increpantis vel detestantis, imperantis, retrahentis vel alienantis, commendantis, alloquentis, promittentis vel terrentis, consolantis vel sub racione condonantis vel irridentis, vocantis vel exoptantis, increpantis vel auxilium postulantis, vel per alium modum quemcumque … . is said as the words of one who exhorts or incites, warns or repels, commands or withdraws or turns off, commends, addresses, promises or frightens, consoles or condones with reason or derides, calls or desires, warns or asks for help, or in whatever other way' (Wenzel 2013: 148-50).
Quotations such as these suggest that whether a sermon was received by a reader or by a hearer had little bearing on the ways in which artes praedicandi conceptualized their reception: hearing is always presented as the primary mode of reception and speaking as the mode of delivery.This is also evident in the rhetorical techniques that are most frequently recommended to composers of sermons, which Wenzel (2015, xvi) identifies as 'structural control, verbal concordance (created by the repetition of structural markers and of etymologically related words, that is, the classical figure of adnominatio), variation and decoration, the last feature including the use of rhyme and of verses'-features that clearly engage the ear as well as the eye, and work to make the subject matter of the sermon more memorable and engaging.One of the principal elements that DisCommz uses to construct itself as spoken is, like in Sainte Modwenne, the inclusion of a 'high density of phatic elements' in each sentence (T.Hunt 2005: 105).Many of these are markers of textual structure, which serve to indicate the conceptual development of the subject material for the benefit of the audience rather than the formal development of a sermon's thema integral to the artes praedicandi. 11The second paragraph of the prologue illustrates the frequency with which these repeated structural markers occur, indicated in italics (f.115v): Ore, entendez le proces de ceo sermon (c'est a dire coment nus avant irrom en cest compileison).Adeprimes vus parlerom de dis comandemenz Deu, a queus garder nus sumes tuz tenuz si nus volom sauver nos almes.E ceo vus musterom aprés par autorité e par reson.Aprés ceo parlerom de checun commandement par sei.Puis tocherom les biens ke homme receit de Deu par garder les dis commaundemenz.Enaprés parlerom des [oit] beneizuns ke Nostre Seignur doune a oit maneres de gens.Puis parlerom des treis ordres de sauvacion.Auderein vus tocherom les sis choses ke gardent homme en seinteté.Ore entendez ci e vostre quer i metez, e oez ci la lesçun ke meine tuz a salvacion ke la retienent.Now, listen to the structure of this sermon (that is to say, how we will proceed in this compilation).To begin with, we will talk to you about the ten commandments of God, to whose keeping we are all bound if we want to save our souls.
And we will show this to you after through authority and through reason.After this, we will talk about each commandment individually.Then we will touch on the rewards that a person receives from God for keeping the ten commandments.Afterwards, we will talk about the [eight] blessings that Our Lord gives to eight kinds of people.Then we will talk about the three orders of salvation.
In conclusion, we will show you the six things that keep a person in holiness.Now, pay attention here and put your heart into it, and listen here to the less that brings all those who keep it into salvation.
Here, the element of written structure so central to the sermon is reframed with reference to a mode of delivery and reception defined by the spoken word, with each structural marker followed by a verb of speaking and/or listening.The highly structed rhetorical patterning that characterizes DisCommz complements Fleischman's identification of sequence as an unmarked aspect of narrative discourse, which constructs 'a distanced, objective perspective on events that are … sequentially ordered' (emphasis mine ;Fleischman 1990: 55).It also draws attention to the concepts of 'ordre' and 'proces' that Watson and Wogan-Browne (2004: 35-59) identify as central to the Compileison's realization.While Fleischman's reading of narrative discourse through tense-aspect theory focuses primarily on the evocation of past tenses, here the emphasis is firmly on the future: the compiler introduces content to his audience that he subsequently promises to deliver.The use of language in DisCommz to create an impression of the work as aurally received is more explicit and emphatic than in contemporary Anglo-Norman pastoral works with similar functions and envisioned audiences.This can be determined if, recalling Evans's emphasis on the nature of the medieval prologue referenced above, we compare the verbs of delivery and reception in the prologue of DisCommz to those of other contemporary Anglo-Norman pastoral works.The Lumere as lais, a verse analogue that covers many of the same topics as the Compileison, likewise indicates its suitability for both lay and clerical readers and listeners from the outset as follows (Hesketh 1996: ll. 656-8): Jeo vodraie ke tutte gent En fussent tretuz amendez Ki l'averunt oi ou regardez.
I would like all people who have heard this treatise or read it to become corrected.
The form, or the formal argument, should be explained in every book because one ought to know the way in which he or she can have an awareness of the number of parts: in knowing this is more security.I have arranged the named principal parts in six books.… I have numbered the principal parts that are arranged in six books.But nevertheless, each book is divided within itself, in good faith, with chapters and sub-headings, which we indicate with rubrics.
Another widely circulated example, the Manuel dé pechez, likewise links its structure to the convenience it affords a reader (Russell 2019: 1.2.49-52): Par perografs iert destinctés ke nus mustrent divers pechés.Pur ce nul trop hastivement cet escrit lise nomeement!It is divided into paragraphs that tell us about different sins.Therefore, never read this text too quickly!The most significant phatic element used in DisCommz, however, is the use of the adverb si as an intensifier, which recurs throughout the text.It frequently appears in noun + est phrases introducing a definition or an explanation, as in 'la … chose si est' and 'li … commandement si est', or before a pronoun that relates back to the subject of a sentence: Mes ke vout estre privé ami Deu e parfitement servir e plere a Deu si se rende en religion.
But whoever wants to be a close friend of God and to serve and please God perfectly thus turns themselves to religion (f.116r).
Other significant phatic elements used in the text are 'ore' (17 times), 'c'est/ceo est a dire' (23 times), and 'donc' (20 times).Unlike in Sainte Modwenne, however, an argument cannot be made that their frequent use is partially explained by the need to meet the syllable count or rhyming pattern demanded of a verse form, as DisCommz is entirely in prose.
Another way in which DisCommz encodes itself as spoken is the high frequency of verbs of speaking on the part of the narrator and listening on the part of the audience (T.Hunt 2005: 106-7).In the excerpt from the prologue quoted above, for example, every sentence opens with some form of performative verb: parler, mustrer, toucher, and so on.In fact, insofar as the inscribed reception of the text goes, verbs of listening are almost exclusively used to articulate the process of textual reception, namely entender and oier; penser and veer make an occasional appearance.These categories of verb are frequently combined with deictic markers that ground the process of the transfer of knowledge even more firmly in the present moment of textual reception: entendez ci, oiez ci, isci devez vus entendre, and so on.These verbs of speaking and performance present the text as dynamic, invoking its real-time existence.While, like Sainte Modwenne, it contains very few formal techniques of classical rhetoric, it is difficult to see in DisCommz Hunt's assessment of Modwenne as 'soberly written, if not plain in style, making only very sparing use of literary effects' (T.Hunt 2005: 105).

Writing oralities: the narrative voice(s) in DisCommz
The first marked difference between DisCommz and the analogues to which we have been comparing it occurs in the use of the first-person narrative voice.Unlike Sainte Modwenne and the pseudepigraphic prologues of the Lumere as lais and Manuel dé pechez, the first person is used relatively rarely in DisCommz, appearing at only eight points in the entire text.Even the opening address to the audience, while framed as the personal greeting of the compiler, is voiced in the third person 'a ses treschers freres e suers' ('to all his dearest brothers and sisters').This sparing use of the first-person also markedly differs from other 'narrative' genres of Anglo-Norman literature: the lai, for example, is characterized by 'the strong presence of a narrative voice' that is used to establish the framework within which the plot of the text takes place (Marnette 2013: 26).The fact that DisCommz uses the first person far less frequently than other contemporary analogues ensures that its use is more emphatic and noticeable whenever it does occur.The first person tends to be used for emphasis, whenever the audience is being urged that a particular concept is very important and, furthermore, is paired with verbs articulating speaking or exegesis: Jeo vus lo e conseil ke vus vus gardez de totes mençonges ausi come de la mort de alme e come de la peine de enfern.(f.118v) I insist and advise you that you guard yourselves from all lies just like from the death of the soul and like from the pain of hell.
E pur ceo vus lo jeo ke vus regardez vos defautes de cors e de alme! (f.120r) And because of this I insist that you consider your sins in body and in soul!' Perhaps as a consequence of this portrayal of the narrator as an educator and moral guide, the voice of the audience is another area in which DisCommz departs from the practices seen in other contemporary Anglo-Norman texts.In examples like the Manuel and Lumere, the audience is present, but unvoiced: a textual community that participates in literary reception intellectually, but not verbally.In other examples, like Sainte Modwenne, the audience plays a more active role in directing the text, albeit one that is still 'silent', through what Hunt defines as 'the motif concerning the pleasure and concurrence of the audience': that is, Modwenne's performance will continue as long as the audience remains entertained and engaged (T.Hunt 2005: 108).In DisCommz, by contrast, the narrator not only speaks to his audience, but voices his audience, writing and speaking them into the text through question-and-answer interactions defined by verbs of speech.These instances feature in both reported and direct speech, as in these examples: Si vus demandez, donc, coment vus poez aver la vie pardurable: jeo vus di coment.(f.115r) If you're asking, then, how you can have the eternal life: I'm telling you how.
Ore poez vus penser e demander, 'quel est cest humilité, pur laquelle Deu done si haute beneiçon?'A ceo, vus dit seint Bernard issi … (f.121r ) Now, you might think and ask, 'what is this humility, for which God gives such a great gift?' To this, saint Bernard [of Clairvaux] says to you as follows … Here, the compiler weaves audience participation into the text through the act of speech, directing not only what they say, but why they might say it.The use of the present tense in these examples is another way in which these verbs of speaking ground the text more firmly in the moment of its present reading silently or aloud: the long-dead Bernard of Clairvaux is not encountered through reading his work, nor is he quoted by the compiler: he speaks directly to the audience 'ore', in their present, facilitating their engagement with the text more readily through the invocation of narrative immediacy.This quotation from Bernard of Clairvaux further demonstrates another aspect of narrative voice defining DisCommz's style: its polyvocality.The compiler's use of spoken register to dramatize and vivify his source material is evident from the opening section of DisCommz, as in his adaptation of Matthew 19:16 onwards: E Nostre Seignur en le Evangelie nus dit coment.Car un homme demanda de lui ceo ke vus demandez e dit a Jhesu Crist, sicome conte le Evangelie: 'Mestre, ke fra jeo pur aver la vie pardurable?' E li duz Jhesu li respondi e dist: 'Si vus volez a tele vie venir, gardez les dis comaundemenz.'E li noma queus des comandemanz.E il dit e respondi a Nostre Sire: 'Ces ai jeo', dit il, 'parfitement gardé de ma juvente.
And Our Lord tells us how [to earn eternal life] in the Gospels.For a man asked of him what you are asking and said to Jesus Christ, as the Gospels tell it: 'Sir, what must I do to have eternal life?'And sweet Jesus replied to him and said: 'If you want to come into such a life, keep the ten commandments.'And he listed several of the commandments.And the man said and replied to Our Lord: 'I have', he said, 'kept these perfectly since my childhood (f.116r).
As above, textual reception is presented here through a dynamic that is fundamentally one of speaking and listening, one in which the audience not only hears the text readwhether in their own voice or in someone else's-but is presented as voicing it themselves, asking questions that the voice of the narrator/preacher then resolves.The Anglo-Norman demander indicates a more active and vocal mode of enquiry than can be easily reflected in translation, occurring in contexts related not only to asking or enquiring, but also to proclamation and to summons or calls.12The compiler's use of the present tense ensures the timelessness of its performed reception, engaging the audience and ensuring that every reading or hearing of DisCommz is encoded in the text.This asking is in turn found in the Gospel passage dramatized for the audience, with the repetition of demander articulating the comparison that the compiler draws between his present audience and the wealthy man of the Gospel account.Another oral feature characteristic of DisCommz that is exemplified is the inclusion of speech tags alongside quotations from auctoritates, which are even mirrored in the Latin sources cited: Li novime commandement si est cest: 'Non desiderabis uxorem proximi tui'.Ceo est a dire, Deu dit: 'jeo vus defend ke vus homme ne coveitez nule femme, ne vus femme nul homme, hors de matrimonie e hors d'esposailles'.Car Dampnedeu dit en la Evangelie: 'ky regarde une femme par entende de coveiter la femme: il en ad fet fornication ou avoutrie en soen quer pur sa corumpné volunté mauveise, ke est contrarie a la volunté Deu'-car Deu, ki est present a checune pensee e a checune volunté, juge la volunté pur fet.'Qui viderit', inquit, 'mulierem etc.' The ninth commandment is this: 'Do not desire your neighbour's wife'.That is to say that God says, 'I command you all that you men must not desire any woman, nor you women any man, outside of marriage and outside of wedlock'.For God says in the Gospel, 'whoever looks at a woman with the intention of desiring the woman: he has then committed fornication or adultery in his heart because of his sinful, corrupted will, which is opposed to the will of God'because God, who is present at every thought and every desire, judges the intention as the act.'Whoever sees', he says, 'a woman, and so on' (f.120r).
Here, the compiler's treatment of his biblical sources for the Ninth Commandment exemplifies the techniques most frequently used in DisCommz: absolutes (rendering the Latin uxorem proximi tui as 'nule femme'), repetition, juxtaposition of synonyms ('hors de matrimonie e hors d'esposailles'), hyperbole, double (and sometimes even triple) negatives, and intensifiers (such as the emphatic 'si').
The way in which the text frames its imparting of information is not only 'oralized' with respect to the delivery of the compilation as a whole.Its treatment of auctoritates-one of the most fundamental components of a successful and well-worked sermon (Wogan-Browne, Watson, et al. 1999: 1-18)-likewise blurs the boundaries between spoken and written knowledge.The compiler's use of auctoritates and the way in which he weaves them into the text is striking: rather than decorating his narrative with the careful placement of authoritative knowledge conveyed in writing, he frames his sources as spoken in nature, using a register of speech and performance to vivify their authors to his audience.Auctoritates, whether scriptural, patristic, or contemporary, are invariably referred to as spoken rather than written, incorporating the voices of their authors into the polyvocal narrative that the treatise strives to create.In the above example, the inclusion of the speech tag 'inquit' in the Latin quotations is to my knowledge unique in vernacular pastoral texts of this period.There is only one instance in DisCommz where the author introduces a source text-a recounting of Saint Benedict's temptation excerpted from Gregory the Great's Dialogues-with the phrase 'nus lissom' ('we read') as opposed to a verb of speaking.By framing his sources as spoken rather than written, the compiler brings them closer to the audience in a sense, perhaps conscious of the educational and social gap between the academic register of Latin texts and the envisioned audiences of a vernacular treatise of religious instruction.Furthermore, DisCommz places vernacular translations of Latin quotations before the original citations from auctoritates, an inversion of precedence that is virtually unprecedented in contemporary examples of the genre.For example, the principal vernacular source of the Compileison, the Middle English guide for anchorites, Ancrene wisse, always gives the Latin quotations before their English translations, and sometimes does not even fully translate the original Latin (Millett 2005).The compiler's consistent maintenance of Anglo-Norman as the primary language of religious instruction facilitates the role of language that Nicholas Watson has written on solely with regard to later medieval works composed in Middle English: the construction of a textual community 'seen not only as a linguistic unity but constituted by that unity' (Watson 1999: 337).It can thus be argued that by presenting its mode of teaching as primarily-and almost solely-oral, the compilation places its agency more in the hands of its speaker, as both he or she and the quoted 'authority' speak simultaneously in the present tense.The text gains further power by inverting the expected hierarchy of the Latin original and the translation into the Anglo-Norman vernacular, which additionally turns the precedence traditionally given to the written word over the spoken one on its head.It is perhaps telling that the disclaiming of authority that Malcolm Parkes considered 'a constituent device' of the medieval compilatio-the compiler's assertion that he 'add [ed] no matter of his own by way of exposition' but merely excerpted and rearranged the work of others-is absent from the Compileison, whose narrator, while deferring to the truth of Holy Scripture, never denies his own presence within the text (Parkes 1991: 59-62).In this way, the compiler's use of oral techniques to combine texts within a text recalls what Sylvia Huot defines as 'polytextual reading' in the combination of texts within a book: 'a type of reading taught in devotional manuals for the laity … [in which] the reading of one text becomes a process of reading multiple "virtual" texts' (Huot 2005: 203).Polytextual reading, she continues, models '[an] associative and free-wheeling approach to the text as a gateway to literary proliferation', a dynamic process that occurs as much in the mind of the audience as in the activity of the compiler (2005: 204).But whereas Huot primarily focuses on the ways in which books of hours model a polytextual reading practice that is individual, silent, and devout, the polytextuality-and polyvocality-of DisCommz more closely corresponds with the intentions of Anglo-Norman pastoral literature as outlined by Waters: 'rather than shaping their audiences primarily as readers, these texts imagine and address them as students, as discipuli' (Waters 2015: 5).Through speaking DisCommz, its performers speak and, in speaking, simultaneously learn and teach.

Inscribing orality: the manuscript context(s) of DisCommz
Thus far, we have considered writing orality in a rhetorical sense, considering how the compiler of DisCommz manipulated aspects of language to construct his text as orally performed.But 'written' orality, by definition, simultaneously refers to a literal process: that of inscribing speech on the manuscript page.
The ways in which medieval texts encode orality markedly differ from our own.Modern techniques used to indicate the spoken word function by characterizing its divergence from the written on a visual and a verbal level, using features such as quotation marks, indented lines, and the use of non-standard forms of spelling and grammar to highlight the presence of speech on the printed page.However, these features are notably absent in medieval texts, whose manuscript conventions mean that the spoken word is more integrated intoand, therefore, more difficult to differentiate-from the written word.
The most interesting, and yet least well understood, aspect of DisCommz that evidences the techniques used by its compiler to invoke and express written orality is as relevant to the manuscript text as it is to the printed one: its punctuation.The application and function of punctuation marks potentially offers the most compelling evidence for how orality is constructed and invoked in historical texts.Medieval punctuation is ideal for this type of analysis, as medieval texts were punctuated as much for the ear as for the eye, being concerned more with the rhetorical shape of a passage than with parsing syntax as modern punctuation is today.This role of punctuation is seen to some extent in modern conventions of representing speech, with quotation marks and departures from the expected number and placement of punctuation symbols used to differentiate speech from text in even the most conservative of written documents.Palaeographers, editors, and literary critics have historically overlooked manuscript punctuation 'in the assumption that it [is] meaningless and haphazard, scattered at random through the folios of [a] manuscript' (Calle Martín and Miranda García 2005a: 95).Although there has been a gradual increase in the amount of attention devoted to manuscript punctuation, few studies have been published that attempt to offer a more general understanding of the subject as opposed to isolated studies of individual manuscripts (e.g., de la Cruz Cabanillas 2004;Alonso Almeida 2003;Arakelian 1975;Calle Martín and Miranda García 2005b).How, then, can we analyze the ways in which punctuation yields information pertinent to the oral performance of the text when the symbols themselves are so unfamiliar to us?The evidence that we can flag as indicating written speech follows the same parameters that are suggested for the modern texts above: divergence from what is expected in the number, placement, and type of punctuation marks that occur in specific passages within a text.The most comprehensive overview of medieval punctuation is Malcolm Parkes's Pause and effect: the history of medieval punctuation in the West (1992) which provides an invaluable account of the historical development of punctuation symbols and the mise-en-page of medieval manuscripts.Parkes observes that punctuation marks structure the 'grammar of legibility' of a text by organizing the manner in which a reader receives information from it in the same way that punctuation grammatically organizes written language.He describes manuscript punctuation as either 'deictic' or 'equiparative' in character.Texts punctuated with a variety of symbols that hierarchize the organization of the surface writing-by marking some pauses as longer than others, or indicating independent and subordinate clauses-can be described as 'deictic': by modifying the emphases through the placement of punctuation symbols, 'learned correctors or scribes will not produce a neutral text for the reader, but one where pauses have already been indicated in accordance with a received interpretation' (Parkes 1992: 67).'Equiparative' punctuation shapes a given passage less prescriptively, giving the reader more interpretative agency.However, Parkes almost exclusively uses Latin prose works surviving in multiple witnesses as sources: while these examples have more internal consistency in their punctuation (as Latin was a written language approaching linguistic standardization) and can be cross-referenced with other manuscripts, they do not shed any light on the punctuation of vernacular texts or texts of different genres.In the absence of large-scale studies of the application and use of medieval punctuation, this close reading of DisCommz cannot hope to offer anything more than an example illustrating the overlooked potential that punctuation has for evidence of the methods of writing orality in historical texts.As the following diplomatically transcribed excerpt from DisCommz illustrates, the visual cues to the presence of speech prompted by contemporary grammar and formatting are noticeably absent in medieval texts: Li secund commandement si est ; Non assumes nomen dei tui in vanum.E ceo est tant a dire.ke vus comand e defend dit nostre sire [;] ke vus ne iorrez pas le nun deu en vein.E countre ceo fet homme en treis maners.Ceus ke ierunt sanz mester.e par custume [;] quant lem les veut crere sanz serment.ou quant ceo nen est nul peril.ou quant nul mal nen avient de ceo ke lem ne veut crere le chose ke il dient.Iceus trespassent premerement le secund commandement.Autre manere est ; quant homme iure.e dit faus.e conferme sa menconge par serement ; en commune parole.ou en achat.ou en vente ou autrement sanz iurer sur seinz.La tierce manere si est ; quant homme a escient se pariure sur seinz.quant il dit en contre sa conscience.O quant il iure ke il fet.e ne fet pas la verite [;] mes quide ke ceo seit veir.La est il pariours.Car saver ne puet homme fors chose ke est verite.E quider puet homme de fausete [;] ke ceo seit verite.Si ieo quid ke li rei seit a loundres.e il ni est pas e ieo iure apres ke il est la pur verite [;] ieo sui pariours.E tut autre si est [;] de teus maneres de paroles.
The second commandment is this: Do not take the name of your God in vain.And this is as if to say, that I forbid and prohibit you Our Lord says: that you never swear on God's name in vain.And a person disobeys this in three ways.Those who swear without need, and habitually: whenever other people will believe them without an oath, or when there is no risk or when no harm would come to them if others do not want to believe the things that they say.These people principally disobey the second commandment.The second way is: when a person swears, and tells a lie, and affirms his or her lying under oath: in common speech, or in buying, or in selling, or otherwise without swearing by the saints.The third way is this: when a person knowingly perjures himself or herself on the saints when he or she speaks against their conscience.Or when someone swears to tell, and does not tell, the truth: but thinks that this is just.There he or she is a perjurer.For a person cannot be saved without the thing that is truth, and a person can think of a lie: that it is true.If I were to say that the king is in London, and he is not there and I swear afterwards that he is there in truth: I would be a perjurer.And so too are all: other types of such speech (f.117r-v).
In the above example, punctuation marks in square brackets reflect contemporary emendations that were not added by the original scribe.The importance that contemporary users of the Trinity manuscript of the Compileison accorded its punctuation is indicated by the significant number of emendations made to it.In DisCommz, alterations to punctuation occur more frequently than alterations to the readings of the text on a relative scale: even though the text naturally has fewer punctuation marks than it has words, the punctuation marks are more likely to be emended relative to how many of them there are in the scribe's original copy.
The extract cited above uses the two primary punctuation marks seen in medieval manuscripts: the punctus, a single point indicating a brief pause, and the punctus elevatus, a symbol shaped like an inverted semicolon that indicates a rising tone of voice.The first thing to note here is that the second hand adds several punctus elevati to this passage immediately before a short clause that he wanted to receive particular rhetorical emphasis-such as that one must never swear on the Lord's name in vain, or that swearing falsely under oath makes one a perjurer-or that juxtapose two opposing concepts for rhetorical effect: 'e quider puet homme de fausete [;] ke ceo seit verite'.This person clearly perceived the importance of the rhetorical shape of the passage as emphasizing its intended meaning and more strongly affirming it in the minds of the audience, highlighting with additional punctuation contrasts between truth and lies and the importance of when to swear and not to swear.The original punctuation of the passage also conveys a concern with emphasizing details given by the compiler: the use of punctus to separate clauses that he felt to be particularly important for the audience to grasp, such as 'quant homme iure.e dit faus.e conferme sa menconge par serement' and 'la est il parjours.'The suggestion that the punctuation practice of the manuscript is more evocative of speech than of written text is naturally difficult to assert conclusively, given the limited number of studies on the punctuation practice of medieval texts.But the placement of punctus in superfluous or unexpected places-such as placing one before the conjunction in the opening phrase of the extract, 'e ceo est tant a dire.ke vus comand e defend dit nostre sire'-could suggest that this mark had the function of indicating a change of voice in the narrative, visually and rhetorically separating the voice of God from that of the compiler and, by extension, the person delivering the text.

Conclusion
Punctuation is indubitably important to the study and realization of writing orality-and yet, the reader of this paper might have noticed that the excerpts quoted from DisCommz are presented in a modernized format, using present conventions of mise-en-page and punctuation.This paper thus concludes by inviting us to think of another overlooked context in which writing medieval orality takes place: that of textual editing.The more specialized nature of modern punctuation inevitably prescribes one interpretation of the originally inscribed delivery of medieval texts, whose use of a smaller set of punctuation symbols that were often multipurpose leaves their phrasing far more open to interpretation.It could even be argued that modern and medieval systems of punctation fulfil different functions: the former punctuates for the eye, the latter for the ear; the first is written, the second is oral.Consequently, to edit a medieval text is in a sense to rewrite orality, obscuring historical techniques and concerns of evoking speech in writing as much as illuminating them.
And yet, questions of accessibility, legibility, and purpose are as important to a modern text as to a medieval one.To produce a translation of a diplomatic version of the original Anglo-Norman would result in a form of writing so different to that which is expected that it would inevitably divert an audience's focus from the narrative itself, even when modern symbols are used in a close approximation of the medieval ones.Punctuating for the voice is so unfamiliar a use to present audiences that, even with the concession of using modern symbols in place of the medieval ones, it would produce an edition so jarring and occasionally difficult to parse that its utility to envisioned audiences would be inevitably reduced.While projects such as Histoire ancienne jusqu'à César produced by 'The values of French' ERC-funded collaboration offer parallel online editions that allow readers to tab between diplomatic, semi-diplomatic, and interpretive texts of a given version at their con-venience, this practice can be prohibitively lengthy or costly to realize in print, especially for texts with a significant number of witnesses (Gaunt, Morcos, et al. 2020).What such editions do offer us, however, is the punctuation that originally communicated their narrative voices to their audience(s), alongside an editorial interpretation of the impression that this punctuation left on a modern reader (or listener).In this way, such editions of texts like DisCommz can illustrate what Bolzoni (2004: 2) sees as the defining characteristic of sermons: 'texts that come halfway between the oral and the written and already contain within themselves indications of how they were received through the way they were recorded.In fact, those who transcribed the homilies, and thus enable us to know them, are in a way both the author's double and a representative of the audience'.And in Dis-Commz, what we are left with is one of the clearest articulations of what Waters defines as the principal motivation of Anglo-Norman pastoral texts: 'the effort to transfer interactive teaching into text, and to preserve some of the teacher's characteristics in written form' (Waters 2015: 4).