The Heroic Age

A Journal of Early Medieval Northwestern Europe

Founded 1998   |   ISSN 1526-1867

Issue 13 (August 2010)  |   Issue Editor: Larry Swain

Doing Philology 2: Something 'Old,' Something 'New': Material Philology and the Recovery of the Past

Michael D.C. Drout, Wheaton College, Massachusetts and Scott Kleinman, California State University, Northridge

© 2010 by Michael D. C. Drout and Scott Kleinman. All rights reserved. This edition copyright © 2010 by The Heroic Age. All rights reserved.

Abstract: This is the second of a series of columns on philology.  Christine Franzen's work on the Tremulous Hand of Worcester shows that when "material philology" and a concern for manuscripts and variants is supported by traditional "old philology," important new knowledge about past cultures can be recovered.

§1.  In an issue dedicated to the uses and abuses of manuscripts, it seems appropriate to make the topic of our second column the convoluted relationship between philology and manuscript study.  Our first column highlighted a debatable phrase in the Beowulf manuscript, "mere wio ingasmilts ungyfeðe," and demonstrated how philological methods enabled scholars to reconstruct with confidence an original reading and meaning, "Merewioingas milts ungyfeðe" (the favour of the Merovingian has been denied), which the scribe had garbled, most likely because he himself could not understand it.  This approach to a problematic phrase is fully in the tradition of the "old" philology, in which accumulated knowledge about language, and language change, is applied to enigmatic texts for the purpose of reconstructing their original forms and, hence, deciphering them.  This approach, it is both fair and sad to say, is regarded by many contemporary medievalists, if they think of philology at all, as being old fashioned, while a discussion of the un-interpretable words in their physical, manuscript context would probably be perceived by many as cutting-edge work: sound changes are "old" and manuscript study is "new," and often, even in medieval studies, what is "new" is seen as being superior to, and taking the place of, what is "old."  We think this dichotomous approach to philology is wrong.  In this column, we will explore how the "old," language-based philology and a "new," manuscript-based philology can enrich each other, and we argue that approaches which ignore either end of this spectrum are likely to miss meaningful insights.

The "New Philology"

§2.  It is twenty years since the contributors to a 1990 issue of Speculum proclaimed the "New Philology", and made manuscript study its centerpiece.  For the New Philologists, the recovery of an "original" reading inappropriately privileges a modern notion of authorial authenticity, dismisses the act of the scribe as "non-authorial," and betrays an urge to correct the manuscript, which is itself an "authorial" intrusion of the modern editor.  Desiring a more theoretically-informed philological praxis, suitably influenced by postmodern developments in contemporary culture, the New Philologists were acutely aware that the theoretical basis of philology was still firmly grounded in the nineteenth century.  They were deeply disturbed by (and, in typical postmodern fashion, wished to disturb) the complacency with which medieval scholars adhered to a nineteenth-century model of editing. In this model multiple manuscripts were compared, and, on the basis of their variant readings, arranged into hierarchical stemmatic diagrams, pointing to a hypothetical original, which was then treated as the basis for interpretation.  Instead, the New Philologists argued, textual meaning should be studied using the forms in the manuscripts themselves.

§3.  At the heart of the New Philologists’ plan for a revision of philology was a focus on the study of manuscript variation.  Drawing on the ideas of medieval French scholars Paul Zumthor and Bernard Cerquiglini, New Philologists saw the medieval text as inherently unstable.1  Lacking a codified system of writing, medieval textual culture was characterized by idiosyncrasy and multiplicity born of its fundamentally oral nature.  Even where not actually transmitted orally, medieval texts were seen to lack the systematic consistency that earlier generations of philologists attempted to impose on them.  Once a reading closest to the text’s origins was chosen, remaining manuscript readings were dismissed as mistakes of oral transmission or scribal corruption, and the diverse conditions which had given rise to the manuscript forms were supposedly given no further thought.  The New Philologists argued that the pursuit of an ideal original form, particularly by the editors of medieval French texts, had over-determined not only the forms of the texts we study but also our ideas about medieval authorship.  Cerquiglini demonstrates this point with a passage from Villehardouin’s Conquest of Constantinople.2  The standard edition by Edmond Faral follows the text of MS O, giving variant readings from MSS BCDE in the bottom margin.  The reader is at best discouraged from considering the variant readings.  None of the variants is grammatically remarkable on its own or substantially changes the meaning of the passage; but a (reconstituted) version as found in, say, MSS CDE, when placed alongside the version found in MS O, displays important rhetorical differences:

MS O: Icil Alexis si prist son frere l’empereor, si li traist les iaulz de la teste et se fist empereor en tel traïson con vos avez oï.  (This Alexis, [he] took his brother the emperor; [he] pulled his eyes out of his head and made himself emperor through the treachery you have just heard about.)

MSS CDE: Cil prist son frere empereor, et li traist les iaulz de la teste et se fist empereor en tel traïson com oés.  (He took his brother, and [he] pulled his eyes out of his head and made himself emperor through the treachery you’re hearing about.)

§4.  As Fleischman points out, the O version contrasts with the others in foregrounding Alexis, and in conveying the action through two independent si clauses, while the CDE version uses the present tense to convey an immediacy between the narrator and the audience.  For the New Philology, the fundamentally medieval quality of literary works comes from the different rhetorical strategies which arise from variant readings of a text.  In medieval literature, discursive features are always emergent, always developing within the unique conditions under which individual manuscripts were produced.  Furthermore, the reduction of manuscript variations to citations in a critical apparatus removes them from their manuscript context, the material situation which produced them and conditions their meaning.  For the New Philologists, our understanding of the medieval world should therefore come not from an idealized "original" form but from a consideration of the "idiosyncrasies" we find in different manuscripts of a text, because these contain the true "authorial" acts in medieval literature.  As Cerquiglini puts it, "l’écriture médiévale ne produit pas de variantes, elle est variance" (medieval writing does not produce variants, it is variance) (1989, 111).  According to this vision, the scope of philological enquiry and insight should be comparative, not archeological.  The New Philologists asserted that traditional philological methods were particularly inadequate for the analysis of discourse.  Although they did not call for a complete rejection of the tools assembled by previous generations, how much they felt that such tools could lapse into obscurity was generally left unclear (perhaps for strategic or rhetorical reasons) in the Speculum "New Philology" issue.

§5.  Scholarship over the past two decades has shown that the New Philologists' critique of the old philology was in many ways over-stated.  Although philological theory may have in 1990 been largely frozen in the methods of nineteenth-century editors, its practice was by the second half of the twentieth century considerably more sophisticated.3  That significant insight could be derived from studying manuscript forms themselves was nothing new—indeed the study of medieval dialects relied on it—and the comparative study of manuscript variations would not be appropriate where only a single manuscript survived.  Furthermore, the complete de-privileging of "authorial authenticity" was counter-intuitive to all but the most hardened deconstructionists when applied to texts whose authors were known—not anonymous—or where scribal "slips" were transparently unintentional.  However, the New Philologists’ emphasis on the materiality of the text has proved to have lasting influence, shifting our focus towards understanding the contexts in which texts (and not just the originals) were produced, used, and reproduced.  It has also generated substantial new interest in medieval paleographical and codicological practices and neglected forms of medieval textuality such as glossing and marginalia.  Thus even though the revolution proclaimed in the 1990 Speculum article has not come to pass, "New Philology" did shift the emphasis of scholarship and helped to open up new approaches to medieval literature.  The "material philology" that has emerged in recent years goes a long way towards synthesizing the tools and methods of "old philology" with those of "New Philology", and out of that synthesis creating valuable new knowledge about medieval texts and cultures.

§6.  So what is "material philology", and how does it relate to older philological methods?  We note a number of general tendencies in a material philologist's approach:

§7.  We believe that the purpose of philology in this approach is still fundamentally archeological and consistent with the goals of the "old philology," which we would summarize as using a deep knowledge of language and language change to recover information about past cultures.  "Material philology," instead of focusing only on the long-lost cultures of which medieval manuscript texts are but later "ruins," seeks to improve our understanding of those slightly more recent cultures which produced the actual manuscripts we now possess.

§8.  One of the areas where the material turn in philology has been particularly fruitful is in the study of the period following the Norman Conquest, during the periods of "transitional" and early Middle English.  Literary output in English was limited between the Conquest and the middle of the thirteenth century, and much of what was produced consisted of copies of Old English texts.  These were often of little interest to Old English scholars because they were linguistically "corrupt" versions of earlier Anglo-Saxon productions.  Alternatively, they were treated out of context as surviving witnesses to the textual cultures of the pre-Conquest period.  Elaine Treharne points out the problematic nature of this approach in a discussion of the surviving manuscripts from Worcester:

Some of the manuscripts, such as Hatton 113 and 114 and Junius 121, are the major witnesses to texts composed earlier; in this instance, it is Wulfstan's works manifested in this codex that make it such an important testimony to the dissemination of his homiletic and legal tracts; it is this manuscript that forms the basis for Bethurum's and Napier's critical editions.  Of less interest to scholars, though, is the actual milieu of this manuscript and its contemporaries.  That is, the manuscripts copied in the second half of the eleventh century are investigated less for the production and use of Old English in its precise time, and more for what they tell us about the texts' authors within those authors' own time.  This raises important questions about the manner in which scholars treat the evidence that is extant for a study of vernacular literature: that again, we privilege the origins rather than the duplication and adaptation of material to the extent that we skew the data, obfuscating particular cultural, political, and historical contexts and undermining the role of literature itself (Treharne 2007, 19–20).

§9.  To this it should be added that the relatively small numbers of texts from this period, their Anglo-Saxon origins, and the frequently didactic or religious nature of their content discouraged Middle English scholars from giving them much consideration.  Recently, however, the period has come under greater scrutiny, and evidence is emerging for a rich vernacular literary culture in England during the period.  In this column we will focus on the literary communities of the West Midlands through the work of a thirteenth-century scribe known as the Tremulous Hand of Worcester.  The glosses of this scribe have been extensively studied by Christine Franzen, and we will examine the methods of her most recent publication here (Franzen 2007, 13–31).  However, we will also try to bring Franzen's findings together with recent work by a number of other scholars, some working in more traditional philological paradigms and others working from a more "materialist" perspective.  We hope to show that the combination of the "old" and the "new" yields some truly astonishing results.

§10.  The Tremulous Hand of Worcester is known from his shaky hand, apparently the result of a congenital condition, which grew worse over the course of his career as a scribe.4  Examples of his hand survive in glosses found in over twenty Anglo-Saxon and post-Conquest manuscripts, often glossing the same manuscripts in layers over a period of time.  As many of his glosses appear to make sense of, or update the language of, Old English words, he is an important witness to the continued readership of Old English in Worcester as late as the thirteenth century.  Marginal annotations and nota signs suggest that his interest was in the content of the texts he read as well as their language.  He also wrote down a number of more sustained texts: Worcester Cathedral, MS F.174 contains a copy of Ælfric's Grammar and Glossary, the so-called Worcester Fragments (consisting of a lament for Bede and The Soul's Address to the Body); and Bodleian Library, MS Junius 121 contains a copy of the Nicene Creed in his hand.  Some manuscripts also contain fragments of worksheets containing English-Latin word pairs based on his glosses.

§11.  Although the Tremulous Hand's work collectively amounts to a significant amount of text, its survival largely as disparate marginalia has meant that this output has rarely received sustained attention.  But Franzen shows in her recent work that we can learn a great deal about the literary culture he was working in from this material.  She notices that the earliest layer of glossing by the Tremulous Hand shows marked linguistic and palaeographical similarities to the work of the scribe of London, British Library, MS Cotton Nero A.xiv, which contains a thirteenth-century copy of Ancrene Wisse, or Guide for Anchoresses.  This text was extensively copied in the Middle Ages, was a key to the study of Middle English dialects in the early part of the twentieth century, and has in recent years been the subject of considerable scrutiny by scholars, especially those interested in female readership.  Franzen concludes from the resemblances between the Tremulous Hand and the text of the Nero manuscript that the two scribes must have had some sort of common training.

§12.  In keeping with our column's goal of demonstrating how scholars "do" philology, we will now provide a simplified explanation of the data and methods Franzen uses to achieve her insight.  Franzen observes that the earliest stage of the Tremulous Hand's glosses and marks show little evidence of his characteristic tremor.  She calls this layer the D (dark and neat) layer.  This earliest stage is found in only three manuscripts: Bodleian Library, MS Hatton 115, and parts of Bodleian Library MSS, Hatton 113 and Junius 121.5  Most of the D layer material is in English and consists of updated spellings, inflections, punctuation and vocabulary, which Franzen thinks was intended to prepare exemplars for copying.  However, at some point the Tremulous Hand began glossing the texts in Latin.  There are over 5,000 Latin glosses from this stage, which Franzen calls the B layer since the writing still has a "bold" appearance while beginning to show signs of trembling.  The last stage of glossing she calls the M (mature) layer, which shows the most trembling and from its position indicates that the B layer glosses were already in place.  There are around 3,200 vernacular glosses (mainly in English with a few in French), nearly half from the D layer.  The Tremulous Hand's method appears to have been to copy words he found puzzling into the margin, apparently to call attention to them.  As he did so, he occasionally updated the spellings (e.g. <tilig> for tylig, <muchele> for mycele).6  In other cases, the Tremulous Hand followed the Old English word he put in the margin with a Latin gloss (e.g. <athæfdon> retinerent for æthæfdon), which leads Franzen to believe that he was attempting to collect equivalents for some sort of glossary.  Where he substituted new words, it was usually because the Old English word was obsolescent: <euer> for a; <lawe> for æ; <ischeawede> for æteowde; <seið> for cwyð; <pine> for wite; <heste> for bebod; <þenchen> for smeagan; <glorie> for wuldor (the last an example of a French loanword).  Franzen notes that "one of the difficulties with his vernacular glosses, then, is that they represent at least two completely different facets of his work: some are early Middle English or French lexical substitutions for obsolescent Old English words and represent up-to-date, thirteenth-century vocabulary and spelling, while others may be indications of puzzlement, perhaps partially re-spelled obsolete or obsolescent Old English words" (Franzen 2007, 15).

§13.  Earlier assessments of the Tremulous Hand's understanding of Old English and his objectives in glossing it have tended to see his work as archaistic, but they have notably relied on the later sustained texts which belong to the M layer.  Thus Jeremy Smith, comparing the Worcester Fragments and the Nero manuscript of Ancrene Wisse, noted the former's heavy use of the typically Old English <æ> and <eo> spellings—even in words where they did not occur in Old English.  He concluded that the Tremulous Hand consciously wished to reproduce Anglo-Saxon orthography, whereas the Nero scribe "converts the language of his exemplar into a language which he, and presumably his local readership, found easier to understand" (Smith 1991, 56f).  Franzen, however, argues that the similarities between the D layer of Tremulous Hand glossing and the Nero text of Ancrene Wisse work against such a simple comparison.  In his earliest work, it appears that the Tremulous Hand was working to match the language in his glosses to that of the spoken language at the time he was writing them.  Later in his career, and when he wrote the Worcester Fragments, he appears to have used much more archaic forms.  Franzen observes that we cannot know the true extent of his archaism until we acquire an understanding of what the Tremulous Hand's early Middle English dialect was like, an understanding which would allow us "to separate out the old from the new" (Franzen 2007, 17).  As we will see, this rather traditional philological undertaking is what reveals the similarities between the Tremulous Hand's dialect and that of the Nero scribe, as well as the startling changes in the former's discursive practices and relationship to his Old English exemplars over the course of his career.

§14.  Franzen assembles an impressive array of data in order to arrive at this point, and we can only present a small and simplified portion of it here.  We have attempted to pick out some of the most significant items and present them in a way which illustrates traditional philological methods and tools with a bare minimum of examples and technical terminology.  Philologists typically begin to describe a dialect in terms of its phonology, or sound system.  One of the most significant developments in the transition between Old English and early Middle English is the change of Old English sound æ to a, which is then reflected in the spelling (so Old English cæt becomes cat).  The glosses of the Tremulous Hand clearly demonstrate this development, but not uniformly.

Spelling of OE æ D layer B and M layers
a 18 42
æ 2 74 (16 unetymological)

By "unetymological" we mean that the Old English word did not have æ in the place where the Tremulous Hand uses that spelling.  Not only do the B and M layers preserve Old English æ almost three-quarters of the time, but almost a fifth of its spellings appear to be attempts at "archaic" spellings with that letter.  The D layer by contrast, preserves Old English æ only twice and contains no unetymological spellings.

§15.  This pattern, in varying proportions, is reflected through the grammar and vocabulary of the Tremulous Hand's work.  A more complicated example can be found in the morphology, or system of grammatical inflections, in the Tremulous Hand's dialect.  Old English possessed a rich array of verb endings which were heavily levelled in early Middle English.  For instance, the infinitive endings –an and –ian appear most often as –(i)en in early Middle English.7  In the Tremulous Hand's glosses, we see a variety of forms:

Spelling of OE Infinitive –an, -ian D layer B and M layers
-en 44 65
-ien 2 65
-ene 3 2
-n 3 4
-an 12 (-on 3 times)
-ian (-onne 1 time)

§16.  Here again, we see that the D layer contrasts with the B and M layers in avoiding archaistic spellings.  Other forms of the verb display similar patterns.  For instance, the third person singular present tense ending was generally – or – in Old English.  In early Middle English, the "back vowels" a, o, and u (those formed at the back of the mouth) merged in unstressed positions to a neutral sound, spelled e.  Hence both Old English inflections appear most often as –.8

Spelling of OE 3rd sg. pres. D layer B and M layers
-eð 23 23
-að 10
-æð 1

The merger between a, o, and u explains the infinitive forms –on(ne) for Old English –(i)an, and the change of Old English æ to a explains the unetymological but archaistic form –æð in the table above.

§17.  In addition to the phonological and morphological changes which took place in the spoken language between Old English and Middle English, there were also numerous changes in orthographic habits (often influenced by written French) which did not necessarily reflect differences in pronunciation.  Here are some typical examples:

Old English Spelling Tremulous Hand Spelling D layer B and M layers

sc sch 22 12
sc 11
ch 1

c c(c)h 54 42
c 23

cw cw 6 1
qu 8 1
cqu 1

f- (initial) f- 44 102
v- 20 10
u- 3

-f- (medial) -u- 32 25
-f- 14

§18.  New spellings for the common palatal sounds (Old English sc and c) are used consistently in the D layer, while over a third of the spellings in the B and M layers use archaic Old English spellings.  Although the number of forms of OE cw is not statistically significant for the B and M layers, it is noticeable that the D layer's slight preference for the use of Latin qu for cw fits this pattern.  Over half the D layer's spellings of initial f- use u or v, whereas this figure is only about 10% for the B and M layers.  Medially, the D layer always has -u-, whereas the B and M layers preserve the Old English spelling over half the time.  Overall, the pattern is clear.  Of the 189 D-layer forms almost three quarters are up-to-date spellings.  The figure is just over a third for the B and M layers, suggesting a much greater effort to reproduce the Old English system of orthography at this later date.

§19.  Franzen argues on the basis of this and other data that the scribe "had a clear and consistent sense of how to represent his own early Middle English dialect in writing, and his scribal habits at this earliest stage, the D layer, would have to be described as translation.  I can only assume he had been taught to write the vernacular or else had a very clear idea in his own mind of how to represent it."  His later copied texts, whether consciously so or not, "are linguistically conservative in spelling, inflections, and vocabulary, while his earliest English glosses, those from the D layer, show him to be linguistically progressive" (Franzen 2007, 19).

§20.  It is clear then, that the archaizing tendencies Smith finds in the Worcester Fragments and Franzen finds in the Tremulous Hand's later glosses obscure any resemblances to the more "progressive" language of the scribe of the Nero Ancrene Wisse.  However, evidence for similarities in the dialects of the two scribes should be apparent in the earlier D layer spellings.  This in fact proves to be the case.  The two scribes share a number of dialect features common in the West Midlands.  We give a few selected by Franzen below:

West Midland Dialect Features Common to the Tremulous Hand and the Nero Scribe

Old English Sound Tremulous Hand and Nero Spellings
a before nasals <o>: 'mon'
a before ȝ <aw>: 'lawe'
a before l + consonant <a>: 'alle'
a before ld <o>: 'tolde'
æ (Mercian e)9 <e>: 'eppel', 'efter', 'feste'
w + æ (Mercian e) <a>: 'was', 'iwat'
Anglian e by smoothing (West Saxon ea) <e>: 'werke'
y <u>: 'muchel', 'dude', 'sunfule'
long a <o>: 'holi'
long a before ȝ <o>: 'owene'
long a before ht <ou>: 'nouht'
short and long o <o>: 'iwrouht'
long y <u>: swuðe

§21.  These features are typical West Midland ones and are to be expected from writers in the region.  However, Franzen also looks for unusual words or spellings.  Although she finds that the Nero scribe and the D layer agree on virtually everything for which there is evidence in the latter, their agreement was not always exact.  The following examples, reorganized and much condensed from the data Franzen gives, illustrate the similarities and differences between the Nero and Tremulous scribes.

  1. Nero/D elde.  The Old English word for "old" was eald or ald.  In late Old English ea > a (by way of æ) and the vowel was lengthened before l + a consonant.  Long a became long o around 1250 (as in tolde), so the spelling of elde with an e is surprising.  That both scribes use it suggests that they were closely related in dialect and/or training.
  2. Nero/D eihte "eight," beih "bowed."  eie, eien "eye, eyes."  These words had ea or e in Old English: eahta, béah, éage, éagan.  The i spelling could represent a palatalisation of the velar ȝ between vowels (a development which took place earlier in Old English) in words like eie, but its transference to words like beih and eihte would have to be based on analogy between the sounds of the vowel before velar consonants.  That the Nero and Tremulous scribes both do this suggests a similar training.
  3. Forms of siggen 'say': Nero -i-, D -i- or -e-, Soul and Body -i- or -e-.  The Old English form was secgan.  The development of e > i before a velar consonant is common in the West Midlands, shared by both scribes.  However, the Tremulous Hand has a tendency to preserve the older vowel, possibly an archaistic scribal practice.
  4. Old English short eo before consonant groups or Old English long eo (e.g. feor, eorð, seolke, seolc, créopen, féond).  Nero and D regularly spell this vowel <eo>.  However, both scribes spell the vowel in Mercian Old English seolf "self" with <u> (Nero sulf, suluen, sulue; Soul sulfen, suluen), which may derive from the West Saxon form sylf.  Since they do not reproduce the <y> of the West Saxon, their spelling probably represents a common pronunciation, rather than an imitation of Old English spelling conventions.
  5. Both Nero and D use the <eo> spelling in a variety of unetymological locations.  In forms of wepen "weep," Nero uses <eo> for OE long e in all cases but one.  B has weopen beside wepinde.  Soul has weopinde.  For Old English néd "need," Nero uses both nede(e) and neode, but TH seems to favour eo (29x) over D nede (2x), M node (1x).  Both Nero and D regularly use <eo> to represent OE o in words derived from OE sorh "sorrow," wolcen "sky."  Both scribes appear to be over-using Old English spelling conventions.
  6. Mercian long æ before a labial is spelled <ea> by both Nero and D (a spelling also found in Old English): eaubruche "breach of the law" (Old English æwbryce).  This was apparently a lax e sound, equivalent to that which had developed from Old English long ea.  This sound is spelt inconsistently by both scribes: Nero edmod(e), B edmod "humble" (Old English eadmod); Nero bileaue "faith" (Old English geléafa), D vnbileaue.  After palatal consonants this sound was often retracted to a, but Nero and D show this only twice: from Old English scéawian we get Nero schawe "see" (1x), D ischawed "seen."  The <ea> spellings are characteristic of the so-called AB language of the West Midlands.10

§22.  The same can be said of Tremulous Hand's lexical choices when he substitutes more up-to-date words for obsolescent Old English ones.  They are frequently (but not always) items found in the Nero scribe's vocabulary.  Here are some examples:

  1. euer: always for a(a) by the Nero scribe, in the D layer.
  2. kecchen: regularly substituted for lecchen by the Nero scribe; the D layer has kecchen for gelæccan.
  3. nimen: the Nero scribe uses this form where other manuscripts have taken.  The later form is never used in the Tremulous Hand's glosses.
  4. wið: Nero substitutes wið for Old English mid "with."  In the D layer wið is glossed using Latin contra, but glosses Old English bearneacnigendum with "mid childe."

§23.  Finally, the hands of two scribes share certain paleographical features:

  1. Both the Nero scribe and the D layer have a very short descender on the letter ƿ (wynn, the Old English w) which curves slightly to the left at the bottom.
  2. They use an identical form of the insular g, written on the line, with a nearly straight top stroke.
  3. Both have very straight minims (the vertical strokes in i, n, m, and u) and vertical strokes in þ, l, h, b.

§24.  Franzen argues that the similarities between the two scribes are sufficient to make a case that they moved within a common literary community and had a similar training (she does not suggest that there is only one scribe).  Note that we can only see this relationship between the scribes through a close philological analysis of their work and that this analysis allows us to then reconstruct more of their cultural context.

§25.  Although Franzen's systematic look at the phonology, morphology, orthography, and lexis of the Tremulous Hand is entirely traditional, in certain important ways it is shaped by the arguments of the "New Philology."  To begin with, it takes two scribes, neither of whom were the originators of their texts, and shows how they engaged in authorial acts and discursive choices in their use of language.  The evidence is drawn not from idealised, edited texts but from the manuscripts themselves, and it is their linguistic idiosyncrasies which provide the important insights: that there was a well developed sense of how to write English in the West Midlands during the early thirteenth century, but that at least one writer felt able to depart from it for specific purposes, presumably relating to his subject matter.  Although Franzen does not focus on the Nero scribe, this scribe's copy of Ancrene Wisse forms part of a matrix of exemplars of a text which has been cited as ideal for the type of variation—study the New Philologists advocate.11  Nevertheless, we should point out that Franzen's conclusions run counter to some of the more extreme statements of New Philological theory.  In shedding light on the career and literary choices of the Tremulous Hand, Franzen builds him more substantially into an original "author."  More importantly, she reveals as a rhetorical overstatement the New Philologists' claim that medieval vernaculars lacked a codified system of writing.  In the Tremulous Hand, we have a scribe who was in fact caught between two literary codes.  This is a point to which we will return.

§26.  What Franzen leaves relatively unexplored is the question of why the Tremulous Hand chose to use archaic language in some of his writings and not others.  In the case of the Worcester Fragments, it is not hard to guess.  They are in essence laments for the passing of Old English literary greats.  But why he chose more archaic spellings in his later glosses is a greater mystery.  She notes in passing that many of the texts he glosses were homilies, and her observation that the Tremulous Hand looks as if he was preparing texts for copying provides an important inroad to understanding his scribal practice.  In order to understand this practice more fully, we can compare the work of the Tremulous Hand to the archaizing tendencies found in other collections of homilies during the early Middle English period.  We will briefly turn, therefore, to a recent study by Bella Millett, who, like Elaine Treharne, is explicitly influenced by the material turn in the New Philology with its emphasis on manuscript variation.  Millett has recently attempted to identify contexts for two collections of homilies which are roughly contemporary with the Tremulous Hand: the "Trinity Homilies" found in Cambridge, Trinity College, MS B. 14. 52 (335) and the "Lambeth Homilies" in London, Lambeth Palace, MS 487.  Scholarly interest in these texts has been limited by the same ideological conditions that have affected other early Middle English texts, as we discussed above.  As Millett points out, "what work has been done has focused mainly on their relationship to Old English preaching (demonstrable in the case of the Lambeth collection, which includes material from Ælfric and Wulfstan I, although harder to establish for the Trinity sermons), and both collections have been characterized as stylistically backward-looking, the product of a period when a 'renewed emphasis on preaching' in English 'tended to find expression in harking back to older homiliaries'" (Millett 2007, 43–44).

§27.  Here we have a prima facie context for the Tremulous Hand's archaizing glosses: the perpetuating of a glorious tradition of vernacular sermon-making through the copying of old texts and, in some cases, the creation of new ones in the old vein.  But let us not forget our earlier observation that the Tremulous Hand was caught between two linguistic codes, and that he increased his archaising tendencies over the course of his career, sometimes when he returned to texts he had already glossed.  Perhaps he was becoming more "backward-looking," "antiquarian," or "nostalgic" in his attitude towards the sermons.  Elaine Treharne has remarked that such descriptions of the compilers of the Trinity and Lambeth Homilies is misleading.  Both collections contain material which reflects the "changed theological and pastoral context" of the post-Conquest period (Millett 2007).  Millett attempts to describe this context in more detail through a comparison of the variations between the Trinity and Lambeth Homilies.  A brief look at her evidence helps to shed light on the questions invoked by Franzen's study of the Tremulous Hand of Worcester.

§28.  Lambeth 487 was produced in the West Midlands sometime between the very late twelfth century and 1225, and Trinity B. 14. 52 was made in the East Midlands slightly earlier.  Millett identifies this as a period of "radical change in preaching and pastoral care," in which the "scholastic" style of preaching in the Paris schools converged with reforms designed to "improve the morals and education of the secular clergy" (Millett 2007, 45).  Canons from the Third and Fourth Lateran Councils were designed to strengthen pastoral care, and a growing number of preaching aids began to appear during this period.  Stylistically, the "scholastic" method of preaching made increasing use of schematic division and subdivision, a feature which is to be found in a number of sermons shared between both the Trinity and Lambeth collections.  In particular, "the use of divisiones (usually biblical distinctiones) to structure the sermon as a whole rather than simply the subdivisions of its argument," is a sign that some of these sermons were composed at quite a late date (Millett 2007, 48).  The number of points within a divisio also increases over time, a phenomenon which Millett observes in the sixteen-point list of "conditions of confession" in Ancrene Wisse (Millett 2007, 49).  Following up on such phenomena, Millett concludes that five of the shared sermons are likely to be late compositions, perhaps based on twelfth century Latin models of contemporary developments in preaching.

§29.  In an attempt to characterize the environment in which these sermons were produced, Millett finds little internal evidence for their use in a pastoral context.  A few references, such as Trinity 11's statement about the expulsion of penitents at the beginning of Lent by "ech bishup þe in his minister beð" (each bishop who is in his minster), suggest that it might have been a monastic, cathedral, or collegiate church.  The intended audience seems to have been mixed, sometimes indicating the clergy, sometimes the laity, or both.  The address to "leoue broðre and sustre" (dear brothers and sisters) in Lambeth 1 and 2 seems to indicate a partially female audience.  External evidence about pastoral care in thirteenth century England suggests that even these communities included considerably diverse groups of people of all levels of education, to whom the sermons could have been addressed.  According to Millett, "a diocesan pastoral context...might account for the diversity of the sermons of the Trinity and Lambeth collections; and its more cosmopolitan cultural milieu would help to explain their incorporation of both recent developments in continental preaching and (in the case of Lambeth 487) preaching inherited from an older native tradition" (Millett 2007, 60).  Millett argues that this last possibility links these collections of homilies with the Ancrene Wisse group of texts which expanded after 1215 in the context of continuing clerical reforms.  She points out that "the Trinity collection, the Lambeth collection, and the works of the Ancrene Wisse group could be seen as "marking consecutive, and to some extent connected, stages of a revival of vernacular religious prose produced by a broader English movement of pastoral reform, dating back at least to the Third Lateran Council of 1179" (Millett 2007, 61).

§30.  We may at this point return to Franzen's own assessment of the implications of her analysis of the language of the Tremulous Hand and Nero scribes.  She suggests that

...these two scribes, writing a consistent and nearly identical form of early West Midlands Middle English using a nearly identical script, may have had a common training in the writing of English.  That is, somewhere in Worcestershire (probably in Worcester itself) in the first half of the thirteenth century (very likely after 1215 and possibly in the second quarter), there may have been a centre for the production of vernacular manuscripts in which scribes were trained to produce, perhaps among other things, up-to-date English books in the local dialect from older, and sometimes much older, English material.  The most intriguing aspect of this is that two scribes who may have been trained in such a centre seem to be dealing with English material from very different sources and religious milieux and for very different audiences.  Whether these two scribes produced their surviving work while at such a centre or elsewhere, having moved on to join, for example, the bishop's household or the Benedictine priory at Worcester, is not clear.  Nor is it clear what religious affiliation, if any, such a centre may have had.  But further work on the dialects of these scribes and on book production in the West Midlands in the thirteenth century may help to answer some of these questions (Franzen 2007, 28).

§31.  Let us consider some of the similarities in this assessment to Millett's conclusions about the Trinity and (particularly) Lambeth Homilies.  We have a somewhat mobile authorship which caters flexibly to a diverse audience, reproducing and supplementing older material.  These "authors" were trained not only to write a contemporary form of Middle English but also to adapt and update pre-Conquest texts according to the requirements of their audiences and the latest developments in pastoral care.  Such a "training centre" implies that there was considerably more use of, and interest in, vernacular texts—both Old English and new ones—than the surviving evidence has indicated to date.  More importantly, the textual culture which begins to emerge requires us to re-assess our views about the relationship of early Middle English writers to pre-Conquest literature.  In this milieu, the Tremulous Hand appears to be less of an aberration—a lone antiquarian working at a time when Old English texts were mostly unintelligible—than may at first appear.  A materially philological approach thus helps us to understand the cultural environment which produced the Tremulous Hand and the only other individual "antiquarian" from the period, Laȝamon, who adapted Wace's Roman de Brut into English.

§32.  Franzen's final words highlight the important role that philology must have in shedding further light on the vernacular literary milieu of the transition from Old English and of the early Middle English period.  It is a dual role, in which scribal dialect and the history of book production are integral parts.  This, we believe, highlights the perfect integration of the "old" philology with the "New", and the resultant synthesis of a "material philology" that has the potential to make significant contributions to medieval studies.  But for "material philology" to be more than just looking at manuscripts, it requires the hard-won knowledge base of the "old" philology, and no ideological critique or shift in emphasis can make those methods any less essential.  Thus it is vitally important that philological knowledge be preserved, and that a new generation of scholars be trained in the painstaking and complex methods of traditional philology.  Otherwise, brilliant works like Franzen's will become merely the aberrations of lonely antiquarians working in a dead tradition, with the majority of living scholars unable even to understand, much less analyze or critique, the knowledge about lost culture that has been recovered.


1.   Influential works by Paul Zumthor are Essai de poétique médiévale (1972), in which he first introduced his concept of textual mobility, or mouvance, and La lettre et la voix: De la 'littérature' médiévale, Collection Poétique (1987).  Much of the theoretical basis for the "New Philology" was laid out by Bernard Cerquiglini in his short "critical history of philology," Éloge de la variante (1989), translated by Betsy Wing as In Praise of the Variant (1999).  [Back]

2.   Cerquiglini 1989, 105–108.  Our presentation of his material here is adapted from and partially inspired by Suzanne Fleischman's discussion in "Philology, Linguistics, and the Discourse of the Medieval Text" (1990, 23).  [Back]

3.   It must be acknowledged that there may have been significant differences in the practices followed by scholars in the English and French vernaculars.  Whereas the New Philologists were primarily scholars of French literature, our perspective is based on that of English.  Contemporary Anglo-Norman scholarship, where French and English scholarly traditions intersect, very often has to rely on antiquated and reductive editorial practices.  [Back]

4.   This career has not been dated more precisely than sometime between the late twelfth century and 1250.  Franzen argues that the degeneration of his hand, if caused by a congenital tremor, might have occurred in as few as five or ten years (2007, 7).  [Back]

5.   These, it should be noted, are the manuscripts cited by Treharne above.  [Back]

6.   For clarity, we depart from Franzen's representation of the glosses and glossed items ('tylig': 'tilig'), using angular brackets for the Tremulous Hand's spelling and italics for that of the glossed form.  [Back]

7.   In Old English, an additional inflection was sometimes added to the infinitive –an when the word followed the particle to, e.g. to nimanne (to take), and the Tremulous Hand's spellings sometimes reflect this.  [Back]

8.   The inflection also occurs with þ in place of ð in both Old English and Middle English, and we have merged these spellings in the figures above.  Under certain conditions, the vowel was lost in Old English, producing endings such as –ð, -þ, and –t.  Since such spellings in the Tremulous Hand's glosses are not noticeably archaic, we omit them from our reproduction of Franzen's data.  [Back]

9.   The Old English Anglian dialect region consisted of Northumbrian in the north and Mercian in the west.  The West Midlands fell within the Mercian area of Old English.  Since most Old English was written, to varying degrees, in the southern West Saxon dialect, this is the most likely model for linguistic archaizing in early Middle English.  [Back]

10.   The AB language was a standardized literary dialect of the West Midlands, represented by texts found in MSS Cambridge, Corpus Christi College, 402 (A) and Oxford, Bodleian Library, Bodley 34 (B).  The consistent spellings in the AB texts demonstrate a continuity in writing traditions from the West Midlands from the Anglo-Saxon period until the thirteenth century.  For discussion, see Tolkien 1929, 104–26.  [Back]

11.   See Bella Millett's online essay "What is mouvance" (  Millett's electronic edition of Ancrene Wisse attempts to account for this by establishing "a context for the edited text within the broader historical development of the work" (  [Back]

Works Cited

Cerquiglini Bernard, 1989.  Éloge de la variante.  Paris: Cerf.  [Back]

Fleischman Suzanne.  1990.  Philology, linguistics, and the discourse of the medieval text. Speculum 65:19–37.  [Back]

———.  2007.  The Tremulous Hand of Worcester and the Nero scribe of the Ancrene Wisse. Medium Ævum 72:13–31.  [Back]

Millett, Bella.  2007.  The pastoral context of the Trinity and Lambeth Homilies. In Essays in manuscript geography: vernacular manuscripts of the English West Midlands from the Conquest to the sixteenth century, ed. Wendy Scase. Turnhout: Brepols.  [Back]

———.  2008.  What is mouvance  [Back]

Smith, Jeremy J.  1991.  Tradition and innovation in south-west-Midland Middle English.  In Regionalism in late mediaeval manuscripts and texts: essays celebrating the publication of 'A linguistic atlas of late mediaeval English,' ed. Felicity Riddy.  Woodbridge: Boydell.  [Back]

Tolkien, J. R. R.  1929.  Ancrene Wisse and Hali MeiðhadEssays and Studies of the English Association 14:104–26.  [Back]

Treharne, Elaine.  2007.  Bishops and their texts in the later eleventh century: Worcester and Exeter. In Essays in manuscript geography: vernacular manuscripts of the English West Midlands from the Conquest to the sixteenth century, ed. Wendy Scase. Turnhout: Brepols.  [Back]

Zumthor, Paul.  1972.  Essai de poétique médiévale.  Collection Poétique. Paris: Seuil.  [Back]

———.  1987.  La lettre et la voix: de la 'littérature' médiévale.  Collection Poétique.  Paris: Seuil.  [Back]