On Saying Yes in Cotton Vitellius C v:

A Case of Final Vowel Elision in Old English?


Phillip Wallage and Wim van der Wurff

Northumbria University and Newcastle University


The question.

This note concerns one word in one manuscript version of an Old English text — the Homily on the Second Sunday of Lent that is part of the second series of Ælfric's Catholic Homilies. In the Dictionary of Old English, manuscript BL Cotton Vitellius C v is shown as having, near the start of the homily, the following exchange between Christ and the Canaanite woman of Matthew 15:21-8:1

    (1) he hyre & wyrde; nis na <god þæt man nyme his bearna hlaf>
    he her answered not-is not good that one takes his children's bread
    & awyrpe hundum; ðat wif & wyrde. gys leof drihten. swa þa hwelpas
    and throws to-dogs the woman answered yes dear Lord and even the pups
    etað of þam crumum þe feallað of heora hlafordes mysan
    eat of the crumbs that fall from their Lord's table
    ("He answered her, 'It is not good that one should take one's children's bread and throw it to the dogs.' The woman answered, 'Yes, dear Lord, the pups even eat the crumbs that fall from their Lord's table.'")

The relevant word in this passage is gys, the forebear of present-day yes. The problem with it is simple: the two existing editions of this homily do not agree on its exact form. The EETS edition by Malcolm Godden, on which (1) is based, specifies that Cotton Vitellius C v has the reading gys in this passage.2 However, a separate edition of the entire manuscript Vitellius C v by Winifred Temple gives the reading gyse.3 The question we explore is therefore the following: did the scribe of Cotton Vitellius C v intend the Canaanite woman to say gys or gyse at this point, i.e. which is the correct reading of the manuscript?


2. Why do we ask?

One reason for asking about gys(e) is that the twelve further tokens of this word that occur in Old English all end in –e.4 If that was the historical form, then the word has undergone final vowel loss in the development leading to present-day yes. In itself, there is nothing surprising about this. Elision of final unstressed –e is attested in many other words and is in fact one of the major phonological changes in the history of the language. However, the handbooks firmly and unanimously date the loss of –e to the Middle English period.5 The only exception is when unstressed –e occurs before a syllable that is itself unstressed and starts with a vowel. In such cases, there is occasional elision already in Old English, as in examples like sægdic for sægde ic 'said I'.6 The form gys(e) in Vitellius C v is followed by a stressed syllable starting with a consonant, so the reading gys would make it a very early instance of generalised vowel elision while the reading gyse would be unremarkable. Hence, the correct reading of this word matters for accurately dating the unfolding of an important process in the historical phonology of English.

A further reason for asking is that the presence or absence of a final vowel in the word gys(e) could have implications for our understanding of the origins of the word. In a recent paper, we have discussed the etymology of yes, suggesting that it derives from the three-word sequence gea is swa 'yes, it is so'.7 At some point in proto-Old English, we propose, this expression became routinised as a positive polarity response and, due to its resulting high frequency, fused into a single word, *geaisswa; this form then underwent the regular Old English sound changes to yield the attested forms. In that article, we argue that this hypothesis is to be preferred over the two other main ideas for the origin of yes found in the literature, i.e. that it goes back to a two-word sequence consisting of gea 'yes' and either swa 'so' or sie, the present subjunctive of the verb be. Regardless of the correctness of these different views, the possible existence of a form in Old English in which the final vowel is absent has obvious repercussions for the debate about the etymology of yes. It could mean that final vowel loss in this word in Cotton Vitellius C v was indeed a very early instance of generalised elision (which, though not completely surprising given the word's likely high frequency, would make it a unique case). But it could also mean that the source of yes has to be sought in a (sequence of) word(s) ending in s and that the final –e forms are due to a secondary development (which would entail that all existing proposals have to be rejected). Either way, it is worth examining the evidence for the word in Cotton Vitellius C v somewhat more carefully.

A final reason for asking is simple curiosity. Both editions tell us that Cotton Vitellius C v has damage at the edges of leaves due to the 1731 fire in the Cotton library.8 Does the form gys(e) occur at one of the damaged places? If so, where exactly is the damage and does it preclude a confident reading or is enough still visible to arrive at some kind of certainty? The passage in (1) is repeated further on in the same homily and both editions show Cotton Vitellius C v as having the form gyse there, which of course adds to the puzzle.


3. The answer

It turns out that the word gys(e) in (1) is the first word in the top line of fol. 72r of Cotton Vitellius C v and that the first four lines of text on this leaf show the effects of burning and scorching. Prolonged visual examination of the manuscript, with the aid of a magnifying glass, reveals a distinct sequence <Gẏ>, with a following letter that is recognisably long s.9 After this, all that can be seen is an indistinct dark spot, of about the width of the ordinary word-spacing on the parts of this page that have not been affected by the action of fire, followed by the clearly legible <leof drihten>. No <e> is discernible between <Gẏf> and <leof drihten>. In spite of the problems caused by fire, inspection of the manuscript therefore appears to confirm the reading gys that is given in the EETS edition.10 The reading gyse given by Temple must then be an error, perhaps induced by knowledge of the usual spelling of the word.11

However, some niggling doubt remains, caused by the dark spot in the manuscript and by the use, confirmed by visual examination, of the form with final –e further on in the same homily (fol. 72v). Following the lead of Carmen Acevedo Butcher, who used a fibre optic light cord to decipher several readings in Cotton Vitellius C v that had been covered by pasting paper in the nineteenth century,12 we decided to use modern technology in the hope that this might also shed light on the word gys(e). And it did. Examination of a high-resolution digital image of the first two words in the manuscript line in question, given in Fig. 1, shows that the correct reading is in fact gyse.


Fig. 1 Close-up of start of top line of Cotton Vitellius C v, fol. 72r
(©British Library Board; reproduced by permission of the British Library)


As can be seen in Fig. 1, there is a small hole in the manuscript immediately before the <l> of leof, into which two parts of the final –e of gyse have disappeared: a tiny fraction of the right-hand part of the closed counter and most of the lower part of the loop. The action of the fire has also caused the line to bulge upwards at this point, which does not help in visual examination. But the digital image is clear enough to conclude confidently that the scribe meant the Canaanite woman here to say not gys but gyse.


4. Implications

While the above demonstration makes only a very small point, there are several implications that follow from it. They can be listed as follows:

  • a. No modification is required to the standard view that final vowel elision started in pre-vocalic position in Old English and did not generalise to pre-consonantal position until the Middle English period;
  • b. The etymon of the word yes definitely ended in a vowel; if we suppose that the word originated in proto-Old English (as seems plausible, given the absence of cognates in other Germanic languages), a gradual process of final vowel reduction and loss can be assumed for it;
  • c. dark or illegible spots in manuscripts may repay closer investigation with the use of digital image technology;
  • d. A tiny corrigendum can be added to Godden's edition of Ælfric's homily for the Second Sunday of Lent, specifying that, in the passage given in (1), manuscript Cotton Vitellius C v, fol. 72r, has the reading gyse;
  • e. The spelling variant <gys> can be removed from the Dictionary of Old English s.v. gȳse;
  • f. The last two sentences in fn. 54 and the whole of fn. 71 in Wallage and van der Wurff, "On Saying 'Yes' in Early Anglo-Saxon England," should be scrapped.

Since we came to the question asked in this note from an interest in the origins of the word yes, our discovery of this problematic manuscript reading was entirely fortuitous. For editors and philologists, the issues involved may invite a more systematic approach and it is easy to see how facts of the type laid out above could be used as one argument for a large-scale coordinated programme of digitally imaging all the physical remains of the Old English language.

1. A. diPaolo Healey, J. Holland, D. Haines, D. McDougall, I. McDougall, N. Speirs and P. Thompson, The Dictionary of Old English on Microfiche: A-G (Toronto, 1994-2008), s.v. gȳse, 1b.

2. Ælfric's Catholic Homilies, Second Series, ed. M. Godden, EETS s.s. 5 (Oxford, 1979), p. 67. Note that, in its main text, this edition follows the seven other existing manuscripts for this homily in giving not gys but gea 'yeah/yes' (and also not ge furðon 'even' but swa ðeah 'nevertheless'). However, the readings of Vitellius C v are specified in the manuscript variant notes.

3. W.M. Temple, An Edition of the Old English Homilies in B.M. MS. Cotton Vitellius C. v, (3 volumes, Ph.D. dissertation, Edinburgh University, 1952; available from http://hdl.handle.net/1842/6686), p. 158. The homily is also printed in the earlier edition by Benjamin Thorpe (The Homilies of the Anglo-Saxon Church: The First Part, Containing the Sermones Catholici, or Homilies of Aelfric in the original Anglo-Saxon, with an English Version, volume 2, ed. B. Thorpe [London, 1846], pp. 111-17); but Thorpe did not collate the Vitellius manuscript and, anyway, omits the passage given in (1).

4. The attested forms are: gyse, gyese, gese, gise and ise. See Dictionary of Old English on Microfiche: A-G, s.v. gȳse.

5. An overview of the literature on this topic, and detailed discussion of the phonological processes involved, can be found in D. Minkova, The History of Final Vowels in English: The Sound of Muting (Berlin, 1991).

6. Further examples are given by Minkova, pp. 155-8. Some of the earliest instances seem to involve fixed expressions, like mid sac and mid socne for mid sace and mid socne 'with the right to receive fines and with the right to adjudicate in disputes'.

7. P. Wallage and W. van der Wurff, "On Saying 'Yes' in Early Anglo-Saxon England," Anglo-Saxon England 42 (2013): 183-215.

8. Godden, p. 67 and Temple, p. i. For a detailed description of the entire manuscript, see N.R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957), pp. 285-91.

9. We would like to thank staff at the British Library Manuscripts Reading Room, in particular Zoe Stansell, for making examination of the manuscript possible.

10. With kind help from Linda van Bergen and Edinburgh University Library staff, we were also able to inspect a microfiche facsimile of the manuscript (Homilies by Ælfric and other Homilies, ed. J. Wilcox, Anglo-Saxon Manuscripts in Microfiche Facsimile vol. 17 [Tempe, AZ, 2009]). This showed a dark blur in between <Gẏſ> and <leof drihten> so did not provide further clues.

11. It does not appear to be an editorial correction, given that Temple explicitly aims "to give a faithful reproduction of the MS. exactly as it stands" (p. i).

12. Carmen Acevedo Butcher, "Recovering Unique Ælfrician Texts Using the Fiber Optic Light Cord: Pope XVII in London, BL Cotton Vitellius C. v," Old English Newsletter 36.3 (2003): 13-22.